Pocket AI – A portable NVIDIA RTX A500 eGPU for AI developers, embedded & industrial applications

Pocket AI NVIDIA RTX 500 eGPU

We’ve previously seen it’s possible to connect an eGPU to a mini PC through a PCIe x16 to M.2 NVMe adapter or a Thunderbolt 3 port, but while it’s fine to install on your desk for gaming or develop AI applications, the eGPU being larger than most mini PCs, it’s a little too big to integrate into products, and potentially inconvenient to carry around. ADLINK Pocket AI portable eGPU changes that with an NVIDIA RTX A500 GPU housed in a 106 x 72 x 25mm box that’s about the size of a typical power bank and connects to a host through a Thunderbolt 3 connector. The company says the upcoming eGPU is mostly designed for AI developers, professional graphics users, and embedded industrial applications, but can also be for gaming. Pocket AI specifications: GPU – NVIDIA RTX 500 Architecture – NVIDIA Ampere GA107 Base clock: 435 MHz Boost clock: 1335 […]

NanoPi R6S RK3588S mini PC & router review – Part 2: Ubuntu 22.04

NanoPi R6S Ubuntu 22.04

NanoPi R6S is both a mini PC and a router based on Rockchip RK3588S processor. I received some samples in November and started the NanoPi R6S review with OpenWrt/FriendlyWrt quickly testing the 2.5GbE interfaces and routing with iperf3, and it worked pretty well. But using a system with an octa-core Cortex-A76/A55 processor and 8GB RAM as an OpenWrt router only feels like a waste of resources, so I wanted to install a more versatile operating system – Ubuntu 22.04 – for further testing. My struggles installing Ubuntu 22.04 on NanoPi R6S FriendlyELEC provides various images on the Wiki either booting from an SD card, installing from a MicroSD to the eMMC flash (aka eFlasher imagers), or flashing through USB with Windows tools. I like the eFlasher images since the OS runs from the internal eMMC flash and no special tools are needed. I had just used the FriendlyWrt eFlasher image, […]

STMicro adds more STM32U5 Cortex-M33 MCUs with NeoChrom 2.5D GPU, 128KB to 4MB flash, NIST certification, etc…

STM32U5 MCU family 2023

STMicro has added three new families to its STM32U5 Cortex-M33 microcontroller series with the entry-level STM32U535/545 with as little as 128KB flash, the STM32U595/5A5 with up to 4MB of Flash and 2,514KB of RAM, and the STM32U599/5A9 with similar features as the STM32U595/5A5 by adding the new NeoChrom 2.5D GPU plus TFT-LCD and DSI display interfaces. STMicro also announced that STM32U5 series MCUs were the first general-purpose MCUs to receive NIST-embedded random-number entropy source certification. As of now, along with the STM32U575/585 introduced in February 2021 with an optional Chrom-ART 2D GPU, STMicro offers eight product lines as part of the STM32U5 MCU series as shown in the table below. That means the STM32U5 microcontroller can address a wider range of applications with the STM32U535/545 targetting lower-cost applications with less flash and RAM, the “legacy” STM32U575/U585 for mid-range applications, and the STM32U59X/5AX for applications where more storage and memory is […]

StarFive JH7110 RISC-V processor specifications released

StarFive JH7110 block diagram

With the Star64 and VisionFive 2 single board computers, we’ve already got two hardware platforms based on the StarFive JH7110 quad-core RISC-V processor, but somehow we did not get the detailed specifications of the new processor, and some details such as the presence of AI accelerators or exact PCIe specifications were lacking. Some documentation for the StarFive JH7110 processor has been released and answers some of those questions. It’s actually an SoC with six RISC-V cores, of which four 64-bit RISC-V cores run the main OS, plus a 64-bit RISC-V monitoring core, and a 32-bit RISC-V real-time core. The AI accelerators found in the JH7100 (Neural Network Engine and NVDLA) appear to be gone for good, and there are two 1-lane PCIe 2.0 interfaces up to 5 Gbps each. StarFive JH7110 specifications: CPU sub-system Quad-core 64-bit RISC-V SiFive U74 (RV64GC) processor @ up to 1.5 GHz with 32KB D-Cache, 32KB […]

Raspberry Pi 4 gets Vulkan 1.2 conformance, Android Vulkan support

Vulkan deferred shading with shadows

Iglia has done more work on the graphics driver for the VideoCore VI GPU found in Raspberry Pi 4 and other Broadcom BCM2711-based hardware with Vulkan 1.2 conformance, and Roman Stratiienko added Vulkan 3D graphics acceleration to Android, or more exactly LineageOS. Raspberry Pi and Iglia have been collaborating together since the launch of the Raspberry Pi 4 SBC to develop a Khronos conformant Mesa 3D graphics driver for the board, and that’s a long-term project that’s been going on for over two years, and not quite finished yet. Here’s a non-exhaustive timeline of the project so far: February 2020 – Raspberry Pi 4 V3DV driver gets OpenGL ES 3.1 conformance, work on Vulkan starts with the basic triangle demo showcased June 2020 – Vulkan driver source code released with many demos working on Raspberry Pi 4 October 2020 – Iglia gives a project update status presentation for Raspberry Pi […]

Arm unveils Cortex-X3 and Cortex-A715 Armv9 cores, improves Cortex-A510 efficiency

Arm Cortex-A510 Cortex-A715 Cortex-X3

Besides announcing the new Immortalis-G715, Mali-G715, and Mali-G615 GPUs, Arm has also introduced the second-generation of Armv9 cores with the Cortex-A715 and Cortex-X3 cores with respectively a 20% energy-efficiency improvement (and smaller 5% performance uplift) over the Cortex-A710 core and a 25% peak performance boost against the Cortex-X2 flagship core. The announcement also includes a “refresh” of the Cortex-A510 core announced last year with a 5% improvement in efficiency and the same level of performance. The Cortex-X3 will also be used in (Windows 11) laptop processors with the single-thread performance improved by up to 34% in that case. Armv9 mobile SoC (Total Compute) in 2023 Future Armv9 flagship mobile SoC worked on this year, and released in 2023 should have a combination of Cortex-X3, Cortex-A715, and Cortex-A510 cores, an Immortalis-G715 GPU, a new DSU-110 “DynamIQ Shared Unit” that supports 50% more cores in CPU clusters (or up to 12 cores […]

Arm Immortalis-G715 GPU supports hardware-based ray tracing

Arm Mali-G615, Mali-G715, Immortalis G715

Arm has unveiled the new Immortalis family of flagship GPUs with support for hardware-based ray tracing starting with the Immortalis-G715 GPU, as well as two new premium Mali GPUs namely Arm Mali-G615 and Mali-G715. Software-based ray tracing was already implemented on Arm Mali G710 on SoC’s such as the MediaTek Dimensity 9000, but the hardware-based ray tracing in the Immortalis-G715 delivers over 300 percent performance improvements, and only uses 4% of the shader core area. This will be mostly used in games to generate realistic lighting and shadows as can be seen in the “before vs after” video demo below.   While hardware-based ray tracing is only available on the Immortalis-G715 GPU, all three new GPUs feature a new execution engine and support variable rate shading. The Immortalis-G715 flagship GPU will come with 10 or more cores, the Mali-G715 with 6 to 9 cores, and the Mali-G615 with 6 cores […]

Renesas SmartBond DA1470x Bluetooth 5.2 LE SoC features a 2D GPU

SmartBond DA1470x block diagram

I would not usually ask myself “is there a GPU?” in a Bluetooth LE SoC, but that’s exactly what Renesas SmartBond DA1470x Bluetooth 5.2 LE microcontroller offers with a 2D GPU used to accelerate the rendering of user interfaces in wearables, healthcare devices, home appliances with displays, industrial automation and security systems, as well as consoles in e-bikes and gaming equipment. The DA1470x also comes with a Cortex-M33 application core, a Cortex-M0+ sensor node controller, 1.5MB SRAM, and an ultra-low-power hardware VAD for always-on audio processing, as well as various peripherals. Specifications: CPU cores Arm Cortex-M33 core @ up to 160 MHz for the application Arm Cortex-M0+ core acting as a sensor node controller (SNC) GPU – 2D GPU for advanced graphics processing Memory – 1.5MB SRAM Storage 4 kB OTG, 32KB ROM on-chip Decrypt-on-the-fly Octa/Quad SPI flash interface Dedicated QSPI PSRAM and QSPI flash interfaces eMMC interface supporting up […]

EmbeddedTS embedded systems design