View source for PowerVR

{{Short description|Brand of graphics hardware and software}}
{{More citations needed|date=January 2021}}
'''PowerVR''' is a division of [[Imagination Technologies]] (formerly VideoLogic) that develops hardware and software for 2D and [[3D rendering]], and for [[video encoding]], [[video decoding|decoding]], associated [[image processing]] and [[DirectX]], [[OpenGL ES]], [[OpenVG]], and [[OpenCL]] acceleration. PowerVR also develops [[AI accelerator]]s called Neural Network Accelerator (NNA).

The PowerVR product line was originally introduced to compete in the desktop PC market for [[Graphics card|3D hardware accelerators]] with a product with a better [[price–performance ratio]] than existing products like those from [[3dfx Interactive]]. Rapid changes in that market, notably with the introduction of [[OpenGL]] and [[Direct3D]], led to rapid consolidation. PowerVR introduced new versions with [[low-power electronics]] that were aimed at the [[laptop computer]] market. Over time, this developed into a series of designs that could be incorporated into [[system-on-a-chip]] architectures suitable for [[handheld device]] use.

PowerVR accelerators are not manufactured by PowerVR, but instead their [[Semiconductor intellectual property core|IP blocks]] of [[integrated circuit]] designs and [[patent]]s are licensed to other companies, such as [[Texas Instruments]], [[Intel]], [[NEC Corporation|NEC]], [[BlackBerry]], [[Renesas Electronics|Renesas]], [[Samsung]], [[Sony]], [[STMicroelectronics]], [[Freescale]], [[Apple Inc.|Apple]],<ref>{{Cite web|url=https://www.imaginationtech.com/news/imagination-and-apple-sign-new-agreement/|title=Imagination and Apple Sign New Agreement - Imagination|work=Imagination |date=2 January 2020 |access-date=2022-09-03 |author1=93Digital }}</ref> [[NXP Semiconductors|NXP Semiconductors (formerly Philips Semiconductors)]], and many others.

== Technology ==
The PowerVR chipset uses a method of 3D rendering known as ''[[tile-based deferred rendering]]'' (often abbreviated as TBDR) which is tile-based rendering combined with PowerVR's proprietary method of Hidden Surface Removal (HSR) and Hierarchical Scheduling Technology (HST). As the polygon generating program feeds triangles to the PowerVR (driver), it stores them in memory in a [[triangle strip]] or an indexed format. Unlike other architectures, polygon rendering is (usually) not performed until all polygon information has been collated for the current [[Video frame|frame]]. Furthermore, the expensive operations of texturing and shading of pixels (or fragments) is delayed, whenever possible, until the visible surface at a pixel is determined — hence rendering is deferred.

In order to render, the display is split into rectangular sections in a grid pattern. Each section is known as a tile. Associated with each tile is a list of the triangles that visibly overlap that tile. Each tile is rendered in turn to produce the final image.

Tiles are rendered using a process similar to [[ray-casting]]. Rays are numerically simulated as if cast onto the triangles associated with the tile and a pixel is rendered from the triangle closest to the camera. The PowerVR hardware typically calculates the depths associated with each polygon for one tile row in 1 cycle.{{dubious|date=April 2013}}

This method has the advantage that, unlike a more traditional early Z rejection based hierarchical systems, no calculations need to be made to determine what a polygon looks like in an area where it is obscured by other geometry. It also allows for correct rendering of partially transparent polygons, independent of the order in which they are processed by the polygon producing application. (This capability was only implemented in Series 2 including Dreamcast and one MBX variant. It is generally not included for lack of API support and cost reasons.)
More importantly, as the rendering is limited to one tile at a time, the whole tile can be in fast on-chip memory, which is flushed to video memory before processing the next tile. Under normal circumstances, each tile is visited just once per frame.

PowerVR is a pioneer of tile based deferred rendering. Microsoft also conceptualized the idea with their abandoned [[Microsoft Talisman|Talisman]] project. Gigapixel, a company that developed IP for tile-based 3D graphics, was purchased by [[3dfx]], which in turn was subsequently purchased by [[Nvidia]]. Nvidia has now been shown to use tile rendering in the Maxwell and Pascal microarchitectures for a limited amount of geometry.<ref>{{Cite web|url=https://www.anandtech.com/show/10536/nvidia-maxwell-tile-rasterization-analysis|title=Hidden Secrets: Investigation Shows That NVIDIA GPUs Implement Tile Based Rasterization for Greater Efficiency|first=Ryan|last=Smith|website=www.anandtech.com}}</ref>

[[ARM Holdings|ARM]] began developing another major tile based architecture known as [[Mali (GPU)|Mali]] after their acquisition of [[Falanx]].

Intel uses a similar concept in their integrated graphics products. However, its method, called zone rendering, does not perform full [[hidden surface removal]] (HSR) and deferred texturing, therefore wasting fillrate and texture bandwidth on pixels that are not visible in the final image.

Recent advances in hierarchical Z-buffering have effectively incorporated ideas previously only used in deferred rendering, including the idea of being able to split a scene into tiles and of potentially being able to accept or reject tile sized pieces of polygon.

Today, the PowerVR software and hardware suite has ASICs for [[video encoding]], [[video decoding|decoding]] and associated [[image processing]]. It also has virtualisation, and [[DirectX]], [[OpenGL ES]], [[OpenVG]], and [[OpenCL]] acceleration.<ref>[https://www.engadget.com/2011/06/02/texas-instruments-announces-multi-core-1-8ghz-omap4470-arm-proc/ Texas Instruments announces multi-core, 1.8GHz OMAP4470 ARM processor for Windows 8], By Amar Toor, June 2, 2011, Engadget</ref>
Newest PowerVR Wizard GPUs have [[graphics pipeline|fixed-function]] [[Ray-tracing hardware|Ray Tracing]] Unit (RTU) hardware and support hybrid rendering.<ref>{{Cite web|url=https://www.imaginationtech.com/graphics-processors/|title=PowerVR - embedded graphics processors powering iconic products|website=Imagination}}</ref>

==PowerVR Graphics==
{{See also|List of PowerVR products}}

===Series1 (NEC)===
[[File:VideoLogic Apocalypse 3Dx.jpg|thumb|VideoLogic Apocalypse 3Dx (NEC PowerVR PCX2 chip)]]
[[File:NEC D62011GD (PowerVR PCX2).png|thumb|NEC D62011GD (PowerVR PCX2)]]
The first series of PowerVR cards was mostly designed as 3D-only accelerator boards that would use the main 2D video card's memory as framebuffer over PCI. Videologic's first PowerVR PC product to market was the 3-chip Midas3, which saw very limited availability in some OEM [[Compaq]] PCs.<ref>{{cite web|title=Compaq Selects PowerVR 3D Graphics Architecture for Next- Generation, High-Performance Presarios Home PCs|url=http://www.pure.com/press-release/compaq-selects-powervr-3d-graphics-architecture-for-next-generation-high-performance-presarios-home-pcs-7/|archive-url=https://archive.today/20130630053742/http://www.pure.com/press-release/compaq-selects-powervr-3d-graphics-architecture-for-next-generation-high-performance-presarios-home-pcs-7/|url-status=dead|archive-date=June 30, 2013|publisher=Imagination Technologies Limited|access-date=24 April 2013}}</ref><ref>{{cite web|title=VideoLogic Targets PC OEMs with PowerVR 3D Accelerator Card|url=http://www.pure.com/press-release/videologic-targets-pc-oems-with-powervr-3d-accelerator-card-8/|archive-url=https://archive.today/20130630044351/http://www.pure.com/press-release/videologic-targets-pc-oems-with-powervr-3d-accelerator-card-8/|url-status=dead|archive-date=June 30, 2013|publisher=Imagination Technologies Limited}}</ref> This card had very poor compatibility with all but the first Direct3D games, and even most SGL games did not run. However, its internal 24-bit color precision rendering was notable for the time.

The single-chip PCX1 was released in retail as the VideoLogic Apocalypse 3D<ref>{{cite web|title=VideoLogic Launches PowerVR-Based 3D Graphics Card Apocalypse 3D|url=https://www.pure.com/uk/press/november-1996/article_3696|publisher=Imagination Technologies Limited|access-date=24 April 2013|archive-date=5 May 2018|archive-url=https://web.archive.org/web/20180505134842/https://www.pure.com/uk/press/november-1996/article_3696|url-status=dead}}</ref> and featured an improved architecture with more texture memory, ensuring better game compatibility. This was followed by the further refined PCX2, which clocked 6&nbsp;MHz higher, offloaded some driver work by including more chip functionality<ref>{{Cite web |date=August 23, 2017 |title=Back to the start: PowerVR 25 |url=https://blog.imaginationtech.com/back-to-the-start-powervr-25/}}</ref> and added bilinear filtering, and was released in retail on the Matrox M3D<ref>{{cite web|title=Matrox Graphics Inc. Selects PowerVR for new 3D Accelerator Add-In Card Range|url=http://www.pure.com/press/august-1997/article_3724|publisher=Imagination Technologies Limited|access-date=2016-01-20|archive-date=2016-09-09|archive-url=https://web.archive.org/web/20160909133348/http://www.pure.com/press/august-1997/article_3724|url-status=dead}}</ref> and Videologic Apocalypse 3Dx cards. There was also the Videologic Apocalypse 5D Sonic, which combined the PCX2 accelerator with a [[Tseng Labs|Tseng]] ET6100 2D core and ESS Agogo sound on a single PCI board.

The PowerVR PCX cards were placed in the market as budget products and performed well in the games of their time, but weren't quite as fully featured as the [[3DFX Voodoo]] accelerators (due to certain blending modes being unavailable, for instance). However, the PowerVR approach of rendering to the 2D card's memory meant that much higher 3D rendering resolutions could be possible in theory, especially with PowerSGL games that took full advantage of the hardware.

* All models support [[DirectX]] 3.0 and PowerSGL, MiniGL drivers available for select games

{| class="wikitable" style="text-align: center"
|-
!rowspan=2|Model
!rowspan=2|Launch
!rowspan=2|Fab ([[nanometer|nm]])
!rowspan=2|Memory ([[Mebibyte|MiB]])
!rowspan=2|Core clock ([[Hertz|MHz]])
!rowspan=2|Memory clock ([[Hertz|MHz]])
!rowspan=2|Core config<sup>1</sup>
!colspan=4|[[Fillrate]]
!colspan=3|Memory
|-
!MOperations/s
!MPixels/s
!MTexels/s
!MPolygons/s
!Bandwidth ([[Gigabyte|GB]]/s)
!Bus type
!Bus width ([[bit]])
|-
!Midas3
|rowspan=2|1996
|?
|2
|66
|66
|rowspan=3|1:1
|66
|66
|66
|0
|0.24<sup>2</sup>
|[[Synchronous dynamic random-access memory#SDR|SDR]]+FPM<sup>2</sup>
|32+16<sup>2</sup>
|-
!PCX1
|500
|rowspan=2|4
|60
|60
|60
|60
|60
|0
|0.48
|rowspan=2|SDR
|rowspan=2|64
|-
!PCX2
|1997
|350
|66
|66
|66
|66
|66
|0
|0.528
|-
|}

* <sup>1</sup> [[Texture mapping unit]]s: [[render output unit]]s
*<sup>2</sup> Midas3 is 3-chip (vs. single-chip PCX series) and uses a split memory architecture: 1 MB 32-bit SDRAM (240 MB/s peak bandwidth) for textures and 1 MB 16-bit FPM DRAM for geometry data (and presumably for PCI communication). PCX series has only texture memory.

===Series2 (NEC)===
The second generation '''PowerVR2''' ("PowerVR Series2", chip codename "CLX2") was brought to market in the [[Dreamcast]] console between 1998 and 2001. As part of an internal competition at [[Sega]] to design the successor to the [[Sega Saturn|Saturn]], the PowerVR2 was licensed to NEC and was chosen ahead of a rival design based on the [[3dfx]] [[Voodoo2]]. It was called "the Highlander Project" during development.<ref>{{cite magazine |title=Power VR Prepares Highlander |magazine=[[Next Generation (magazine)|Next Generation]]|issue=34|publisher=[[Imagine Media]] |date=October 1997|page=20}}</ref> The PowerVR2 was paired with the [[Hitachi SH-4]] in the Dreamcast, with the SH-4 as the [[T&L]] geometry engine and the PowerVR2 as the rendering engine.<ref name="Unified">{{cite journal|last1=Hagiwara |first1=Shiro |last2=Oliver |first2=Ian |title=Sega Dreamcast: Creating a Unified Entertainment World |journal=IEEE Micro |publisher=[[Institute of Electrical and Electronics Engineers]] |volume=19 |number=6 |date=November–December 1999 |pages=29–35 |url=http://computer.org/micro/articles/dreamcast_2.htm |url-status=dead |archive-url=https://web.archive.org/web/20000823204755/http://computer.org/micro/articles/dreamcast_2.htm |archive-date=2000-08-23 |doi=10.1109/40.809375|url-access=subscription }}</ref> The PowerVR2 also powered the [[Sega Naomi]], the upgraded [[arcade system board]] counterpart of the Dreamcast.

However, the success of the Dreamcast meant that the PC variant, sold as Neon 250, appeared a year late to the market,<ref>{{Cite web |last=Andrawes |first=Michael |title=The Future & History of Tile Rendering - Imagination Technologies / STMicro PowerVR Series 3: KYRO |url=https://www.anandtech.com/show/558 |access-date=2024-02-20 |website=www.anandtech.com}}</ref> in late 1999. The Neon 250 was nevertheless competitive with the [[RIVA TNT2]] and [[Voodoo3]].<ref>">{{cite web | title=Sharky Extreme | website=sharkyextreme.com | date=2000-10-11 | url=http://sharkyextreme.com/hardware/reviews/video/neon250/15.shtml | archive-url=https://web.archive.org/web/20001011035118/http://sharkyextreme.com/hardware/reviews/video/neon250/15.shtml | archive-date=2000-10-11 | url-status=dead | access-date=2021-01-18}}</ref> The Neon 250 features inferior hardware specifications compared to the PowerVR2 part used in Dreamcast, such as a halved tile size, among others.

* All models are fabricated with a 250&nbsp;nm process
* All models support [[DirectX]] 6.0 
* PMX1 supports PowerSGL 2 and includes a MiniGL driver optimized for ''[[Quake III Arena]]''

{| class="wikitable" style="text-align: center"
|-
!rowspan=2|Model
!rowspan=2|Launch
!rowspan=2|Memory ([[Mebibyte|MiB]])
!rowspan=2|Core clock ([[Hertz|MHz]])
!rowspan=2|Memory clock ([[Hertz|MHz]])
!rowspan=2|Core config<sup>1</sup>
!colspan=4|[[Fillrate]]
!colspan=3|Memory
|-
!MOperations/s
!MPixels/s
!MTexels/s
!MPolygons/s
!Bandwidth ([[Gigabyte|GB]]/s)
!Bus type
!Bus width ([[bit]])
|-
!CLX2<ref name="Unified"/>
|1998
|8
|100
|100
|rowspan=2|1:1
|3200
|3200 <sup>2</sup> <br /> 100 <sup>3</sup>
|3200 <sup>2</sup> <br /> 100 <sup>3</sup>
|7 <sup>4</sup>
|0.8
|rowspan=2|SDR
|rowspan=2|64
|-
!PMX1
|1999
|32
|125
|125
|125
|125
|125
|0
|1
|-
|}

* <sup>1</sup> [[Texture mapping unit]]s: [[render output unit]]s
*<sup>2</sup> Fillrate for opaque polygons.
*<sup>3</sup> Fillrate for translucent polygons with hardware sort depth of 60.
*<sup>4</sup> [[Hitachi SH-4]] geometry engine calculates [[T&L]] for more than 10&nbsp;million triangles per second. CLX2 rendering engine throughput is 7&nbsp;million triangles per second.

===Series3 (STMicro)===

In 2000, the third generation '''PowerVR3''' STG4000 ''KYRO'' was released, manufactured by new partner [[STMicroelectronics]]. The architecture was redesigned for better game compatibility and expanded to a dual-pipeline design for more performance. The refresh STM PowerVR3 KYRO II, released later in 2001, likely had a lengthened pipeline to attain higher clock speeds<ref>{{Cite web|url=https://www.anandtech.com/show/735|title=STMicroelectronics Kyro II 64MB|first=Matthew|last=Witheiler|website=www.anandtech.com}}</ref> and was able to rival the more expensive ATI [[Radeon]] DDR and NVIDIA [[GeForce 2]] GTS in some benchmarks of the time, despite its modest specifications on paper and lack of hardware [[transform and lighting]] (T&L), a fact that Nvidia especially tried to capitalize on in a confidential paper they sent out to reviewers.<ref>{{cite web |url=http://dumpster.hardwaretidende.dk/dokumenter/nvidia_on_kyro.pdf |title=Whay you should know |date=2001 |website=dumpster.hardwaretidende.dk |access-date=2021-01-18}}</ref> As games increasingly started to include more geometry with this feature in mind, the KYRO II lost its competitiveness.

The KYRO series had a decent featureset for a budget-oriented GPU in their time, including a few Direct3D 8.1-compliant features such as 8-layer multitexturing (not 8-pass) and Environment Mapped Bump Mapping (EMBM); Full Scene Anti-Aliasing (FSAA) and Trilinear/Anisotropic filtering were also present.<ref>{{Cite web|url=https://www.imaginationtech.com/news/press-release/imagination-technologies-powervr-in-stmicroelectronics-kyro-pc-graphics-accelerator-unveiled/|archive-url=https://web.archive.org/web/20211127164455/https://www.imaginationtech.com/news/press-release/imagination-technologies-powervr-in-stmicroelectronics-kyro-pc-graphics-accelerator-unveiled/|url-status=dead|archive-date=November 27, 2021|title=Imagination Technologies' PowerVR™ in STMicroelectronics' KYRO™ PC Graphics Accelerator Unveiled|website=Imagination}}</ref><ref>{{Cite web|url=https://www.imaginationtech.com/news/press-release/stmicrolectronics-announces-next-generation-kyro-ii-3d-graphics-accelerator/|title=STMicrolectronics announces next generation KYRO II ™ 3D Graphics Accelerator|website=Imagination|access-date=2021-01-18|archive-date=2021-01-22|archive-url=https://web.archive.org/web/20210122030918/https://www.imaginationtech.com/news/press-release/stmicrolectronics-announces-next-generation-kyro-ii-3d-graphics-accelerator/|url-status=dead}}</ref><ref>{{Cite web|url=https://www.imaginationtech.com/news/press-release/powervr-technologies-debuts-kyro-ii-se-graphics-processor-at-cebit-2002/|title=PowerVR Technologies Debuts KYRO II SE™ Graphics Processor at CeBIT 2002|website=Imagination|access-date=2021-01-18|archive-date=2021-01-22|archive-url=https://web.archive.org/web/20210122184445/https://www.imaginationtech.com/news/press-release/powervr-technologies-debuts-kyro-ii-se-graphics-processor-at-cebit-2002/|url-status=dead}}</ref> KYRO II could also perform Dot Product (Dot3) Bump Mapping at a similar speed as GeForce 2 GTS in benchmarks.<ref>{{Cite web|url=https://aceshardware.com/Spades/read.php?article_id=25000237|archive-url=https://web.archive.org/web/20020202062747/https://aceshardware.com/Spades/read.php?article_id=25000237|url-status=dead|archive-date=February 2, 2002|title=Ace's Hardware|date=February 2, 2002}}</ref> Omissions included hardware T&L (an optional feature in Direct3D 7), Cube Environment Mapping and legacy 8-bit paletted texture support. While the chip supported [[S3TC]]/DXTC texture compression, only the (most commonly used) DXT1 format was supported.<ref>{{Cite web|url=https://www.beyond3d.com/content/reviews/13/4|title=Beyond3D - Imagination Technologies Videologic Vivid! 32MB KYRO|website=www.beyond3d.com}}</ref> Support for the proprietary PowerSGL API was also dropped with this series.

16-bit output quality was excellent compared to most of its competitors, thanks to rendering to its internal 32-bit tile cache and downsampling to 16-bit instead of straight use of a 16-bit framebuffer.<ref>{{cite web |url=https://www.vogons.org/download/file.php?id=20346 |title=Data|website=www.vogons.org|access-date=2021-01-18}}</ref> This could play a role in improving performance without losing much image quality, as memory bandwidth was not plentiful. However, due to its unique concept on the market, the architecture could sometimes exhibit flaws such as missing geometry in games, and therefore the driver had a notable amount of compatibility settings, such as switching off the internal Z-buffer. These settings could cause a negative impact on performance.

A second refresh of the KYRO was planned for 2002, the STG4800 KYRO II SE. Samples of this card were sent to reviewers but it does not appear to have been brought to market. Apart from a clockspeed boost, this refresh was announced with a "EnT&L" HW T&L software emulation, which eventually made it into the drivers for the previous KYRO cards starting with version 2.0. The STG5500 KYRO III, based upon the next-generation '''PowerVR4''', was completed and would have included hardware T&L but was shelved due to STMicro closing its graphics division.

<gallery mode="packed" heights="150px">
File:Hercules 3D Prophet 4000XT 64MB PCI.png|Hercules 3D Prophet 4000XT 64MB PCI with the KYRO chipset.
File:STMicroelectronics@250nm@PowerVR Series3@STG-4000@ST Kyro@STG4000-X A3S F 42775.1 9224L0149 MALTA DSC04968.jpg|The Hercules 3D Prophet 4000XT aside a Kyro chipset
File:STMicroelectronics@250nm@PowerVR Series3@STG-4000@ST Kyro@STG4000-X A3S F 42775.1 9224L0149 MALTA DSCx6 polysilicon@5x.jpg|Die shot of the Kyro chipset
File:STMicroelectronics STG4500 (PowerVR Kyro II).png|KYRO II.
File:STMicroelectronics@180nm@PowerVR Series3@STG-4500@ST Kyro II@STG4500-X A2S F40557.1 9H27L0111 KOREA DSCx5 polysilicon-layer@5x.jpg|Die shot of the Kyro II
</gallery>
* All models support [[DirectX]] 6.0
{| class="wikitable" style="text-align: center"
|-
!rowspan=2|Model
!rowspan=2|Launch
!rowspan=2|Fab ([[nanometer|nm]])
!rowspan=2|Memory ([[Mebibyte|MiB]])
!rowspan=2|Core clock ([[Hertz|MHz]])
!rowspan=2|Memory clock ([[Hertz|MHz]])
!rowspan=2|Core config<sup>1</sup>
!colspan=4|[[Fillrate]]
!colspan=3|Memory
|-
!MOperations/s
!MPixels/s
!MTexels/s
!MPolygons/s
!Bandwidth ([[Gigabyte|GB]]/s)
!Bus type
!Bus width ([[bit]])
|-
!STG4000 KYRO
|2000<ref>{{Cite web |title=Már kapható a PowerColor Evil Kyro 64MB |url=https://www.hwsw.hu/hirek/6549/mar-kaphato-a-powercolor-evil-kyro-64mb.html |access-date=2022-05-08 |website=HWSW |language=hu}}</ref>
|250
|32/64
|115
|115
|rowspan=3|2:2
|230
|230
|230
|0
|1.84
|rowspan=3|SDR
|rowspan=4|128
|-
!STG4500 KYRO II
|2001
|rowspan=2|180
|32/64
|175
|175
|350
|350
|350
|0
|2.8
|-
!STG4800 KYRO II SE
|2002
|rowspan=2|64
|200
|200
|400
|400
|400
|0
|3.2
|-
!STG5500 KYRO III
|Never Released
|130
|250
|250
|4:4
|1000
|1000
|1000
|0
|8
|DDR
|-
|}

* <sup>1</sup> [[Texture mapping unit]]s: [[render output unit]]s

===Series4 (STMicro)===
PowerVR achieved great success in the mobile graphics market with its low power '''PowerVR MBX'''. MBX, and its SGX successors, were licensed a number of the top mobile semiconductor manufacturers in their mobile [[System on a chip|SoC]] chipsets, including [[Intel]], [[Texas Instruments]], [[Samsung]], [[NEC]], [[NXP Semiconductors]], [[Freescale]], [[Renesas]], [[SiRF]], [[Marvell Technology|Marvell]], and Sunplus.<ref>{{cite web |title=Imagination Technologies {{!}} Imagination Technologies Extends Its Lead In Mobile Graphics PowerVR SGX |url=https://www.realwire.com/releases/imagination-technologies-extends-its-lead-in-mobile-graphics-powervr-sgx |website=RealWire |access-date=4 May 2023 |date=10 February 2008}}</ref>

These mobile chipsets with MBX IP in turn were used in several high-end cellphones and smartphones, including the original [[iPhone]] and [[iPod Touch]] (with Samsung S5L8900), [[Nokia N95]] and [[Motorola RIZR Z8]] (with TI [[OMAP]] 2420), and the [[Sony Ericsson]] [[Sony Ericsson P1|P1]] and [[Sony Ericsson M600|M600]] ([[Nexperia (processor)#Nexperia PNX4008 SoC|NXP Nexperia PNX4008]]). It was also used in some [[Personal digital assistant|PDA]]s such as the [[Axim X50 series|Dell Axim X50V]] and [[Dell Axim#X51 family|X51V]] featuring the [[Intel 2700G]] co-processor, as well as in set-top boxes featuring the MBX Lite-powered Intel CE 2110.

There were two variants: MBX and MBX Lite. Both had the same feature set, where the MBX was optimized for speed and MBX Lite was optimized for low power consumption. The MBX could also be paired up with options to include either a full or lite [[Floating-point unit|FPU]], and/or full or lite VGP (Vector Graphics Processor).

{|class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Year
!rowspan=2|Die Size (mm<sup>2</sup>){{efn|name=fn1|Official Imgtec data}}
!rowspan=2|Core config
!colspan=2|[[Fillrate]] @ 200&nbsp;MHz
!rowspan=2|Bus width ([[bit]])
!colspan=2|[[Application Programming Interface|API]] (version)
|-
!MTriangles/s{{efn|name=fn1}}
![[Pixel|MPixel/s]]{{efn|name=fn1}}
![[DirectX]]
![[OpenGL]]
|-
|MBX Lite
|rowspan=2|Feb 2001
|4 @ 130&nbsp;nm?
|rowspan=2|0/1/1/1
|1.0
|100
|rowspan=2|64
|rowspan=2|7.0, VS 1.1
|rowspan=2|1.1
|-
|MBX
|8 @ 130&nbsp;nm?
|1.68
|150
|}

===Series5 (SGX)===
PowerVR's Series5 SGX series features [[pixel shader|pixel]], [[vertex shader|vertex]], and [[geometry shader]] hardware, supporting [[OpenGL ES]] 2.0 and [[DirectX]] 10.1 with Shader Model 4.1.

The SGX GPU core is included in several popular [[system-on-a-chip|systems-on-chip]] (SoC) used in many portable devices. Apple uses the [[Apple A4|A4]] (manufactured by Samsung) in their [[iPhone 4]], [[iPad]], [[iPod Touch]], and [[Apple TV]], and in the [[Apple Watch]] as part of [[Apple S1]]. [[Texas Instruments]]' [[Texas Instruments OMAP|OMAP]] 3 and 4 series SoC's are used in the [[Kindle Fire HD|Amazon's Kindle Fire HD 8.9"]], [[Nook HD|Barnes and Noble's Nook HD(+)]], [[BlackBerry PlayBook]], [[Nokia N9]], [[Nokia N900]], [[Sony Ericsson Vivaz]], [[Motorola Droid|Motorola Droid/Milestone]], [[Motorola Defy]], Motorola RAZR D1/D3, Droid Bionic, [[Archos 70]], [[Palm Pre]], [[Samsung Galaxy SL]], [[Galaxy Nexus]], [[Open Pandora]], and others. Samsung produces the ''[[Hummingbird Processor|Hummingbird]]'' SoC and uses it in their [[Samsung Galaxy S series|Samsung Galaxy S]], [[Samsung Galaxy Tab|Galaxy Tab]], [[Samsung Wave S8500]] [[Samsung Wave II S8530]] and Samsung Wave III S860 devices. Hummingbird is also in [[Meizu M9]] smartphone.

[[Intel]] used a number of SGX products in its [[Mobile Internet device#Menlow platform (2008)|Menlow]], [[Mobile Internet device#Moorestown platform (2010)|Moorestown]], [[Medfield]] and [[Mobile Internet device#Clover Trail+ platform (2012)|Clover Trail+]] [[Atom (system on a chip)|Atom]]-based [[Mobile Internet device|MID]] platforms. Using the SGX graphics chipsets helped Intel to successfully achieve the ultra-low power budgets required for passively cooled devices, such as smartphones, tablets and netbooks.<ref>{{cite web | last=Shimpi | first=Anand Lal | title=Intel's Medfield & Atom Z2460 Arrive for Smartphones: It's Finally Here | website=AnandTech | url=https://www.anandtech.com/show/5365/intels-medfield-atom-z2460-arrive-for-smartphones | access-date=2021-01-18}}</ref> However, the significant difference in graphics architecture resulted in poor driver support.<ref>{{cite web |title=Intel Is Planning To Drop PowerVR Graphics |url=https://www.phoronix.com/news/MTA1NDU |website=www.phoronix.com |access-date=4 May 2023 |language=en}}</ref>

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Year
!rowspan=2|Die Size (mm<sup>2</sup>){{efn|name=fn1}}
!rowspan=2|Core config{{efn|name=fn2|USSE ([[Unified Shader Model|Universal]] Scalable Shader Engine) lanes/[[Texture mapping unit|TMUs]]}}
!colspan=2|[[Fillrate]] @ 200&nbsp;MHz
!rowspan=2|Bus width ([[bit]])
!colspan=3|[[Application Programming Interface|API]] (version)
!rowspan=2|[[FLOPS|GFLOPS]] @ 200&nbsp;MHz
|-
!MTriangles/s{{efn|name=fn1}}
![[Pixel|MPixel/s]]{{efn|name=fn1}}
![[OpenGL ES]]
![[OpenGL]]
![[Direct3D]]
|-
|SGX520
|rowspan=2|Jul 2005
|2.6@65&nbsp;nm
|1/1
|7
|100
|rowspan=6|32-128
|rowspan=6|2.0
|{{N/A}}
|{{N/A}}
|0.8
|-
|SGX530
|7.2@65&nbsp;nm
|rowspan=2|2/1
|rowspan=3|14
|rowspan=2|200
|{{N/A}}
|{{N/A}}
|rowspan=3|1.6
|-
|SGX531
|Oct 2006
|{{dunno}}
|{{N/A}}
|{{N/A}}
|-
|SGX535
|rowspan=2|Nov 2007
|{{dunno}}
|2/2
|rowspan=3|400
|rowspan=2|2.1
|9.0c
|-
|SGX540
|{{dunno}}
|rowspan=2|4/2
|20
|{{N/A}}
|rowspan=2|3.2
|-
|SGX545
|Jan 2010
|12.5@65&nbsp;nm
|40
|3.2
|10.1
|-
|}

===Series5XT (SGX)===
<!-- PowerVR SGXMP redirects here -->
PowerVR Series5XT SGX chips are multi-core variants of the SGX series with some updates. It is included in the [[PlayStation Vita]] portable gaming device with the MP4+ Model of the PowerVR SGX543, the only intended difference, aside from the + indicating features customized for Sony, is the cores, where MP4 denotes 4 cores (quad-core) whereas the MP8 denotes 8 cores (octo-core). The [[Allwinner A31]] (quad-core mobile application processor) features the dual-core SGX544 MP2. The [[Apple Inc.|Apple]] [[iPad 2]] and [[iPhone 4S]] with the [[Apple A5|A5]] SoC also feature a dual-core SGX543MP2. The [[iPad (3rd generation)]] [[Apple A5X|A5X]] SoC features the quad-core SGX543MP4.<ref>[http://www.anandtech.com/show/4216/apple-ipad-2-gpu-performance-explored-powervr-sgx543mp2-benchmarked Apple iPad 2 GPU Performance Explored: PowerVR SGX543MP2 Benchmarked], by Anand Lal Shimpi, 2011/03/12, Anandtech</ref> The [[iPhone 5]] [[Apple A6|A6]] SoC features the tri-core SGX543MP3. The [[iPad (4th generation)]] [[Apple A6X|A6X]] SoC features the quad-core SGX554MP4. The [[Exynos]] variant of the [[Samsung Galaxy S4]] sports the tri-core SGX544MP3 clocked at 533&nbsp;MHz.

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn3|USSE2 (Universal Scalable Shader Engine 2) lanes/[[Texture mapping unit|TMUs]]}}
!colspan=3|[[Fillrate]] @ 200&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=4|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 200&nbsp;MHz<br>(per core)
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
|SGX543
|Jan 2009
|rowspan=3|1-16
|rowspan=2|5.4@32&nbsp;nm
|rowspan=2|4/2
|rowspan=3|35
|rowspan=3|3.2
|{{dunno}}
|rowspan=3|128-256
|{{dunno}}
|rowspan=3|2.0
|2.0?
|rowspan=3|1.1
|[[Direct3D feature levels|9.0 L1]]
|rowspan=2|6.4
|-
|SGX544
|Jun 2010
|{{dunno}}
|{{dunno}}
|0.0
|rowspan=2|[[Direct3D feature levels|9.0 L3]]
|-
|SGX554
|Dec 2010
|8.7@32&nbsp;nm
|8/2
|{{dunno}}
|{{dunno}}
|2.1
|12.8
|}

These GPU can be used in either single-core or multi-core configurations.<ref name=omap>{{cite web | last=Klug | first=Brian | title=TI Announces OMAP4470 and Specs: PowerVR SGX544, 1.8 GHz Dual Core Cortex-A9 | website=AnandTech | url=https://www.anandtech.com/show/4413/ti-announces-omap-4470-and-specs-powervr-sgx544-18-ghz-dual-core-cortexa9 | access-date=2021-01-18}}</ref>

=== Series5XE (SGX) ===
Introduced in 2014, the PowerVR GX5300 GPU<ref>{{Cite web |title=PowerVR Series5XE GX5300 GPU - Imagination Technologies |url=https://www.imaginationtech.com/product/powervr-series5xe/ |access-date=2016-06-22 |website=Imagination Technologies |language=en-GB}}</ref> is based on the SGX architecture and is the world's smallest Android-capable graphics core, providing low-power products for entry-level smartphones, wearables, IoT and other small footprint embedded applications, including enterprise devices such as printers.

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn3|USSE2 (Universal Scalable Shader Engine 2) lanes/[[Texture mapping unit|TMUs]]}}
!colspan=3|[[Fillrate]] @ 200&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=4|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 200&nbsp;MHz<br>(per core)
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
|GX5300
|Jul 2014
|{{dunno}}
|0.55@28&nbsp;nm
|{{dunno}}
|{{dunno}}
|{{dunno}}
|{{dunno}}
|{{dunno}}
|{{dunno}}
|2.0
|{{dunno}}
|{{dunno}}
|{{dunno}}
|0.8
|}

===Series6 (Rogue)===
PowerVR Series6 GPUs<ref>{{Cite web |title=PowerVR Series6 - Imagination Technologies |url=https://www.imaginationtech.com/products/catalog/?query=product&filter-search=&filter-series=series6 |access-date=2016-06-22 |website=Imagination Technologies |language=en-GB}}</ref> are based on an evolution of the SGX architecture codenamed ''Rogue''. [[ST-Ericsson]] (now defunct) announced that its [[NovaThor|Nova]] application processors would include Imagination's next-generation PowerVR Series6 architecture.<ref>{{Cite news |date=15 February 2011 |title=Imagination partners drive mobile and embedded graphics to new level |url=http://imgtec.com/news/Release/index.asp?NewsID=617 |url-status= |access-date= |archive-url= https://web.archive.org/web/20130118223747/http://imgtec.com/news/Release/index.asp?NewsID=617|archive-date=2013-01-18}}, Imagination Technologies Ltd.</ref> MediaTek announced the quad-core MT8135 [[system on a chip]] (SoC) (two ARM [[Cortex-A15]] and two ARM [[Cortex-A7]] cores) for tablets.<ref>{{Cite news|url=http://www.mediatek.com/_en/03_news/01-2_newsDetail.php?sn=1115|title=MediaTek Introduces Industry Leading Tablet SoC, MT8135|url-status=dead|archive-url=https://web.archive.org/web/20130801180055/http://www.mediatek.com/_en/03_news/01-2_newsDetail.php?sn=1115|archive-date=2013-08-01}}, MediaTek Inc.</ref> Renesas announced its R-Car H2 SoC includes the G6400.<ref>{{Cite news|url=http://am.renesas.com/applications/automotive/cis/cis_highend/rcar_h2/index.jsp|title=R-Car H2}}, Renesas Electronics Corporation Ltd</ref> [[Allwinner Technology]] A80 SoC, (4 Cortex-A15 and 4 Cortex-A7) that is available in the Onda V989 tablet, features a PowerVR G6230 GPU.<ref>{{Cite web|url=https://www.cnx-software.com/2014/07/01/cubieboard-8-specs-pictures/|title=Pictures and Specs for CubieBoard 8 Development Board Powered by AllWinner A80 SoC|first=Jean-Luc|last=Aufranc|date=July 1, 2014}}</ref> The [[Apple A7]] SoC integrates a [[graphics processing unit]] (GPU) which [[AnandTech]] believes to be a PowerVR G6430 in a four cluster configuration.<ref name="AnandTech-iPhone5s-GPU">{{cite web |first=Anand |last=Lal Shimpi |url=http://anandtech.com/show/7335/the-iphone-5s-review/7 |title=The iPhone 5s Review: GPU Architecture |date= September 17, 2013 |publisher=AnandTech |access-date= September 18, 2013}}</ref>

Intel also continued its use of PowerVR graphics exclusively in its ultra-low-power [[Mobile Internet device#Merrifield platform (2013)|Merrifield]] and [[Mobile Internet device#Moorefield platform (2014)|Moorefield]] smartphone Atom platforms.<ref>{{cite web |last1=Shimpi |first1=Anand Lal |title=Intel Talks Atom Z3460/Z3480 (Merrifield), Z3560/Z3580 (Moorefield) and LTE at MWC 2014 |url=https://www.anandtech.com/show/7789/intel-talks-merrifield-moorefield-and-lte-at-mwc-2014 |website=www.anandtech.com |access-date=4 May 2023}}</ref>

PowerVR Series 6 GPUs have 2 TMUs/cluster.<ref>{{Cite web|url=https://www.anandtech.com/show/7335/the-iphone-5s-review|title=The iPhone 5s Review|first=Anand Lal|last=Shimpi|website=www.anandtech.com}}</ref>

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4|USC (Unified Shading Cluster) lanes/[[Texture mapping unit|TMUs]] per cluster}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 600&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=5|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 600&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)|Vulkan]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
|G6100
|Feb 2013
|1
|??@28&nbsp;nm 
|1/4
|16
|?
|colspan=2 rowspan=3|2.4
|128
|{{dunno}}
|rowspan="6" | 1.1
| rowspan="6" |3.1
|2.x
| rowspan="6" |1.2
|[[Direct3D feature levels|9.0 L3]]
|38.4 / 57.6
|-
|G6200
|Jan 2012
|rowspan=2|2
|??@28&nbsp;nm 
|rowspan=2|2/2
|rowspan=2|32
|?
|?
|{{dunno}}
| rowspan="5" |3.2
| rowspan="5" |10.0
|76.8 / 76.8
|-
|G6230
|Jun 2012
|??@28&nbsp;nm 
|?
|?
|{{dunno}}
|76.8 / 115.2
|-
|G6400
|Jan 2012
|rowspan=2|4
|??@28&nbsp;nm 
|rowspan=2|4/2
|rowspan=2|64
|?
|rowspan=2 colspan=2|4.8
|?
|{{dunno}}
|153.6 / 153.6
|-
|G6430
|Jun 2012
|??@28&nbsp;nm 
|?
|?
|{{dunno}}
|153.6 / 230.4
|-
|G6630
|Nov 2012
|6
|??@28&nbsp;nm 
|6/2
|96
|?
|colspan=2|7.2
|?
|{{dunno}}
|230.4 / 345.6
|-
|-
|}

=== Series6XE (Rogue) ===
PowerVR Series6XE GPUs<ref>{{Cite web |title=PowerVR Series6XE GPU Family - Imagination Technologies |url=https://www.imaginationtech.com/product/powervr-series6xe/ |access-date=2016-06-22 |website=Imagination Technologies |language=en-GB}}</ref> are based around Series6 and designed as entry-level chips aimed at offering roughly the same fillrate compared to the Series5XT series. They however feature refreshed API support such as Vulkan, OpenGL ES 3.1, OpenCL 1.2 and DirectX 9.3 (9.3 L3).<ref name="AnandTech PowerVR Series6XE architecture" /> Rockchip and Realtek have used Series6XE GPUs in their SoCs.

PowerVR Series 6XE GPUs were announced on January 6, 2014.<ref name="AnandTech PowerVR Series6XE architecture">[http://www.anandtech.com/show/7630/imagination-technologies-announces-entrylevel-powervr-series6xe-gpu-family Imagination Technologies Announces Entry-Level PowerVR Series6XE GPU Family], January 6, 2014, AnandTech</ref>

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] 
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=5|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]](@ 600&nbsp;MHz)
FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)|Vulkan]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
|G6050
|Jan 2014
|rowspan=2|0.5
|??@28&nbsp;nm 
|?/?
|?
|?
|??
|?
|?
|{{dunno}}
|rowspan="4" | 1.1
| rowspan="4" |3.1
| rowspan="4" |3.2
| rowspan="4" |1.2
| rowspan="4" |[[Direct3D feature levels|9.0 L3]]
| rowspan="2" |19.2 / 38.4
|-
|G6060
|Jan 2014
|??@28&nbsp;nm 
|?/?
|?
|?
|??
|?
|?
|{{dunno}}
|-
|G6100 (XE)
|Jan 2014
|rowspan=2|1
|??@28&nbsp;nm 
|?/?
|?
|?
|??
|?
|?
|{{dunno}}
|rowspan="2" |38.4 / 76.8
|-
|G6110
|Jan 2014
|??@28&nbsp;nm 
|?/?
|?
|?
|??
|?
|?
|{{dunno}}
|-
|-
|}

===Series6XT (Rogue)===
PowerVR Series6XT GPUs<ref>{{Cite web |title=PowerVR Series6XT GPU Family - Imagination Technologies |url=https://www.imaginationtech.com/product/powervr-series6xt/ |access-date=2016-06-22 |website=Imagination Technologies |language=en-GB}}</ref> aims at reducing power consumption further through die area and performance optimization providing a boost of up to 50% compared to Series6 GPUs. Those chips sport PVR3C triple compression system-level optimizations and Ultra HD deep color.<ref name="Imagination’s new generation PowerVR Series6XT architecture">[http://www.anandtech.com/show/7629/imagination-technologies-announces-powervr-series6xt-architecture-available-for-immediate-licensing Imagination Technologies Announces PowerVR Series6XT Architecture], January 6, 2014, Imagination</ref> The Apple [[iPhone 6]], [[iPhone 6 Plus]] and [[iPod Touch (6th generation)]] with the [[Apple A8|A8]] SoC feature the quad-core GX6450.<ref name="Chipworks-A8">{{cite web |publisher=Chipworks |url=http://www.chipworks.com/en/technical-competitive-analysis/resources/blog/inside-the-iphone-6-and-iphone-6-plus/ |title=Inside the iPhone 6 and iPhone 6 Plus |date=September 19, 2014 |access-date=September 24, 2014 |archive-url=https://web.archive.org/web/20150503155004/http://www.chipworks.com/en/technical-competitive-analysis/resources/blog/inside-the-iphone-6-and-iphone-6-plus/ |archive-date=May 3, 2015 |url-status=dead }}</ref><ref name="AnandTech-Chipworks">{{cite web |first=Ryan |last=Smith |url=http://www.anandtech.com/show/8562/chipworks-a8 |title=Chipworks Disassembles Apple's A8 SoC: GX6450, 4MB L3 Cache & More |publisher=AnandTech |date= September 23, 2014 |access-date= September 24, 2014}}</ref> An unannounced 8 cluster variant was used in the Apple A8X SoC for their iPad Air 2 model (released in 2014). The MediaTek MT8173 and Renesas R-Car H3 SoCs use Series6XT GPUs.

PowerVR Series 6XT GPUs were unveiled on January 6, 2014.<ref name="Anandtech-Series6XT">{{cite web |url=http://www.anandtech.com/show/7629/imagination-technologies-announces-powervr-series6xt-architecture-available-for-immediate-licensing |title=Imagination Technologies Announces PowerVR Series6XT Architecture |date=January 6, 2014 |first=Ryan |last=Smith |publisher=AnandTech}}</ref>

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 450&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=5|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 450&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)|Vulkan]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
|GX6240
|rowspan="4" | Jan 2014
|rowspan="2" | 2
|??@28&nbsp;nm 
|rowspan="2" | 2/4
|rowspan="2" | 64/128
|?
|colspan="2" | 
|?
|{{dunno}}
|rowspan="5" | 1.1
|rowspan="5" | 3.1
|rowspan="5" | 3.3
|rowspan="5" | 1.2
|rowspan="5" | 10.0
|rowspan="2" | 57.6/115.2
|-
|GX6250
|??@28&nbsp;nm 
|35
|colspan="2" | 2.8
|128
|{{dunno}}
|-
|GX6450
|4
|19.1mm2@28&nbsp;nm 
|4/8
|128/256
|?
|colspan="2" | 3.6
|?
|{{dunno}}
|115.2/230.4
|-
|GX6650
|6
|??@28&nbsp;nm 
|6/12
|192/384
|?
|colspan="2" | 
|?
|{{dunno}}
|172.8/345.6
|-
|GXA6850
|Unannounced
|8
|38mm2@28&nbsp;nm 
|8/16
|256/512
|?
|colspan="2" | 
|128
|{{dunno}}
|230.4/460.8
|-
|}

=== Series7XE (Rogue) ===
PowerVR Series 7XE GPUs were announced on 10 November 2014. When announced, the 7XE series contained the smallest [[OpenGL ES#Android Extension Pack|Android Extension Pack]] compliant GPU.

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=5|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] (FP32)<br>@ 650&nbsp;MHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)|Vulkan]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
| GE7400
| rowspan="2"| Nov 2014
| 0.5
|
|
|
|
| colspan="2" |
|
|
| rowspan="2" | 1.1
| rowspan="2" | 3.1
| rowspan="2" | 
| rowspan="2" |  1.2 embedded profile
| rowspan="2" | [[Direct3D feature level|9.0 L3]]
| 20.8
|-
| GE7800
| 1
|
|
|
|
| colspan="2" |
|
|
| 41.6
|-
|}

=== Series7XT (Rogue) ===
PowerVR Series7XT GPUs<ref>{{Cite web |title=PowerVR Series7XT GPU Family - Imagination Technologies |url=https://www.imaginationtech.com/product/powervr-series7xt/ |access-date=2016-06-22 |website=Imagination Technologies |language=en-GB}}</ref> are available in configurations ranging from two to 16 clusters, offering dramatically scalable performance from 100 GFLOPS to 1.5 TFLOPS. The GT7600 is used in the Apple iPhone 6s and iPhone 6s Plus models (released in 2015) as well as the Apple iPhone SE model (released in 2016) and the Apple iPad model (released in 2017) respectively. An unannounced 12 cluster variant was used in the Apple A9X SoC for their iPad Pro models (released in 2015).

PowerVR Series 7XT GPUs were unveiled on 10 November 2014.

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz 
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=5|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 650&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)|Vulkan]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenCL]]
![[Direct3D]]
|-
| GT7200
| rowspan="4" | Nov 2014
| 2
|
| 2/4
| 64/128
| 
| colspan="2" | 
|
|
| rowspan="6" | 1.1
| rowspan="6" | 3.1
| rowspan="6" | 3.3<br>(4.4 optional)
| rowspan="6" | 1.2 embedded profile<br>(FP optional)
| rowspan="6" | 10.0<br>(11.2 optional)
| 83.2/166.4
|-
| GT7400
| 4
|
| 4/8
| 128/256
|
| colspan="2" | 
|
|
| 166.4/332.8
|-
| GT7600
| 6
|
| 6/12
| 192/384
|
| colspan="2" | 
|
|
| 249.6/499.2
|-
| GT7800
| 8
|
| 8/16
| 256/512
|
| colspan="2" | 
|
|
| 332.8/665.6
|-
| GTA7850
| Unannounced
| 12
|
| 12/24
| 384/768
|
| colspan="2" | 
|
|
| 499.2/998.4
|-
| GT7900
| Nov 2014
| 16
|
| 16/32
| 512/1024
|
| colspan="2" | 
|
|
| 665.6/1331.2
|-
|}

=== Series7XT Plus (Rogue) ===
PowerVR Series7XT Plus GPUs are an evolution of the Series7XT family and add specific features designed to accelerate computer vision on mobile and embedded devices, including new INT16 and INT8 data paths that boost performance by up to 4x for OpenVX kernels. Further improvements in shared virtual memory also enable OpenCL 2.0 support. The GT7600 Plus is used in the Apple iPhone 7 and iPhone 7 Plus models (released in 2016) as well as the Apple iPad model (released in 2018).

PowerVR Series 7XT Plus GPUs were announced on International CES, Las Vegas – 6 January 2016.

Series7XT Plus achieve up to 4x performance increase for vision applications.
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 650&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GT7200 Plus
| rowspan="2" | January 2016
| 2
|?
| 2/4
| 64/128
| 
| colspan="2" | 2.6
|
|
| rowspan="3" | 1.1
| rowspan="3" | 3.2
| rowspan="2" | 3.3 (4.4 optional)
| rowspan="3" | 1.0.1
| rowspan="3" | 2.0
| rowspan="2" | ??
| 83.2/166.4
|-
| GT7400 Plus
| 4
| ?
| 4/8
| 128/256
|
| colspan="2" | 5.2
|
|
| 166.4/332.8
|-
|GT7600 Plus
|June 2016
|6
|??@10&nbsp;nm
|6/12
|192/384
|
| colspan="2" | 7.8
|
|
|4.4
|12
|249.6/499.2
|}
The GPUs are designed to offer improved in-system efficiency, improved power efficiency and reduced bandwidth for vision and computational photography in consumer devices, mid-range and mainstream smartphones, tablets and automotive systems such as advanced driver assistance systems (ADAS), infotainment, computer vision and advanced processing for instrument clusters.

The new GPUs include new feature set enhancements with a focus on next-generation compute:

Up to 4x higher performance for OpenVX/vision algorithms compared to the previous generation through improved integer (INT) performance (2x INT16; 4x INT8)
Bandwidth and latency improvements through shared virtual memory (SVM) in OpenCL 2.0
Dynamic parallelism for more efficient execution and control through support for device enqueue in OpenCL 2.0

=== Series8XE (Rogue) ===
PowerVR Series8XE GPUs support OpenGL ES 3.2 and Vulkan 1.x and are available in 1, 2, 4 and 8 pixel/clock configurations,<ref>{{cite web |title=PowerVR Series8XE GPU Family |url=https://www.imaginationtech.com/products/catalog/?query=product&filter-search=&filter-series=series8 |access-date=26 August 2018}}</ref> enabling the latest games and apps and further driving down the cost of high quality UIs on cost sensitive devices.

PowerVR Series 8XE were announced February 22, 2016 at the Mobile World Congress 2016. They are an iteration of the Rogue microarchitecture and target entry-level SoC GPU market.
New GPUs improve the performance/mm² for the smallest silicon footprint and power profile, while also incorporating hardware virtualization and multi-domain security.<ref>{{Cite web|url=https://www.imaginationtech.com/news/press-release/latest-imagination-powervr-series8xe-gpus-set-new-standard-for-performance-power-and-area-in-cost-sensitive-markets/|title=Latest Imagination PowerVR® Series8XE GPUs set new standard for performance, power and area in cost-sensitive markets|website=Imagination|access-date=2021-01-18|archive-date=2021-05-10|archive-url=https://web.archive.org/web/20210510051108/https://www.imaginationtech.com/news/press-release/latest-imagination-powervr-series8xe-gpus-set-new-standard-for-performance-power-and-area-in-cost-sensitive-markets/|url-status=dead}}</ref> Newer model were later released in January 2017, with a new low end and high end part.<ref name='8XE_plus'>{{cite news|last1=Smith|first1=Ryan|title=Imagination Announces PowerVR Series8XE Plus & New Series8XE Designs for Midrange Market|url=http://www.anandtech.com/show/11028/powervr-8xe-plus-announced|access-date=17 January 2017|publisher=Anandtech|date=17 January 2017}}</ref>
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 650&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GE8100<ref>https://www.imaginationtech.com/product/ge8100/</ref>
| January 2017
| rowspan="2" | 0.25 USC
|
| ?
| ?
| 
| colspan="2" | 0.65
|
|
| rowspan="5" | 1.1
| rowspan="5" | 3.2
| rowspan="5" | ?
| rowspan="5" | 1.1
| rowspan="5" | 1.2 EP
| rowspan="5" | 9.3 (optional)
| rowspan="2" | 10.4 / 20.8
|-
| GE8200<ref>{{cite web | url=https://www.imaginationtech.com/product/ge8200/ | title=IMG GE8200 GPU - Imagination | work=Imagination }}</ref>
| rowspan="3" | February 2016
|
| ?
| ?
| 
| colspan="2" | 1.3
|
|
|-
| GE8300<ref>{{cite web | url=https://www.imaginationtech.com/product/ge8300/ | title=IMG GE8300 GPU - Imagination | work=Imagination }}</ref>
| rowspan="2" | 0.5 USC
|
| ?
| ?
| rowspan="2" | 0.5
| colspan="2" rowspan="2" | 2.6
|
|
| rowspan="2" | 20.8 / 41.6
|-
| GE8310 
|
| ?
| ?
|
|
|-
| GE8430 
| January 2017
| 2 USC
|
| ?
| ?
|
| colspan="2" | 5.2
|
|
| 83.2 / 166.4
|}

=== Series8XEP (Rogue) ===
PowerVR Series8XEP were announced January 2017. There are an iteration of the Rogue microarchitecture and target the mid range SoC GPU market, targeting 1080p. The Series8XEP remains focused on die size and performance per unit
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 650&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GE8320<ref>https://www.imaginationtech.com/product/ge8320/</ref>
| rowspan="3" | January 2017
| rowspan="2" | 1 USC
|
| ?
| ?
| 
| colspan="2" rowspan="3" | 2.6
|
|
| rowspan="3" | 1.1
| rowspan="3" | 3.2
| rowspan="3" | ?
| rowspan="3" | 1.1
| rowspan="3" | 1.2 EP
| rowspan="3" | ?
| rowspan="2" | 41.6 / 83.2
|-
| GE8325 
|
| ?
| ?
|
|
|
|-
| GE8340 
| 2 USC
|
| ?
| ?
|
|
|
| 83.2 / 166.4
|}

=== Series8XT (Furian) ===
Announced on 8 March 2017, Furian is the first new PowerVR architecture since Rogue was introduced five years earlier.<ref>{{Cite news|url=http://www.anandtech.com/show/11186/imagination-announces-powervr-furian-gpu-architecture|title=Imagination Announces PowerVR Furian GPU Architecture: The Next Generation of PowerVR|last=Smith|first=Ryan|access-date=2017-03-08}}</ref>

PowerVR Series 8XT were announced March 8, 2017. It is the first series GPU's based on the new Furian architecture. According to Imagination, GFLOPS/mm² is improved 35% and Fill rate/mm<sup>2</sup> is improved 80% compared to the 7XT Plus series on the same node.{{cn|date=April 2024}} Specific designs have not been announced as of March 2017. Series8XT features 32-wide pipeline clusters.
{| class="wikitable" style="text-align: center;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Clusters
! rowspan="2" |Die Size (mm<sup>2</sup>)
! rowspan="2" |Cluster config{{efn|name=fn4}}
! rowspan="2" |SIMD lane
! colspan="3" |[[Fillrate]] @ 650&nbsp;MHz
! rowspan="2" |Bus width<br />([[bit]])
! rowspan="2" |[[Heterogeneous System Architecture|HSA]]-features
! colspan="6" |[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] @ 650&nbsp;MHz<br>FP32/FP16
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GT8525
| March 2017 
| 2
|
|2/?
|64
|
| colspan="2"| 5.2
|
|
| rowspan="2" | 1.1
| rowspan="2" | 3.2
| rowspan="2" | ?
| rowspan="2" | 1.1
| rowspan="2" | 2.0
| rowspan="2" | ?
|62.4/124.8
|-
| GT8540<ref>{{Cite web|url=https://arstechnica.com/tech-policy/2017/05/imagination-technologies-apple-ip-dispute-resolution-process/|title=Imagination Technologies Can't Resolve Apple IP Spat, Opens Formal Dispute|last=Fiveash|first=Kelly|date=4 May 2017|website=Arstechnica|access-date=8 January 2018|quote=Starting in 2019, Apple will no longer use firm's designs.}}</ref>
| January 2018
| 4
|
|4/?
|128
|
| colspan="2"| 10.4
|
|
|124.8/249.6
|}

=== Series9XE (Rogue) ===
Announced in September 2017, Series9XE family of GPUs benefit from up to 25% Bandwidth savings over the previous generation GPUs. The Series9XE family is targeted for set-top boxes (STB), digital TVs (DTV) and low end smartphones SoCs
Note: Data in table is per cluster.<ref>{{Cite web |date=January 9, 2018 |title=Making the best even better: PowerVR Series9XE and 9XM – the ultimate GPUs for today's embedded platforms |url=https://blog.imaginationtech.com/making-the-best-even-better-the-powervr-series9xe-and-9xm-the-ultimate-gpus-for-todays-embedded-platforms/}}</ref>
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] (FP32)<br>@ 650&nbsp;MHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GE9000
| rowspan="2" | September 2017
| rowspan="2" | 0.25
|
| 16/1
| 
| 
| colspan="2" | 0.65
|
|
| rowspan="6" | 1.1
| rowspan="6" | 3.2
| rowspan="6" | 
| rowspan="6" | 1
| rowspan="6" | 1.2 EP
| rowspan="6" | 
| rowspan="2" | 10.4
|-
| GE9100 
|
| 16/2
| 
|
| colspan="2" rowspan="2" | 1.3
|
|
|-
| GE9115
| January 2018
| rowspan="3" | 0.5
|
| 32/2
| 
|
|
|
| rowspan="3" | 20.8
|-
| GE9210
| September 2017
|
| rowspan="2" | 32/4
| 
|
| colspan="2" rowspan="2" | 2.6
|
| 
|-
| GE9215<ref>https://www.imaginationtech.com/product/ge9215/</ref>
| January 2018
|
| 
|
|
| 
|-
| GE9420 
| September 2017
| 
|
| 
| 
|
| colspan="2" | 
|
|
| 
|}

=== Series9XM (Rogue) ===
The Series9XM family of GPUs achieve up to 50% better performance density than the previous 8XEP generation. The Series9XM family targets mid-range smartphone SoCs.
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] (FP32)<br>@ 650&nbsp;MHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GM9125<ref>{{cite web | url=https://www.imaginationtech.com/product/gm9125/ | title=IMG GM9125 GPU - Imagination | work=Imagination }}</ref>
| rowspan="6" | September 2017
| rowspan="3" | 1
|
| rowspan="3" | 64/4
| 
| 
| colspan="2" | 1.3
|
|
| 1.3
| rowspan="6" | 3.2
| rowspan="6" | 
| rowspan="6" | 1
| 2.0
| rowspan="6" | 
| rowspan="3" | 41.6
|-
| GM9220
|
| 
| 
| rowspan="2" colspan="2" | 2.6
|
|
| 1.1
| rowspan="4" | 1.2 EP
|-
| GM9226<ref>{{cite web | url=https://www.imaginationtech.com/product/gm9226/ | title=IMG GM9226 GPU - Imagination | work=Imagination }}</ref>
|
| 
| 
|
|
| 1.3
|-
| GM9240
| rowspan="3" | 2
|
| rowspan="3" | 128/4
| 
|
| colspan="2" | 2.6
| 
|
| 1.1
| rowspan="3" | 83.2
|-
| GM9446<ref>{{cite web | url=https://www.imaginationtech.com/product/gm9446/ | title=IMG GM9446 GPU - Imagination | work=Imagination }}</ref>
|
| 
|
| colspan="2" | 5.2
|
|
| rowspan="3" | 1.3
|-
| GM9740<ref>{{cite web | url=https://www.imaginationtech.com/product/gm9740/ | title=IMG GM9740 GPU - Imagination | work=Imagination }}</ref>
|
| 
|
| colspan="2" | 2.6
|
|
| 3.0
|}

=== Series9XEP (Rogue) ===
The Series9XEP family of GPUs was announced on December 4, 2018.<ref name=":0">{{Cite news|url=https://pcper.com/2018/12/powervr-9xep-9xmp-and-9xtp-gpus-launched/|title=PowerVR 9XEP, 9XMP, and 9XTP GPUs Launched|website=PC Perspective|date=4 December 2018|language=en-US|access-date=2019-05-30}}</ref> The Series9XEP family supports PVRIC4 image compression.<ref name=":1">{{Cite web |date=2018-10-31 |title=Introducing PVRIC4 – taking image compression to the next level |url=https://blog.imaginationtech.com/introducing-pvric4-taking-image-compression-to-the-next-level/ |access-date=2019-05-30 |website=Imagination |language=en-GB}}</ref> The Series9XEP family targets set-top boxes (STB), digital TVs (DTV) and low end smartphones SoCs.
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] (FP32)<br>@ 650&nbsp;MHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GE9608<ref>{{cite web | url=https://www.imaginationtech.com/product/ge9608/ | title=IMG GE9608 GPU - Imagination | work=Imagination }}</ref>
| rowspan="4" | December 2018
| rowspan="3" | 0.5
|
| rowspan="3" | 32/?
| 
| 
| colspan="2" | 1.3
|
|
| rowspan="4" | 1.3
| rowspan="4" | 3.2
| rowspan="4" | 
| rowspan="4" | 1
| rowspan="4" | 1.2 EP
| rowspan="4" | 
| rowspan="3" | 20.8
|-
| GE9610
|
| 
|
| colspan="2" | ?
|
|
|-
| GE9710
|
| 
|
| colspan="2" | ?
|
|
|-
| GE9920<ref>{{cite web | url=https://www.imaginationtech.com/product/ge9920/ | title=IMG GE9920 GPU - Imagination | work=Imagination }}</ref>
| 1
|
| 64/?
| 
|
| colspan="2" | 5.2
|
|
| 41.6
|-
|}

=== Series9XMP (Rogue) ===
The Series9XMP family of GPUs was announced on December 4, 2018.<ref name=":0" /> The Series9XMP family supports PVRIC4 image compression.<ref name=":1" /> The Series9XMP family targets mid-range smartphone SoCs.
{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] (FP32)<br>@ 650&nbsp;MHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GM9740
| December 2018
| 2
|
| 128/?
| 
| 
| colspan="2"| ?
|
|
| 1.1
| 3.2
| 
| 1
| 1.2 EP
| 
| 83.2
|-
|}

=== Series9XTP (Furian) ===
The Series9XTP family of GPUs was announced on December 4, 2018.<ref name=":0" /> The Series9XTP family supports PVRIC4 image compression.<ref name=":1" /> The Series9XTP family targets high-end smartphone SoCs. Series9XTP features 40-wide pipeline clusters.

{| class="wikitable" style="text-align: center;"
|-
!rowspan=2|Model
!rowspan=2|Date
!rowspan=2|Clusters
!rowspan=2|Die Size (mm<sup>2</sup>)
!rowspan=2|Core config{{efn|name=fn4}}
!rowspan=2|SIMD lane
!colspan=3|[[Fillrate]] @ 650&nbsp;MHz
!rowspan=2|Bus width<br />([[bit]])
!rowspan=2|[[Heterogeneous System Architecture|HSA]]-features
!colspan=6|[[Application programming interface|API]] (version)
! rowspan="2" | [[FLOPS|GFLOPS]] (FP32)<br>@ 650&nbsp;MHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| GT9524<ref>{{cite web | url=https://www.imaginationtech.com/product/gt9524/ | title=IMG GT9524 GPU - Imagination | work=Imagination }}</ref>
| December 2018
| ?
|
| ?/?
| 
| 
| colspan="2" | 5.2
|
|
| 1.3
| 3.x
| 
| 1
| 3.0
| 
| 156
|-
|}

=== IMG A-Series (Albiorix) ===
The A-Series GPUs offer up to 250% better performance density than the previous Series 9. These GPUs are no longer called PowerVR, they are called IMG.<ref>{{Cite web |title=IMG A-Series GPU |url=https://www.imaginationtech.com/products/gpu/img-a-series-gpu/ |access-date=2020-01-04 |website=Imagination |language=en-GB}}</ref>
{| class="wikitable sortable" style="text-align: center;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Clusters
! rowspan="2" |Die Size (mm<sup>2</sup>)
! rowspan="2" |Core config{{efn|name=fn4}}
! rowspan="2" |SIMD lane
! colspan="3" |[[Fillrate]]<br>@ 1&nbsp;GHz
! rowspan="2" |Bus width<br />([[bit]])
! rowspan="2" |[[Heterogeneous System Architecture|HSA]]-features
! colspan="6" |[[Application programming interface|API]] (version)
! rowspan="2" |[[FLOPS|GFLOPS (FP32)]]<br>@ 1&nbsp;GHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenGL]]
![[OpenVX]]
![[OpenCL]]
![[Direct3D]]
|-
|-
| IMG AXE-1-16<ref>{{Cite web |date=2019 |title=IMG AXE-1-16 GPU |url=https://www.imaginationtech.com/product/img-axe-1-16/ |access-date=3 January 2020 |publisher=Imagination Technologies Limited}}</ref>
| rowspan="7" | December 2019
| ?
|
| rowspan="3" | 1/?
| 
| 
| colspan="2" | 1
|
|
| rowspan="7" | 1.3
| rowspan="7" | 3.x
| rowspan="7" | ?
| rowspan="7" | ?
| rowspan="7" | 3.0
| rowspan="7" | ?
| 16
|-
|IMG AXE-2-16<ref>{{Cite web |title=Find out about the PowerVR IMG AXE-2-16 embedded GPU IP Core |url=https://www.imaginationtech.com/product/img-axe-2-16/ |access-date=2020-01-04 |website=Imagination |language=en-GB}}</ref>
|?
|
|
|
| colspan="2" | 2
|
|
|16
|-
| IMG AXM-8-256<ref>{{Cite web |title=Find out about the PowerVR IMG AXM-8-256 embedded GPU IP Core |url=https://www.imaginationtech.com/product/img-axm-8-256/ |access-date=2020-01-04 |website=Imagination |language=en-GB}}</ref>
|1
|
| 
|
| colspan="2" | 8
|
|
|256
|-
|IMG AXT-16-512<ref>{{Cite web |title=Find out about the PowerVR IMG AXT-16-512 embedded GPU IP Core |url=https://www.imaginationtech.com/product/img-axt-16-512/ |access-date=2020-01-04 |website=Imagination |language=en-GB}}</ref>
|2
|
|2/?
|
|
| colspan="2" | 16
|
|
|512
|-
|IMG AXT-32-1024<ref>{{Cite web |title=Find out about the PowerVR IMG AXT-32-1024 embedded GPU IP Core |url=https://www.imaginationtech.com/product/img-axt-32-1024/ |access-date=2020-01-04 |website=Imagination |language=en-GB}}</ref>
|4
|
|4/?
|
|
| colspan="2" | 32
|
|
|1024
|-
|IMG AXT-48-1536
|6
|
|6/?
|
|
| colspan="2" | 48
|
|
|1536
|-
|IMG AXT-64-2048
|8
|
|8/?
|
|
| colspan="2" | 64
|
|
|2048
|}

=== IMG B-Series ===
The B-Series GPUs offer up to 25% lower die space and 30% lower power than the previous A-Series.
{| class="wikitable" style="text-align: center;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Clusters
! rowspan="2" |Die Size (mm<sup>2</sup>)
! rowspan="2" |Core config{{efn|name=fn4}}
! rowspan="2" |SIMD lane
! colspan="3" |[[Fillrate]]<br>@ 1&nbsp;GHz
! rowspan="2" |Bus width<br />([[bit]])
! rowspan="2" |[[Heterogeneous System Architecture|HSA]]-features
! colspan="3" |[[Application programming interface|API]] (version)
! rowspan="2" |[[FLOPS|GFLOPS (FP32)]]<br>@ 1&nbsp;GHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenCL]]
|-
|-
| IMG BXE-1-16<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxe-1-16/ | title=IMG BXE-1-16 - Imagination | work=Imagination }}</ref>
| rowspan="33" | October 2020
|
|
| rowspan="3" | 1/?
|
|
| colspan="2" | 1
|
|
| rowspan="33" | 1.3
| rowspan="33" | 3.x
| rowspan="33" | 3.0
| 16
|-
|IMG BXE-2-32<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxe-2-32/ | title=IMG BXE-2-32 - Imagination | work=Imagination }}</ref>
|
|
|
|
| colspan="2" | 2
|
|
| rowspan="2" | 32
|-
| IMG BXE-4-32
|
|
|
|
| colspan="2" | 4
|
|
|-
|IMG BXE-4-32 MC2<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxe-4-32-mc2/ | title=IMG BXE-4-32 MC2 - Imagination | work=Imagination }}</ref>
|
|
| 2/?
|
|
| colspan="2" | 8
|
|
| 64
|-
|IMG BXE-4-32 MC3
|
|
| 3/?
|
|
| colspan="2" | 12
|
|
| 96
|-
|IMG BXE-4-32 MC4
|
|
| 4/?
|
|
| colspan="2" | 16
|
|
| 128
|-
|IMG BXM-4-64 MC1<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxm-4-64-mc1/ | title=IMG BXM-4-64 MC1 - Imagination | work=Imagination }}</ref>
|
|
| 1/?
|
|
| colspan="2" | 4
|
|
| 64
|-
|IMG BXM-4-64 MC2<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxm-4-64-mc2/ | title=IMG BXM-4-64 MC2 - Imagination | work=Imagination }}</ref>
|
|
| 2/?
|
|
| colspan="2" | 8
|
|
| 128
|-
|IMG BXM-4-64 MC3
|
|
| 3/?
|
|
| colspan="2" | 12
|
|
| 192
|-
|IMG BXM-4-64 MC4
|
|
| 4/?
|
|
| colspan="2" | 16
|
|
| rowspan="2" | 256
|-
|IMG BXM-8-256<ref>https://www.imaginationtech.com/product/img-bxm-8-256/</ref>
| 1
|
| 1/?
|
|
| colspan="2" | 8
|
|
|-
|IMG BXS-1-16
|
|
|
|
|
| colspan="2" | 1
|
|
| 16
|-
|IMG BXS-2-32
|
|
|
|
|
| colspan="2" | 2
|
|
| 32
|-
|IMG BXS-2-32 MC2
|
|
|
|
|
| rowspan="2" colspan="2" | 4
|
|
| 64
|-
|IMG BXS-4-32 MC1
|
|
|
|
|
|
|
| 32
|-
|IMG BXS-4-32 MC2
|
|
|
|
|
| colspan="2" | 8
|
|
| 64
|-
|IMG BXS-4-32 MC3
|
|
|
|
|
| colspan="2" | 12
|
|
| 96
|-
|IMG BXS-4-32 MC4
|
|
|
|
|
| colspan="2" | 16
|
|
| 128
|-
|IMG BXS-4-64 MC1<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxs-4-64-mc1/ | title=IMG BXS-4-64 MC1 - Imagination | work=Imagination }}</ref>
|
|
|
|
|
| colspan="2" | 4
|
|
| 64
|-
|IMG BXS-4-64 MC2<ref>https://www.imaginationtech.com/product/img-bxs-4-64-mc2/</ref>
|
|
|
|
|
| colspan="2" | 8
|
|
| 128
|-
|IMG BXS-4-64 MC3
|
|
|
|
|
| colspan="2" | 12
|
|
| 192
|-
|IMG BXS-4-64 MC4
|
|
|
|
|
| colspan="2" | 16
|
|
| rowspan="2" | 256
|-
|IMG BXS-8-256<ref>https://www.imaginationtech.com/product/img-bxs-8-256/</ref>
| 1
|
|
|
|
| colspan="2" | 8
|
|
|-
|IMG BXS-16-512
| 2
|
|
|
|
| colspan="2" | 16
|
|
| 512
|-
|IMG BXS-32-1024 MC1<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxs-32-1024-mc1/ | title=IMG BXS-32-1024 MC1 - Imagination | work=Imagination }}</ref>
| 4
|
|
|
|
| colspan="2" | 32
|
|
| 1024
|-
|IMG BXS-32-1024 MC2<ref>https://www.imaginationtech.com/product/img-bxs-32-1024-mc2/</ref>
| 8
|
|
|
|
| colspan="2" | 64
|
|
| 2048
|-
|IMG BXS-32-1024 MC3
| 12
|
|
|
|
| colspan="2" | 96
|
|
| 3072
|-
|IMG BXS-32-1024 MC4
| 16
|
|
|
|
| colspan="2" | 128
|
|
| 4096
|-
|IMG BXT-16-512
| 2
|
| 2/?
|
|
| colspan="2" | 16
|
|
| 512
|-
|IMG BXT-32-1024 MC1<ref>https://www.imaginationtech.com/product/img-bxt-32-1024-mc1/</ref>
| 4
|
| 4/?
|
|
| colspan="2" | 32
|
|
| 1024
|-
|IMG BXT-32-1024 MC2<ref>{{cite web | url=https://www.imaginationtech.com/product/img-bxt-32-1024-mc2/ | title=IMG BXT-32-1024 MC2 - Imagination | work=Imagination }}</ref>
| 8
|
| 8/?
|
|
| colspan="2" | 64
|
|
| 2048
|-
|IMG BXT-32-1024 MC3
| 12
|
| 12/?
|
|
| colspan="2" | 96
|
|
| 3072
|-
|IMG BXT-32-1024 MC4
| 16
|
| 16/?
|
|
| colspan="2" | 128
|
|
| 4096
|}

=== IMG C-Series (Photon) ===
Imagination Technologies announced on the 4th of November 2021 the new C-series GPU architecture.<ref>{{Cite web|title=company blog post|work=Imagination |date=4 November 2021|url=https://imaginationtech.com/news/imagination-launches-the-most-advanced-ray-tracing-gpu/ |author1=93Digital }}</ref>

{| class="wikitable" style="text-align: center;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Clusters
! rowspan="2" |Die Size (mm<sup>2</sup>)
! rowspan="2" |Core config{{efn|name=fn4}}
! rowspan="2" |SIMD lane
! colspan="3" |[[Fillrate]] @ 1&nbsp;GHz
! rowspan="2" |Bus width<br />([[bit]])
! rowspan="2" |[[Heterogeneous System Architecture|HSA]]-features
! colspan="3" |[[Application programming interface|API]] (version)
! rowspan="2" |[[FLOPS|GFLOPS (FP32)]]<br>@ 1&nbsp;GHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenCL]]
|-
|-
| IMG CXM-2-64<ref>{{cite web | url=https://www.imaginationtech.com/product/img-cxm-2-64/ | title=IMG CXM-2-64 - Imagination | work=Imagination }}</ref>
| rowspan="3" | November 2021
|
|
| rowspan="2" | 1/?
|
|
| colspan="2" | 2
|
|
| rowspan="3" | 1.3
| rowspan="3" | 3.x
| rowspan="3" | 3.0
| rowspan="2" | 64
|-
|IMG CXM-4-64<ref>https://www.imaginationtech.com/product/img-cxm-4-64/</ref>
|
|
|
|
| colspan="2" rowspan="2" | 4
|
|
|-
|IMG CXM-4-128<ref>{{cite web | url=https://www.imaginationtech.com/product/img-cxm-4-128/ | title=IMG CXM-4-128 - Imagination | work=Imagination }}</ref>
|
|
| 2/?
|
|
|
|
| 128
|}

=== IMG D-Series ===
Imagination Technologies announced on the 11th of January 2023 the new D-series GPU architecture.<ref>{{cite web | url=https://www.imaginationtech.com/news/imaginations-img-dxt-gpu-unlocks-scalable-premium-ray-tracing-for-all-mobile-gamers/ | title=Imagination's IMG DXT GPU unlocks scalable, premium ray tracing for all mobile gamers - Imagination | work=Imagination | date=11 January 2023 | last1=Balan | first1=Cosmin }}</ref>

{| class="wikitable" style="text-align: center;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Clusters
! rowspan="2" |Die Size (mm<sup>2</sup>)
! rowspan="2" |Core config{{efn|name=fn4}}
! rowspan="2" |SIMD lane
! colspan="3" |[[Fillrate]] @ 1&nbsp;GHz
! rowspan="2" |Bus width<br />([[bit]])
! rowspan="2" |[[Heterogeneous System Architecture|HSA]]-features
! colspan="3" |[[Application programming interface|API]] (version)
! rowspan="2" |[[FLOPS|GFLOPS (FP32)]]<br>@ 1&nbsp;GHz
|-
!MPolygons/s
! ([[Pixel|GP]]/s)
! ([[Texel (graphics)|GT]]/s)
![[Vulkan (API)]]
![[OpenGL ES]]
![[OpenCL]]
|-
|-
| IMG DXT-8-256<ref>{{cite web | url=https://www.imaginationtech.com/product/img-dxt-8-256/ | title=IMG DXT-8-256 - Imagination | work=Imagination }}</ref>
| rowspan="7" | January 2023
|
|
| 1/?
|
|
| colspan="2" | 8
|
|
| rowspan="7" | 1.3
| rowspan="7" | 3.x
| rowspan="7" | 3.0
| 256
|-
|IMG DXT-48-1536 0.5RT2<ref>{{cite web | url=https://www.imaginationtech.com/product/img-dxt-48-1536/ | title=IMG DXT-48-1536 0.5RT2 - Imagination | work=Imagination }}</ref>
|
|
| rowspan="2" | ?/?
|
|
| colspan="2" rowspan="2" | 48
|
|
| rowspan="2" | 1536
|-
|IMG DXTP-48-1536<ref>{{cite web | url=https://www.imaginationtech.com/product/dxtp-48-1536-gpu/ | title=Imagination DXTP-48-1536 GPU - Imagination | work=Imagination }}</ref>
|
|
| 
|
|
|
|-
|IMG DXTP-64-2048<ref>{{cite web | url=https://www.imaginationtech.com/product/dxtp-64-2048-gpu/ | title=Imagination DXTP-64-2048 GPU - Imagination | work=Imagination }}</ref>
|
|
| ?/?
|
|
| colspan="2"| 64
|
|
| 2048
|-
|IMG DXT-72-2304 RT3<ref>{{cite web | url=https://www.imaginationtech.com/product/img-dxt-72-2304/ | title=IMG DXT-72-2304 RT3 - Imagination | work=Imagination }}</ref>
|
|
| rowspan="2" | ?/?
|
|
| colspan="2" rowspan="2" | 72
|
|
| rowspan="2" | 2304
|-
|IMG DXD-72-2304 MC1<ref>{{cite web | url=https://www.imaginationtech.com/product/img-dxd-72-2304-mc1/ | title=IMG DXD-72-2304 MC1 - Imagination | work=Imagination }}</ref>
|
|
|
|
|
|
|-
|IMG DXD-72-2304 MC2<ref>{{cite web | url=https://www.imaginationtech.com/product/img-dxd-72-2304-mc2/ | title=IMG DXD-72-2304 MC2 - Imagination | work=Imagination }}</ref>
|
|
| ?/?
|
|
| colspan="2"| 144
|
|
| 4608
|}

'''''Notes'''''
{{notelist}}
* All models support Tile based deferred rendering (TBDR)

== PowerVR Vision & AI ==

=== Series2NX ===
The Series2NX family of [[AI accelerator|Neural Network Accelerators (NNA)]] was announced on September 21, 2017.

Series2NX core options:
{| class="wikitable" style="text-align: center;"
|-
! Model
! Date
! Engines
! 8-bit TOPS
! 16-bit TOPS
! 8-bit MACs
! 16-bit MACs
! [[FLOPS|APIs]]
|-
| AX2145<ref>{{Cite web |title=PowerVR AX2145 Neural Network Accelerator (NNA) IP Core |url=https://www.imaginationtech.com/products/ai/ |access-date=2019-05-30 |website=Imagination |language=en-GB}}</ref>
| rowspan="2" | September 2017
|?
| 1
|0.5
|512/clk
|256/clk
| rowspan="2" | IMG DNN
Android NN
|-
| AX2185<ref>{{Cite web |title=PowerVR AX2185 Neural Network Accelerator (NNA) IP Core |url=https://www.imaginationtech.com/products/ai/ |access-date=2019-05-30 |website=Imagination |language=en-GB}}</ref>
|8
| 4.1
|2.0
|2048/clk
|1024/clk
|}

=== Series3NX ===
The Series3NX family of [[AI accelerator|Neural Network Accelerators (NNA)]] was announced on December 4, 2018.<ref name=":2">{{Cite web|url=https://www.anandtech.com/show/13673/imagination-announces-powervr-series3nx-neural-network-accelerator|title=Imagination Goes Further Down the AI Rabbit Hole, Unveils PowerVR Series3NX Neural Network Accelerator|last=Oh|first=Nate|website=www.anandtech.com|access-date=2019-05-30}}</ref>

Series3NX core options:
{| class="wikitable" style="text-align: center;"
|-
! Model
! Date
! Engines
! 8-bit TOPS
! 16-bit TOPS
! 8-bit MACs
! 16-bit MACs
! [[FLOPS|APIs]]
|-
|AX3125
| rowspan="5" |December 2018
|?
|0.6
|?
|256/clk
|64/clk
| rowspan="5" |IMG DNN
Android NN
|-
| AX3145
|?
| 1.2
|?
|512/clk
|128/clk
|-
| AX3365
|?
| 2.0
|?
|1024/clk
|256/clk
|-
|AX3385
|?
|4.0
|?
|2048/clk
|512/clk
|-
|AX3595
|?
|10.0
|?
|4096/clk
|1024/clk
|}
Series3NX multi-core options
{| class="wikitable" style="text-align: center;"
|-
! Model
! Date
! Cores
! 8-bit TOPS
! 16-bit TOPS
! 8-bit MACs
! 16-bit MACs
! [[FLOPS|APIs]]
|-
|UH2X40
| rowspan="4" |December 2018
|2
|20.0
|?
|8192/clk
|2048/clk
| rowspan="4" |IMG DNN
Android NN
|-
| UH4X40
|4
| 40.0
|?
|16384/clk
|4096/clk
|-
| UH8X40
|8
| 80.0
|?
|32768/clk
|8192/clk
|-
|UH16X40
|16
|160.0
|?
|65536/clk
|16384/clk
|}

==== Series3NX-F ====
The Series3NX-F family of [[AI accelerator|Neural Network Accelerators (NNA)]] was announced alongside the Series3NX family. The Series3NX-F family combines the Series 3NX with a Rogue-based GPGPU (NNPU), and local RAM. This allows support for programmability and floating-point.<ref name=":2" />

== Implementations ==
The PowerVR [[Graphics processing unit|GPU]] variants can be found in the following table of systems on chips ([[System on a chip|SoC]]). Implementations of PowerVR accelerators in products are listed [[List of PowerVR products|here]].

{| class="wikitable" style="text-align: center"
|-
! Vendor 
!Date!! [[System on a chip|SOC]] name !! PowerVR chipset !! [[Clock rate|Frequency]]!! [[FLOPS|GFLOPS]] (FP16)
|- 
| rowspan="16" |[[Texas Instruments]]
|
| OMAP 3420
| rowspan="16" |[[#Series5 (SGX)|SGX530]]
| ?
| ?
|-
|
| OMAP 3430
| ?
| ?
|-
|
| OMAP 3440
| ?
| ?
|-
|
| OMAP 3450
| ?
| ?
|-
|
| OMAP 3515
| ?
| ?
|-
|
| OMAP 3517
| ?
| ?
|-
|
| OMAP 3530
| 110&nbsp;MHz
|  0.88
|-
|
| OMAP 3620
| ?
| ?
|-
|
| OMAP 3621
| ?
| ?
|-
|
| OMAP 3630
| ?
| ?
|-
|
| OMAP 3640
| ?
| ?
|-
|
| Sitara AM335x<ref>{{Cite web |title=AM3358 data sheet, product information and support {{!}} TI.com |url=https://www.ti.com/product/AM3358 |access-date=2025-01-01 |website=www.ti.com}}</ref>
| 200&nbsp;MHz
| 1.6
|-
|
| Sitara AM3715
| ?
| ?
|-
|
| Sitara AM3891
| ?
| ?
|-
|
| DaVinci DM3730
| 200&nbsp;MHz
| 1.6
|-
|
| Integra C6A8168
| ?
| ?
|-
| [[NEC]]
|
| EMMA Mobile/EV2
|[[#Series5 (SGX)|SGX530]]
| ?
| ?
|-
| rowspan="2" | [[Renesas Electronics|Renesas]]
|
| SH-Mobile G3
| rowspan="2" |[[#Series5 (SGX)|SGX530]]
| ?
| ?
|-
|
| SH-Navi3 (SH7776)
|?
|?
|-
| rowspan="2" | [[Sigma Designs]]
|
| SMP8656
| rowspan="2" |[[#Series5 (SGX)|SGX530]]
| ?
| ?
|-
|
| SMP8910
|?
|?
|-
| rowspan="3" | [[MediaTek]]
|
| MT6513
| rowspan="3" |[[#Series5 (SGX)|SGX531]]
| rowspan="3" | 281&nbsp;MHz
| rowspan="3" | 2.25
|-
|2010
| MT6573
|-
|2012
| MT6575M
|-
| rowspan="3" | [[Trident Microsystems|Trident]]
|
| PNX8481
| rowspan="3" |[[#Series5 (SGX)|SGX531]]
| ?
| ?
|-
|
| PNX8491
|?
|?
|-
|
| HiDTV PRO-SX5
|?
|?
|-
| rowspan="9" | [[MediaTek]]
|
| MT6515
| rowspan="9" |[[#Series5 (SGX)|SGX531]]
| rowspan="9" | 522&nbsp;MHz
| rowspan="9" | 4.2
|-
|2011
| MT6575
|-
|
| MT6517
|-
|
| MT6517T
|-
|2012
| MT6577
|-
|
| MT6577T
|-
|
| MT8317
|-
|
| MT8317T
|-
|
| MT8377
|-
| rowspan="2" | [[NEC]]
|
| NaviEngine EC-4260
| rowspan="2" |[[#Series5 (SGX)|SGX535]]
| rowspan="2" |?
| rowspan="2" |?
|-
|
|NaviEngine EC-4270
|-
| rowspan="8" | [[Intel]]
|
| CE 3100 (Canmore)
| rowspan="8" |[[#Series5 (SGX)|SGX535]]
| ?
| ?
|-
|
| SCH US15/W/L (Poulsbo)
| ?
| ?
|-
|
| CE4100 (Sodaville)
| ?
| ?
|-
|
| CE4110 (Sodaville)
| rowspan="2" | 200&nbsp;MHz
| rowspan="2" | 1.6
|-
|
| CE4130 (Sodaville)
|-
|
| CE4150 (Sodaville)
| rowspan="3" |400&nbsp;MHz
| rowspan="3" |3.2
|-
|
| CE4170 (Sodaville)
|-
|
| CE4200 (Groveland)
|-
| [[Samsung]]
|
| APL0298C05
|[[#Series5 (SGX)|SGX535]]
| ?
| ?
|-
| rowspan="2" | [[Apple Inc.|Apple]]
| rowspan="2" |April 3, 2010
| [[Apple A4]] ([[iPhone 4]])
| rowspan="2" |[[#Series5 (SGX)|SGX535]]
| 200&nbsp;MHz
| 1.6
|-
| [[Apple A4]] ([[IPad (1st generation)|iPad]])
| 250&nbsp;MHz
| 2.0
|-
| [[Ambarella Inc.|Ambarella]]
|
| iOne
|[[#Series5 (SGX)|SGX540]]
| ?
| ?
|-
| rowspan="3" |[[Renesas Electronics|Renesas]]
|
| SH-Mobile G4
| rowspan="3" |[[#Series5 (SGX)|SGX540]]
| ?
| ?
|-
|
| SH-Mobile APE4 (R8A73720)
|?
|?
|-
|
| R-Car E2 (R8A7794)
|?
|?
|-
| [[Ingenic Semiconductor]]
|
| JZ4780
|[[#Series5 (SGX)|SGX540]]
| ?
| ?
|-
| rowspan="4" | [[Samsung]]
|2010
| Exynos 3110
| rowspan="4" |[[#Series5 (SGX)|SGX540]]
| rowspan="3" | 200&nbsp;MHz
| rowspan="3" | 3.2
|-
|2010
| S5PC110
|-
|
| S5PC111
|-
|
| S5PV210
|?
|?
|-
| rowspan="2" | [[Texas Instruments]]
| rowspan="2" |Q1 2011
| OMAP 4430
| rowspan="2" |[[#Series5 (SGX)|SGX540]]
| 307&nbsp;MHz
| 4.9
|-
| OMAP 4460
| 384&nbsp;MHz
| 6.1
|-
| [[Intel]]
|Q1 2013
| Atom Z2420
|[[#Series5 (SGX)|SGX540]]
| 400&nbsp;MHz
| 6.4
|-
| rowspan="3" | [[Actions Semiconductor]]
|
| ATM7021
| rowspan="3" |[[#Series5 (SGX)|SGX540]]
| rowspan="3" |  500&nbsp;MHz
| rowspan="3" |  8.0
|-
|
| ATM7021A
|-
|
| ATM7029B
|-
| [[Rockchip]]
|
| RK3168
|[[#Series5 (SGX)|SGX540]]
| 600&nbsp;MHz
| 9.6
|-
| rowspan="3" | [[Apple Inc.|Apple]]
|November 13, 2014 
| [[Apple S1]] ([[Apple Watch (1st generation)]])
|[[#Series5XT (SGX)|SGX543]]
| ?
| ?
|-
|March 11, 2011
| [[Apple A5]] ([[iPhone 4S]], [[iPod Touch (5th generation)]])
| rowspan="2" |[[#Series5XT (SGX)|SGX543]] MP2
| 200&nbsp;MHz
| 12.8
|-
|March 2012
| [[Apple A5]] ([[iPad 2]], [[IPad Mini (1st generation)|iPad mini]])
| 250&nbsp;MHz
| 16.0
|-
| [[MediaTek]]
|
| MT5327
|[[#Series5XT (SGX)|SGX543]] MP2
| 400&nbsp;MHz
| 25.6
|-
| [[Renesas Electronics|Renesas]]
|
| R-Car H1 (R8A77790)
|[[#Series5XT (SGX)|SGX543]] MP2
| ?
| ?
|-
| rowspan="2" | [[Apple Inc.|Apple]] 
|September 12, 2012
| [[Apple A6]] ([[iPhone 5]], [[iPhone 5C]])
|[[#Series5XT (SGX)|SGX543]] MP3
| rowspan="2" | 250&nbsp;MHz
| 24.0
|-
|March 7, 2012
| [[Apple A5X]] ([[iPad (3rd generation)]])
|[[#Series5XT (SGX)|SGX543]] MP4
| 32.0
|-
| [[Sony]] 
|
| CXD53155GG ([[PlayStation Vita|PS Vita]])
|[[#Series5XT (SGX)|SGX543]] MP4+
| 41-222&nbsp;MHz
| 5.248-28.416
|-
| [[ST-Ericsson]]
|
| Nova A9540
| rowspan="4" |[[#Series5XT (SGX)|SGX544]]
| ?
| ?
|-
|
|
| NovaThor L9540
| ?
| ?
|-
|
|
| NovaThor L8540
| 500&nbsp;MHz
| 16
|-
|
|
| NovaThor L8580
| 600&nbsp;MHz
| 19.2
|-
| rowspan="7" | [[MediaTek]]
|July 2013
| MT6589M
| rowspan="7" |[[#Series5XT (SGX)|SGX544]]
| rowspan="3" | 156&nbsp;MHz
| rowspan="3" | 5
|-
|
| MT8117
|-
|
| MT8121
|-
|March 2013
| MT6589
| rowspan="2" | 286&nbsp;MHz
| rowspan="2" | 9.2
|-
|
| MT8389
|-
|
| MT8125
| 300&nbsp;MHz
| 9.6
|-
|July 2013
| MT6589T
| 357&nbsp;MHz
| 11.4
|-
| [[Texas Instruments]]
|Q2 2012
| OMAP 4470
|[[#Series5XT (SGX)|SGX544]]
| 384&nbsp;MHz
| 13.8
|-
| rowspan="2" | [[Broadcom]]
|
| Broadcom M320
| rowspan="2" |[[#Series5XT (SGX)|SGX544]]
| rowspan="2" | ?
| rowspan="2" | ?
|-
|
| Broadcom M340
|-
| [[Actions Semiconductor]]
|
| ATM7039
|[[#Series5XT (SGX)|SGX544]]
| 450&nbsp;MHz
| 16.2
|-
| rowspan="2" | [[Allwinner Technology|Allwinner]]
|
| Allwinner A31
| rowspan="2" |[[#Series5XT (SGX)|SGX544]] MP2
| rowspan="2" | 300&nbsp;MHz
| rowspan="2" | 19.2
|-
|
| Allwinner A31S
|-
| rowspan="3" | [[Intel]]
| rowspan="3" |Q2 2013
| Atom Z2520
| rowspan="3" |[[#Series5XT (SGX)|SGX544]] MP2
| 300&nbsp;MHz
| 21.6
|-
| Atom Z2560
| 400&nbsp;MHz
| 25.6
|-
| Atom Z2580
| 533&nbsp;MHz
| 34.1
|-
| rowspan="3" |[[Texas Instruments]]
| rowspan="2" |Q2 2013
| OMAP 5430
| rowspan="2" |[[#Series5XT (SGX)|SGX544]] MP2
| rowspan="2" | 533&nbsp;MHz
| rowspan="2" | 34.1
|-
| OMAP 5432
|-
| Q4 2018
| Sitara AM6528<br />Sitara AM6548
| [[#Series5XT (SGX)|SGX544]]
| 
|
|-
| rowspan="2" | [[Allwinner Technology|Allwinner]]
|
| Allwinner A83T
| rowspan="2" |[[#Series5XT (SGX)|SGX544]] MP2
| rowspan="2" | 700&nbsp;MHz
| rowspan="2" | 44.8
|-
|
| Allwinner H8
|-
| [[Exynos|Samsung]]
|Q2 2013
| Exynos 5410
|[[#Series5XT (SGX)|SGX544]] MP3
| 533&nbsp;MHz
| 51.1
|-
| rowspan="11" | [[Intel]]
|
| Atom Z2460
| rowspan="11" |[[#Series5XT (SGX)|SGX545]]
| rowspan="2" | 533&nbsp;MHz
| rowspan="2" | 8.5
|-
|
| Atom Z2760
|-
|
| Atom CE5310
| ?
| ?
|-
|
| Atom CE5315
| ?
| ?
|-
|
| Atom CE5318
| ?
| ?
|-
|
| Atom CE5320
| ?
| ?
|-
|
| Atom CE5328
| ?
| ?
|-
|
| Atom CE5335
| ?
| ?
|-
|
| Atom CE5338
| ?
| ?
|-
|
| Atom CE5343
| ?
| ?
|-
|
| Atom CE5348
| ?
| ?
|-
| [[Apple Inc.|Apple]]
|October 23, 2012
| [[Apple A6X]] ([[iPad (4th generation)]])
|[[#Series5XT (SGX)|SGX554]] MP4
| 300&nbsp;MHz
| 76.8
|-
|[[Apple Inc.|Apple]]
|September, 2016
|[[Apple S1P]] ([[Apple Watch Series 1]]), [[Apple S2]] ([[Apple Watch Series 2]])
|Series6 ([[#Series6XE (Rogue)|G6050]] ?)
|?
|?
|-
| [[Rockchip]]
|
| RK3368
|[[#Series6XE (Rogue)|G6110]]
| 600&nbsp;MHz
| 38.4
|-
| rowspan="6" | [[MediaTek]]
|Q1 2014
| MT6595M
| rowspan="6" |[[#Series6 (Rogue)|G6200]] (2 Clusters)
| rowspan="2" | 450&nbsp;MHz
| rowspan="2" | 57.6
|-
|
| MT8135
|-
| rowspan="2" |Q4 2014
|Helio X10 (MT6795M)
|550&nbsp;MHz
|70.4
|-
|Helio X10 (MT6795T)
|
|
|-
| rowspan="2" |Q1 2014
| MT6595
| 600&nbsp;MHz
| 76.8
|-
| MT6795
| 700&nbsp;MHz
| 89.5
|-
| [[LG Corporation|LG]]
|Q1 2012
| LG H13
|[[#Series6 (Rogue)|G6200]] (2 Clusters)
| 600&nbsp;MHz
| 76.8
|-
| rowspan="2" | [[Allwinner Technology|Allwinner]]
|
| Allwinner A80
| rowspan="2" |[[#Series6 (Rogue)|G6230]] (2 Clusters)
| rowspan="2" | 533&nbsp;MHz
| rowspan="2" | 68.0
|-
|
| Allwinner A80T
|-
| [[Actions Semiconductor]]
|
| ATM9009
|[[#Series6 (Rogue)|G6230]] (2 Clusters)
| 600&nbsp;MHz
| 76.8
|-
| rowspan="2" | [[MediaTek]]
|Q1 2015
| MT8173
| rowspan="2" |[[#Series6XT (Rogue)|GX6250]] (2 Clusters)
| 700&nbsp;MHz
| 89.6
|-
|Q1 2016
| MT8176
| 600&nbsp;MHz
| 76.8
|-
| rowspan="2" | [[Intel]]
| rowspan="2" |Q1 2014
| Atom Z3460
| rowspan="2" |[[#Series6 (Rogue)|G6400]] (4 Clusters)
| rowspan="2" | 533&nbsp;MHz
| rowspan="2" | 136.4
|-
| Atom Z3480
|-
| rowspan="2" | [[Renesas Electronics|Renesas]]
|
| R-Car H2 (R8A7790x)
|[[#Series6 (Rogue)|G6400]] (4 Clusters)
| rowspan="2" | 600&nbsp;MHz
| 153.6
|-
|
| R-Car H3 (R8A7795)
|[[#Series6XT (Rogue)|GX6650]] (6 Clusters)
| 230.4
|-
| [[Apple Inc.|Apple]]
|September 10, 2013
| [[Apple A7]] ([[iPhone 5S]], [[iPad Air]], [[iPad mini 2]], [[iPad mini 3]])
|[[#Series6 (Rogue)|G6430]] (4 Clusters)
| 450&nbsp;MHz
| 115.2
|-
| rowspan="4" | [[Intel]]
| rowspan="2" |Q2 2014
| Atom Z3530
| rowspan="4" |[[#Series6 (Rogue)|G6430]] (4 Clusters)
| 457&nbsp;MHz
| 117
|-
| Atom Z3560
| rowspan="3" |533&nbsp;MHz
| rowspan="3" |136.4
|-
|Q3 2014
| Atom Z3570
|-
|Q2 2014
| Atom Z3580
|-
| rowspan="5" | [[Apple Inc.|Apple]]
|September 9, 2014
| [[Apple A8]] ([[iPhone 6]] / [[iPhone 6 Plus|6 Plus]], [[iPad mini 4]], [[Apple TV (4th generation)|Apple TV HD]], 
[[iPod Touch (6th generation)]])
|[[#Series6XT (Rogue)|GX6450]] (4 Clusters)
| rowspan="2" | 533&nbsp;MHz
| 136.4
|-
|October 16, 2014
| [[Apple A8X]] ([[iPad Air 2]])
|[[#Series6XT (Rogue)|GX6850]] (8 Clusters)
| 272.9
|-
| rowspan="2" |September 9, 2015
| [[Apple A9]] ([[iPhone 6S]] / [[iPhone 6S Plus|6S Plus]], [[iPhone SE (1st generation)]], [[iPad (5th generation)]])
| Series7XT [[#Series7XT (Rogue)|GT7600]] (6 Clusters)
| 600&nbsp;MHz
| 230.4
|-
| [[Apple A9X]] ([[iPad Pro (9.7-inch)]], [[iPad Pro (12.9-inch)]])
| Series7XT [[#Series7XT (Rogue)|GT7800]] (12 Clusters)
| >652&nbsp;MHz
| >500<ref>{{Citation|last=Apple|title=Apple - March Event 2016|date=2016-03-23|url=https://www.youtube.com/watch?v=0wIiDnjz4X4&t=50m45s |archive-url=https://ghostarchive.org/varchive/youtube/20211212/0wIiDnjz4X4| archive-date=2021-12-12 |url-status=live|access-date=2017-09-29}}{{cbignore}}</ref>
|-
|September 7, 2016
| [[Apple A10|Apple A10 Fusion]] ([[iPhone 7]] / [[iPhone 7 Plus|7 Plus]] & [[iPad (6th generation)]])
| Series7XT [[#Series7XT Plus (Rogue)|GT7600 Plus]] (6 Clusters)
| 900&nbsp;MHz
| 345.6
|-
|[[Spreadtrum]]
|2017
|SC9861G-IA
|Series7XT [[#Series7XT (Rogue)|GT7200]]
|
|
|-
|[[MediaTek]]
|Q1 2017
|Helio X30 (MT6799)
|Series7XT [[#Series7XT Plus (Rogue)|GT7400 Plus]] (4 Clusters)
|800&nbsp;MHz
|204.8
|-
|[[Apple Inc.|Apple]]
|June 5, 2017
|[[Apple A10X]] ([[iPad Pro (10.5-inch)]], [[iPad Pro (12.9-inch) (2nd generation)]], [[Apple TV 4K]])
|Series7XT [[#Series7XT Plus (Rogue)|GT7600 Plus]] (12 Clusters)
|>912&nbsp;MHz
|>700<ref>{{Cite news|url=http://www.anandtech.com/show/11515/the-apple-wwdc-2017-live-blog#post0605144637|title=40% Graphics Performance A9X|last1=Humrick|first1=Ryan|last2=Smith|first2=Matt|access-date=2017-09-29|others=check references 44}}</ref>
|-
|[[Socionext]]
|2017
|SC1810
|[[#Series8XE (Rogue)|Series8XE]]
|
|
|-
|[[Synaptics]]
|2017
|Videosmart VS-550 (Berlin BG5CT)
|Series8XE [[#Series8XE (Rogue)|GE8310]]
|
|
|-
|rowspan="11" |Mediatek
|rowspan="2" |2017
|MT6739
|Series8XE [[#Series8XE (Rogue)|GE8100]]
|
|
|-
|MT8167
|rowspan="2" |Series8XE [[#Series8XE (Rogue)|GE8300]]
|
|
|-
|rowspan="4" |2018
|Helio A20 (MT6761D)
|
|
|-
|Helio P22 (MT6762)
|rowspan="3" |Series8XE [[#Series8XE Plus (Rogue)|GE8320]]
|
|
|-
|Helio A22 (MT6762M)
|
|
|-
|Helio P35 (MT6765)
|
|
|-
|2019
|MT6731
|Series8XE [[#Series8XE (Rogue)|GE8100]]
|
|
|-
|rowspan="3"|2020
|Helio A25
|rowspan="3"|Series8XE [[#Series8XE Plus (Rogue)|GE8320]]
|
|
|-
|Helio G25
|
|
|-
|Helio G35
|
|
|-
|rowspan="1"|2022
|Dimensity 930 (MT6855)
|IMG BXM-8-256
|950&nbsp;MHz
|259.2
|-
|
|2023
|Dimensity 7020
|IMG BXM-8-256
|
|
|-
| rowspan="2" |Texas Instruments
|2020
|TDA4VM
|Series8 [[#Series8XE (Rogue)|GE8430]]
|
|
|-
|2023
|AM69<ref>{{Cite web |date=February 2023 |title=AM69x Processors, Silicon Revision 1.0 datasheet (Rev. D) - am69 |url=https://www.ti.com/lit/ds/symlink/am69.pdf?ts=1705886185653 |url-status=live |archive-url=https://web.archive.org/web/20240926183349/https://www.ti.com/lit/ds/symlink/am69.pdf?ts=1705886185653 |archive-date=26 September 2024 |website=Texas Instruments}}</ref>
|IMG BXS-4-64
|800MHz
|50
|-
|Renesas
|2017
|R-Car D3 (R8A77995)
|Series8XE [[#Series8XE (Rogue)|GE8300]]
|
|
|-
|rowspan="4" |Unisoc (Spreadtrum)
|2018
|SC9863A
|Series8XE [[#Series8XE (Rogue)|GE8322]]
|
|
|-
|Q1 2019
|Tiger T310
|Series8XE [[#Series8XE (Rogue)|GE8300]]
|
|
|-
|Q3 2019
|Tiger T710
|rowspan="2"|Series9XM [[#Series9XM (Rogue)|GM9446]]
|
|
|-
|Q1 2020
|Tiger T7510
|
|
|-
|rowspan="2"|Mediatek
|2018
|Helio P90
|rowspan="2"|Series9XM [[#Series9XM (Rogue)|GM9446]]
|
|
|-
|Q1 2020
|Helio P95
|
|
|-
|Synaptics
|Q1 2020
|Videosmart VS680
|Series9XE GE9920
|
|
|-
|Semidrive
|Q2 2020
|X9, G9, V9
|[[#Series9XM (Rogue)|Series9XM]]
|
|
|}

==See also==
* [[List of PowerVR products|List of products featuring PowerVR accelerators]]
* [[Adreno]] – GPU developed by Qualcomm
* [[Mali (GPU)|Mali]] – available as SIP block to 3rd parties
* [[Vivante Corporation#Products|Vivante]] – available as SIP block to 3rd parties
* [[Tegra]] – family of SoCs for mobile computers, the graphics core could be available as SIP block to 3rd parties
* [[VideoCore]] – family of SOCs, by Broadcom, for mobile computers, the graphics core could be available as SIP block to 3rd parties
* [[Atom (system on chip)|Atom family of SoCs]] – with Intel graphics core, not licensed to 3rd parties
* [[List of AMD Accelerated Processing Unit microprocessors|AMD mobile APUs]] – with AMD graphics core, not licensed to 3rd parties

==References==
{{Reflist | refs =

<!-- Not in use
<ref name="7XE-Announce">{{cite web 
  |first=Alexandru
  |last=Voica
  |url=http://blog.imgtec.com/powervr/new-powervr-series7xe-family-targets-the-next-billion-mobile-and-embedded-gpus
  |title=New PowerVR Series7XE family targets the next billion mobile and embedded GPUs
  |publisher=Imagination Technologies
  |date=10 November 2014
  |access-date=10 November 2014}}</ref>
Not in use-->

<!-- Not in use
<ref name="7XT-Announce">{{cite web 
  |first=Alexandru
  |last=Voica
  |url=http://blog.imgtec.com/powervr/powervr-series7xt-gpus-push-graphics-and-compute-performance
  |title=PowerVR Series7XT GPUs push graphics and compute performance to the max
  |publisher=Imagination Technologies
  |date=10 November 2014
  |access-date=10 November 2014}}</ref>
Not in use-->

}}

==External links==
*{{Official website|http://www.imaginationtech.com/}}
*[https://web.archive.org/web/20190217152604/https://www.imgtec.com/?do-download=powervr-graphics-brochure PowerVR Technology Overview]

{{Graphics Processing Unit}}

[[Category:Graphics hardware companies]]
[[Category:Graphics processing units]]
[[Category:Graphics cards]]