Imagination Launches IMG A-Series ”GPU of Everything"

December 13, 2019 Alix Paultre

Imagination Technologies announced the tenth generation of their PowerVR graphics architecture, the IMG A-Series. Claiming to be the fastest GPU IP ever released, the devices leverage the PowerVR GPU architecture. Presented as "The GPU of Everything" the IMG A-Series offers performance scalability that ranges from 1 pixel per clock (PPC) parts for the entry-level, up to 2 TFLOP cores for performance devices, and beyond that to multi-core solutions for cloud applications.

The IMG A-Series offers up to 2.5x the performance, 8x faster machine learning processing, and 60% lower power than current PowerVR devices. The family is available in four configurations:

-IMG AXT-64-2048 with 2.0 TFLOPS, 64 Gpixels, and 8 TOPS of AI performance

-IMG AXT-48-1536 with 1.5 TFLOPS, 48 Gpixels, and 6 TOPS of AI performance

-IMG AXT-32-1024 with 1 TFLOPs, 32 Gpixels, and 4 TOPS of AI performance

-IMG AXT-16-512 with 0.5 TFLOPS, 16 Gpixels, and 2 TOPS of AI performance

-IMG AXM-8-256 with 0.25 TFLOPS, 8 Gpixels, and 1 TOPS of AI performance

-IMG AXE-2-16 with 2 PPC, 16 GFLOPS and 2 Gpixels

-IMG AXE-1-16 with 1 PPC, 16 GFLOPS, and 1 Gpixels

The IMG A-Series' PowerVR architecture uses a tile-based deferred rendering technology, meaning it only draws what is visible on screen. This combined with advanced power signalling using Pro-Active DVFS (Dynamic Voltage and Frequency Scaling) and Deadline Scheduling algorithms, which deliver fast power control with ultra-low latency response. If parts of the GPU aren't fully utilised or needed for processing, they are immediately slowed down or even put to sleep to ensure optimal power efficiency.

Imagination's HyperLane technology, with individual hardware control lanes isolated in memory, enable different tasks to be submitted to the GPU simultaneously for secure GPU multitasking. With Dynamic Performance Control, the GPU can spread its performance across these multiple tasks, executing them all, while maximising GPU use. HyperLane technology can also isolate protected content for rights management. All IMG A-Series GPUs support up to eight HyperLanes.

HyperLane also enables a feature called AI Synergy, which enables SoC designers to improve graphics performance, using its spare resource to enable programmable AI alongside a fixed-function, highly-optimised Imagination neural network accelerator.  A unified AI API in combination with PowerVR neural network accelerators supports the latest API standards including OpenGL ES, Vulkan, and OpenCL.

 

For more information, visit www.imgtec.com

Previous Article
Congatec and Hacarus Present AI Kit Based on Sparse Modeling
Congatec and Hacarus Present AI Kit Based on Sparse Modeling

Congatec and Hacarus unveiled what was presented as the first embedded computing kit for Artificial Intelli...

Next Article
Infineon’s 50- and 60-mm Prime Block Modules Target Drives and UPS Applications
Infineon’s 50- and 60-mm Prime Block Modules Target Drives and UPS Applications

Infineon Technologies Bipolar expanded its portfolio of thyristor/diode modules. The Prime Block 50mm modul...