Last week, Mythic announced a breakthrough with compute-in-memory technology based on a 40 nm process with what the company claims to be the industry’s first Analog Matrix Processor. The M1108 AMP AI accelerator chip targets high-end edge AI applications including smart home, AR/VR, drones, and is said to set a benchmark in the industry for high performance and low power in a single cost-effective device, also available in M.2 and PCIe form factors. The M1108 comes with an array of flash cells, ADCs, a 32-bit RISC-V nano-processor, a SIMD vector engine, SRAM, and a high-throughput Network-on-Chip (NOC) router. With 108 AMP tiles, the M1108 provides up to 35 Trillion-Operations-per-Second (TOPS) enabling ResNet-50 at up to 870 fps. This enables a power-efficient execution of complex AI models such as ResNet-50, YOLOv3, and OpenPose Body25. The industry leader NVIDIA also has a similar AI accelerator chip NVIDIA Xavier AGX which delivers up to 32 TOPS. A wide variety of host processors are […]
With the increasing need for video encoding, there are some breakthrough developments in hardware-accelerated video encoding for Linux. Bootlin has been working on the implementation of Hantro H1 hardware accelerated video encoding to support H.264 encoding on Linux which follows the company’s work on the previously-released open-source VPU driver for Allwinner processors. Hantro H1 Hardware Hantro H1 is a common hardware H.264 encoder, it can also do VP8 and JPEG. It is found in a few ARM SoCs including a lot of Rockchip (RK3288, RK3328, RK3399, PX30, RK1808) and NXP (i.MX 8M Mini). Depending on the version, it can support up to 1080p at 30 or 60 fps. Here we can see different blocks used for encoding. Hantro H1 is a stateless hardware implementation which means it has no microcontroller or firmware running. As can be seen in the diagram, it has a pre-processor that can do things like cropping, rotation, scaling, stabilization, and CSC. It will produce slice NALUs […]
To support embedded design, there are several options when it comes to choosing an operating system (OS). Some of the traditional approaches to building custom Linux systems is to use built systems such as Yocto/OpenEmbedded or Buildroot. The options available for system integration include building everything manually, binary distributions (Debian, Ubuntu, Fedora, etc.), and build systems (Buildroot, Yocto, PTXdist, etc.). The major drawback of build systems is that they are not as easy as a binary distribution and also the build time is more. Why was ELBE born? In the early days, the embedded devices had 4MiB flash and 16MiB of RAM. With these specifications, people started to hack a root file system for their devices. But in some cases, they had to start with building a cross-toolchain first. For this, tools like OpenEmbedded, Buildroot are good as long as they are well maintained. For this, a lot of libraries are used for the development of a system. Now, depending […]
For more than 6 years, the LibIIO library has existed to ease the development of software interfacing Linux Industrial I/O (IIO) devices. It is part of the Linux Kernel and a subsystem that provides support for devices like analog to digital or digital to analog converters (ADCs, DACs). This subsystem includes ADCs, accelerometers, pressure sensors, color, light and proximity sensors, temperature sensors, RF transceivers, and many more. You can use LibIIO natively on an embedded Linux target. It is cross-platform, supporting Linux, Windows, and Mac OS. Analog Devices Inc. was the main company behind LibIIO development, which is currently an active open-source library, which many people have contributed to. What does LibIIO do? LibIIO will identify the channels that belong to each device. It will assign specific attributes, one for the channels and one for the devices. Then, it will also create a context that is a place where all the devices exist, and you can browse through the channels, […]
The story of LoRa began in 2009 when Cycleo, a French company, invented LoRa. LoRa (Long Range) support for Zephyr OS goes back to December 2019. Since then, there has been a huge interest among the community to extend their support for it. More recently, LoRaWAN (low-power wide-area network) support was added to Zephyr OS. This will provide true networking support to Zephyr OS over LoRa. It operates in licensed free Sub Gigahertz frequencies (865 MHz-India, 868 MHz-Europe, etc…). It makes the perfect choice for low data rates and long-range applications. What is LoRaWAN? LoRaWAN is a MAC layer that sits on top of the LoRa. According to the OSI model, LoRaWAN is the MAC (media access control) layer while LoRa is the PHY (physical) layer. LoRa protocol is closed, meaning it is proprietary to Semtech, while LoRaWAN specifications are open to the public. The reference implementation is available at the GitHub repository. The end nodes transmit data to the […]
Igalia has been developing a new open-source Mesa driver for the Raspberry Pi 4 since December 2019 and announced the implementation of the classical triangle Vulkan demo last February. Four months after the announcement of the Vulkan effort for Raspberry Pi 4 (v3dv), they merged with Mesa upstream. This means Raspberry Pi 4’s v3dv Vulkan driver has become part of the official Mesa drivers. Thus, bringing several advantages, like easy to find as it is now available on the official Mesa repository. Bugs can now be filed on the official Mesa repository bug tracker. In June, they passed over 70,000 tests from the Khronos Conformance Test Suite for Vulkan 1.0 and had an implementation of a significant subset of the Vulkan 1.0 API. This does not mean that the driver is ready for production use as they have implemented the full Vulkan 1.0 API. They are now passing over 100,000 tests in the Kronos Conformance Test Suite (CTS) and have implemented the full Vulkan […]
When it comes to RISC-V based SoC, SiFive has always set a benchmark in the RISC-V ecosystem. On 29th October 2020, SiFive confirmed the first-ever RISC-V PC. After an increased demand for AI-focused RISC-V microarchitecture, targeting all applications from artificial intelligence, the internet of things, high-performance computing, and now even desktop PCs. SiFive Freedom U740 powered HiFive Unmatched mini-ITX motherboard comes with a complete development environment which allows developers to create RISC-V based applications from bare-metal to Linux-based systems. “HiFive Unmatched ushers in a new era of RISC-V Linux development with a platform in a PC form factor. Powered by the SiFive Freedom U740, a high-performance multi-core, 64-bit dual-issue, superscalar RISC-V processor.”, SiFive says. It is the world’s fastest native RISC-V development platform. SiFive HiFive Unmatched Board At the heart of the SiFive board is a SiFive FU740 processor coupled with 8 GB DDR4 memory and 32 MB SPI Flash. It comes with a 4x USB 3.2 ports and a 16x […]
Last week, Flex Logix announced the InferX X1 AI Inference Accelerator at Linley Fall Conference 2020. Today, they announced the InferX X1 SDK, PCIe board, and M.2 board. InferX X1 Edge Inference SDK The InferX Edge Inference SDK is simple and easy. The input to the compiler can be an open-source high-level, hardware-agnostic implementation of the neural network model that can be TensorFlow Lite or ONNX model. The compiler takes this model and looks for the available X1 resources and generates a binary executable file. This goes to the runtime which then takes the input stream, for example, a live feed from a camera. The user has to specify which compiler model, then the InferX X1 driver takes it and sends it to hardware. The binary file generated is fed to InferX X1 through the runtime. Then it takes the input data stream with a user-specified model and gives the output back to the host. This gives an advantage to […]
Privacy & Cookies Policy
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.