Silicon Valley adaptive computing bellwether Xilinx announced its entrance into the growing system-on-module (SOM) market today, with a portfolio of palm-sized compute modules for embedded applications that accelerate AI, machine learning and vision at the edge. Xilinx Kria will eventually expand into a family of single board computers based on reconfigurable FPGA (Field Programmable Gate Array) technology, coupled to Arm core CPU engines and a full software stack with an app store, the first of which is specifically is targeted at AI machine vision and inference applications. The Xilinx Kria K26 SOM employs the company's UltraScale multi-processor system on a chip (MPSoC) architecture, which sports a quad-core Arm Cortex A53 CPU, along with over 250 thousand logic cells and an H.264/265 video compression / decompression engine (CODEC). This may sound like alphabet soup as I spit out acronyms, however, the underlying solution is a compelling offering for developers and engineers looking to give new intelligent systems, in industries like security, smart cities, retail analytics, autonomous machines and robotics, the ability to see, infer information and adapt to their deployments in the field. Also on board the Xilinx Kria K26 SOM is 4GB of DDR4 memory and 245 general purpose IO, along with the ability to support 15 cameras, up to 40 Gbps of combined Ethernet throughput, and four USB 2/3 compatible ports.
Vision is one of the hottest area for AI development because vision can or will be used for a wide variety of consumer and industrial applications. Everything from security systems and retail monitoring solutions to manufacturing quality control and autonomous machines require vision AI. Xilinx has been developing chips and tools for AI development leveraging the company's expertise in programmable/adaptable platforms. Now, Xilinx is introducing the Kria System-on-Module (SoM) just for vision AI applications. A module or SoM is a predesigned system or sub-system with the required chips already mounted on a printed circuit board (PCB) that connect to other system components or interfaces.
The $199 Kria KV260 starter kit comes ready with inputs for mutiple cameras, image processing logic, and softare for object detection, to get people started easily on a computer vision product. In the hotly contested field of "edge" artificial intelligence, systems and chip makers are coming up with ways to get their customers to market without months or years of machine learning development. The latest contestant in that regard is programmable-logic maker Xilinx, which on Tuesday unveiled what is planned to be a family of "system-on-module," or SOM, products, chips on boards with pre-built applications for tasks such as object detection. The announcement features a starter kit for $199 containing all the guts of a camera system to detect people and things in the world. Kria, as the product line is called, are Xilinx's first production-ready SOMs.
The configurable, adaptable nature of FPGAs (Field Programmable Gate Arrays), as well system-on-chip architectures that employ them, make the technology invaluable in a multitude of applications, from AI-powered data centers to smart edge devices and the IoT. As part of its evolving strategy, Silicon Valley bellwether Xilinx has been integrating this adaptive technology into platform accelerator solutions for machine learning, as well as domain specific architecture solutions that incorporate various compute resources like ARM cores, high speed IO and even RF functions. One of the issues with modern heterogeneous compute architectures like this today, however, is that they're difficult for the average software developer to work with. You have to have a lot of hardware expertise to understand how to best utilize the various compute resources in modern systems, from CPUs to GPUs and FPGAs. However, today, Xilinx has announced a new, free, open source tool it calls Vitis.
At this week's Xilinx Developers Forum (XDF) in San Jose, California, Xilinx announced "Vitis" – a new framework for developing applications that use Xilinx programmable logic devices such as FPGAs, ACAPs, MPSoCs, RFSoCs, and all the other acronyms they can come up with that refer to what we'd call "FPGAs." With an abundance of grandiosity, the company proclaimed that Vitis was "five years and a total of 1,000 man-years in the making." We don't frequently encounter person-millenia metrics for new product announcements. Even if we assume that they started this in 1984 when Xilinx was founded, they would still have needed at least 28 engineers writing code full time for Vitis over that 35 year span. Spoiler: engineering-eons aside, Vitis IS impressive, and more importantly, is likely to mature to be even more impressive over the coming years.