Editing PowerVR (section)

== 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;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Engines
! rowspan="2" |8-bit TOPS
! rowspan="2" |16-bit TOPS
! rowspan="2" |8-bit MACs
! rowspan="2" |16-bit MACs
! rowspan="2" |[[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;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Engines
! rowspan="2" |8-bit TOPS
! rowspan="2" |16-bit TOPS
! rowspan="2" |8-bit MACs
! rowspan="2" |16-bit MACs
! rowspan="2" |[[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;"
|-
! rowspan="2" |Model
! rowspan="2" |Date
! rowspan="2" |Cores
! rowspan="2" |8-bit TOPS
! rowspan="2" |16-bit TOPS
! rowspan="2" |8-bit MACs
! rowspan="2" |16-bit MACs
! rowspan="2" |[[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" />