Imagination IMG B-Series GPU family scales from IoT to the datacenter

Imagination IMG B-Series GPU: BXT MC4

Last year, Imagination Technologies unveiled IMG A-Series GPU family scaling from low-power IoT to mobile and high-performance server applications with up to 2.5 times the performance of the earlier PowerVR 9-series GPUs, as well as eight times faster AI processing and 60% less power under similar conditions. While I’m not aware of any SoCs announced with the new IMG A-Series GPU yet, the company has already announced the next-gen IMG B-Series GPU family with up to 4 times the multi-core performance thanks to decentralized multi-core technology, 30% lower power consumption, and 2.5 times the fill rate. The company offers four types of IM B-series GPU, each optimized for specific applications IMG BXE for high-resolution displays – From 1 up to 16 pixels per clock (PPC) BXE scales from 720p to 8K for UI rendering and entry-level gaming. IMG BXM designed for mid-range mobile gaming and complex UI solutions for DTV and other markets. IMG BXT four-core high-performance GPU generating 6.0 […]

Imagination Unveils IMG A-Series GPU Designed For Everything from IoT to Mobile and Server

IMG A-Series GPU

Imagination Technologies has just launched IMG A-Series GPU which they claim is “The GPU of Everything” and “The fastest GPU IP ever”. IMG A-Series can be customized and scaled for various applications & markets from automotive, AIoT, set-top box, mobile, and server.  Compared to the company’s earlier PowerVR 9Series GPU, IMG A-Series GPU delivers 2.5 times more performance, eight times faster AI processor, and 60% less power while running complex content with the same process node, area, and under similar conditions. IMG A-Series GPU supports the latest API standards including OpenGL ES, Vulkan, OpenCL, and Imagination provides a Unified AI API for use in combination with PowerVR neural network accelerators. It also offers 5x performance density compared to the best current shipping PowerVR devices and supports PVRIC4 lossless or virtually-lossless compression guaranteeing a 50% bandwidth and footprint reduction. The new GPU also leverages HyperLane Technology with up to eight individual, isolated hardware control lanes for multizone hardware virtualization leading to […]

Imagination iEW220 Low Power Dual-band 802.11n WiFi IP Solution Targets IoT, Automation and Wearables

Imagination-low-power dual-band WiFi IP solution

Imagination Technologies is better known for their PowetVR GPUs and as the previous owner of MIPS technology, but they do have other products as well and the company just announced their second generation 802.11n WiFi IP designed for low power applications Ensigma iEW220 is an IEEE 802.11a/b/g/n 1X1 SISO Wi-Fi IP solution comprised of RF, baseband and MAC, and supporting both 2.4GHz and 5GHz spectrum. iEW220 specifications: RF Features Dual-band 2.4 and 5 GHz RF including PA Power efficient AGC Optimized PLL power consumption Customized PA to minimize Tx power consumption Power optimized DAC/ADC design Integrated transmit/receive switch for RF Single pin RF antenna interface IEEE 802.11 PHY Support (2.4/5 GHz) IEEE 802.11b DSSS/CCK (up to 11 Mbps) IEEE 802.11a/g OFDM (up to 54 Mbps) IEEE 802.11n OFDM MCS 0-7, (20/40 MHz channels, up to 72.2/150 Mbps) Station, soft AP and Wi-Fi direct with up to four associated devices Tx power up to 19 dBm and Rx sensitivity up to […]

SiFive RISC-V Processors to Support Imagination PowerVR GPU and NNA IP Cores


RISC-V is getting more popular and mature as development progress, but if you want a system with RISC-V and a GPU so far you had to use a PCIe graphics card which is not cost-effective nor practical for most applications based on RISC-V processor. What is needed is some GPU IP that will glue with RISC-V core. Developing a new GPU is not an easy task (understatement of the month) so it would make sense to go with solutions available on the market. However, Arm Mali is tied to Arm Cortex cores and Arm is unlikely to want to help RISC-V take away their market share, and Adreno and VideoCore are owned by respectively Qualcomm and Broadcom which are unwilling to provide their GPU IP to third parties. This basically leaves us with Vivante and Imagination. Vivante may have made the most sense since open-source graphics drivers do exist (Etnaviv), but instead, I’ve just learned that SiFive added Imagination Technologies […]

Imagination Unveils PowerVR AX2185 and AX2145 Neural Network Accelerators (NNA)

PowerVR AX2185

Imagination Technologies introduced PowerVR Series2NX neural network accelerator (NNA) last year. At the time, the company claimed the NNA would deliver  twice the performance at half the bandwidth over existing competing solutions, and that it was the first dedicated hardware solution with flexible bit-depth support from 16-bit down to 4-bit. What they did not announce last September were any specific parts. But they’ve just done that today with PowerVR AX2185 and AX2145 NNAs “designed to enable high-performance computation of neural networks at very low power consumption in minimal silicon area”. PowerVR AX2185 PowerVR AX2185 is said to be the highest performing neural network accelerator per mm2 in the market with eight full-width compute engines delivering up to 4.1 TOPS (Tera Operations Per Second), or 2,048 MACs/clock. AX2185 NNA can deliver 3.5 times the performance compared to a desktop GPU. The NNA is best suited for high-end smartphones, smart surveillance and automotive markets. PowerVR AX2145 PowerVR AX2145 is a lower cost […]

Status of Embedded GPU Ecosystem – Linux/Mesa Upstream Support (ELC 2018 Video)

The Embedded Linux Confernce is on-going, and the Linux Foundation has been uploading videos about talks in a timely manner on YouTube. I checked out at RISC-V keynote yesterday, but today I’ve watched a talk by Robert Foss (his real name, not related to FOSS) from Collabora entitled “Progress in the Embedded GPU Ecosystem”, where he discusses open source software support in Linux/Mesa from companies and reverse-engineering support. The first part deals with the history of embedded GPU support, especially when it comes to company support. Intel was the first and offers very good support for their drivers, following by AMD who also is a good citizen. NVIDIA has the Nouveau driver but they did not really backed it up, and Tegra support is apparently sponsored by an aircraft supplier. Other companies have been slower to help, but Qualcomm has made progress since 2015 and now support all their hardware, Broadcom has a “one man team” handling VideoCore IV/V,  and […]

Vulkan 1.1 and SPIR-V 1.3 Specifications Released

The Khronos Group released Vulkan 1.0 specifications in 2015 as a successor of OpenGL ES, compatible with OpenGL ES 3.1 or greater capable GPU, and taking less CPU resources thank to – for instance – better use of multi-core processors with support for multiple command buffers that can be created in parallel. A year later, we saw Vulkan efficiency in a demo, since then most vendors have implemented a Vulkan driver for their compatible hardware across multiple operating systems, including Imagination Technologies which recently released Vulkan drivers for Linux. The Khronos Group has now released Vulkan 1.1 and the associated SPIR-V 1.3 language specifications. New functionalities in Vulkan 1.1: Protected Content – Restrict access or copying from resources used for rendering and display, secure playback and display of protected multimedia content Subgroup Operations – Efficient mechanisms that enable parallel shader invocations to communicate, wide variety of parallel computation models supported Some Vulkan 1.0 extensions are now part of Vulkan 1.1 […]

Imagination Releases PowerVR CLDNN Neural Network SDK and Image for Acer Chromebook R13

Last month, Imagination Technology released their PowerVR CLDNN SDK, an AI-oriented API that leverages OpenCL support in PowerVR GX6250 GPU in order to  create network layers for constructing and running a neural network on PowerVR hardware. Eventually the SDK will support PowerVR Series2NX Neural Network Accelerator, but while waiting for the hardware, the company has provided a firmware that runs only on Mediatek MT8173 based Acer Chromebook R13. The SDK includes a demo taking a live camera feed to identify the object(s) the camera is pointing at, using known network models such as AlexNet, GoogLeNet, VGG-16, or SqueezeNet. All models are Caffe models trained against the ImageNet dataset, a benchmark function is included within the demo. Beside simply playing with the demos, you’ll be able to study the source code to check out various helper functions such as file loading, dynamic library initialisation and OpenCL context management, and read documentation such as the PowerVR CLDNN reference manual explaining all CLDNN API’s functions. If you happen to […]