Movidius Neural Compute Stick Shown to Boost Deep Learning Performance by about 3 Times on Raspberry Pi 3 Board

Intel recently launched Movidius Neural Compute Stick (MvNCS)for low power USB based deep learning applications such as object recognition, and after some initial confusions, we could confirm the Neural stick could also be used on ARM based platforms such as the Raspberry Pi 3. Kochi Nakamura, who wrote the code for GPU accelerated object recognition on the Raspberry Pi 3 board, got hold of one sample in order to compare the performance between GPU and MvNCS acceleration. His first attempt was quite confusing as with GoogLeNet, Raspberry Pi 3 + MvNCS achieved an average inference time of about 560ms, against 320 ms while using VideoCore IV GPU in RPi3 board. But then it was discovered that the “stream_infer.py” demo would only use one core out of the 12 VLIW 128-bit vector SHAVE processors in Intel’s Movidius Myriad 2 VPU, and after enabling all those 12 cores instead of just one, […]

Work on VideoCore V GPU Drivers Could Pave the Way for Raspberry Pi 4 Board

I’ve come across an article on Phoronix this morning, about VideoCore IV GPU used in Broadcom BCM283x “Raspberry Pi” processors, but part of the post also mentioned work related to VC5 drivers for the next generation VideoCore V GPU, written by Eric Anholt, working for Broadcom, and in charge of the open source code related to VideoCore IV GPU for Raspberry Pi. This led me Eric’s blog “This Week in VC4/VC5” and articles such as “2017-07-10: vc5, raspbian performance“, where he explains he committed Mesa drivers for VC5. I’ve just pushed a “vc5” branch to my Mesa tree (https://github.com/anholt/mesa/commits/vc5). This is the culmination of a couple of months of work on building a new driver for Broadcom’s V3D 3.3. V3D 3.3 is a GLES3.1 part, though I’m nowhere near conformance yet. This driver is for BCM7268, a set-top-box SOC that boots an upstream Linux kernel. I’m really excited to be […]

Android Can Now Boot with a Full Open Source Graphics Stack on NXP i.MX6 Boards

While the Android operating systems is itself open source, it still relies on proprietary binary files to leverage GPU acceleration, VPU hardware decoding, wireless connectivity, and so on. It’s been possible to run Android with an open source software graphics stack, but it’s normally terribly slow and barely usable. But Collabora has announced it could now boot Android with a full-graphics stack on iMX6 platforms using no proprietary blobs at all. To do so, they leveraged the work done on Etnaviv open source drivers for Vivante GPUs, and adding the different formats used for  graphical buffers in Android and Mesa library using modifiers representing different properties of buffers. They further explain: Support was added in two places; Mesa and gbm_gralloc. Mesa has had support added to many of the buffer allocation functions and to GBM (which is the API provided by Mesa, that gbm_gralloc uses). gbm_gralloc in turn had support […]

ARM Cortex-A75 & Cortex-A55 Cores, and Mali-G72 GPU Details Revealed

We’ve already seen ARM Cortex A75 cores were coming thanks to leak showing Snapdragon 845 SoC will feature custom Cortex A75 cores, but we did not have many details. But since we live in a world where “to leak is glorious”, we already have some slides originally leaked through VideoCardz with the post now deleted, but Liliputing & TheAndroidSoul got some of the slides before deletion, so let’s see what we’ve got here. ARM Cortex A75 So ARM Cortex-A75 will be  about 20% faster than Cortex A73 for single thread operation, itself already 30% faster than Cortex A72. It will also be the first DynamIQ capable processor together with Cortex A55 with both cores potentially used in big.LITTLE configuration. Cortex A75 performance is only better for peak performance, and remain the same as Cortex-A73 for sustained performance. The chart above does not start at zero, so it appear as though […]

Getting Started with OpenCV for Tegra on NVIDIA Tegra K1, CPU vs GPU Computer Vision Comparison

This is a guest post by Leonardo Graboski Veiga, Field Application Engineer, Toradex Brasil Introduction Computer vision (CV) is everywhere – from cars to surveillance and production lines, the need for efficient, low power consumption yet powerful embedded systems is nowadays one of the bleeding edge scenarios of technology development. Since this is a very computationally intensive task, running computer vision algorithms in an embedded system CPU might not be enough for some applications. Developers and scientists have noticed that the use of dedicated hardware, such as co-processors and GPUs – the latter traditionally employed for graphics rendering – can greatly improve CV algorithms performance. In the embedded scenario, things usually are not as simple as they look. Embedded GPUs tend to be different from desktop GPUs, thus requiring many workarounds to get extra performance from them. A good example of a drawback from embedded GPUs is that they are […]

Android Play Store Tidbits – Blocking Unlocked/Uncertified/Rooted Devices, Graphics Drivers as an App

There’s been at least two or three notable stories about the Play Store this week. It started with Netflix not installing from the Google Play Store anymore on rooted device, with unclocked bootloader, or uncertified devices, and showing as “incompatible”. AndroidPolice contacted Netflix which answered: With our latest 5.0 release, we now fully rely on the Widevine DRM provided by Google; therefore, many devices that are not Google-certified or have been altered will no longer work with our latest app and those users will no longer see the Netflix app in the Play Store. So that means you need to  Google Widevine DRM in your device, which mean many Android TV boxes may stop to work with Netflix. You can check whether you device is certified by opening Google Play and click on settings, Scroll to the bottom and check Device Certification to see if it is Certified or Uncertified […]

Imagination PowerVR “Furian” Series8XT GT8525 GPU Targets High-end Smartphones, Virtual Reality and Automotive Products

Imagination Technologies has unveiled their first GPU based on PowerVR Furian architecture with Series8XT GT8525 GPU equipped with two clusters and designed for SoCs going to into products such as high-end smartphones and tablets, mid-range dedicated VR and AR devices, and mid- to high-end automotive infotainment and ADAS systems. The Furian architecture is said to allow for improvements in performance density, GPU efficiency, and system efficiency, features a new 32-wide ALU cluster design, and can be manufactured using sub-14nm (e.g. 7nm process once available). PowerVR GT8525 GPU supports compute APIs such as OpenCL 2.0, Vulkan 1.0 and OpenVX 1.1. Compared to the previous Series7XT GPU family, Series8XT GT8525 GPU delivers 80% higher fps in Trex benchmark, an extra 50% fps in GFXbench Manhattan benchmark, 50% higher fps in Antutu, doubles the fillrate throughput for GUI, and increases GFLOPs for compute applications by over 50%. GT8525 GPU is available for licensing […]

Think Silicon Ultra Low Power NEMA GPUs are Designed for Wearables and IoT Applications

When you have to purchase a wearable device, let’s say a smartwatch or fitness tracker, you have to make trade offs between user interface and battery life. For example, a fitness tracker such as Xiaomi Mi Band 2 will last about 2 weeks per charge with a limited display, while Android smartwatches with a much better interface need to be recharged every 1 or 2 days. Think Silicon aims to improve battery life of the devices with nicer user interfaces thanks to their ultra-low power NEMA 2D, 3D, and GP GPU that can be integrated into SoCs with ARM Cortex-M and Cortex-A cores. The company has three family of GPUs: NEMA|p pico 2D GPU with one core 4bpp framebuffer, 6bpp texture with/out alpha Fill Rate – 1pixel/cycle Silicon Area – 0.07 mm2 with 28nm process Power Consumption – leakage power GPU consumption of 0.06mW; with compression (TSFSc): 0.03 mW NEMA|t […]

Memfault IoT and embedded debugging platform