Xilinx has announced that it has shipped Versal AI Core series and Versal Prime series devices to multiple tier one customers through the company’s early access program.
Versal is the industry’s first Adaptive Compute Acceleration Platform (ACAP), described as heterogeneous compute devices with capabilities that far exceed those of conventional CPUs, GPUs, and FPGAs.
The integrated, multi-core, heterogeneous compute platform can be changed at both the hardware and software levels to dynamically adapt to the needs of applications and workloads in data centre, automotive, 5G wireless, wired and defence markets.
The Versal ACAP is natively software programmable and its architecture features a flexible, multi-Tbit per second network on chip (NoC). By seamlessly integrating all engines and key interfaces, the NoC ensures that the ACAP is available at boot and easily programmed, whatever a developer’s specialism, be they software developers, data scientists or hardware developers, says Xilinx. Tools, software, libraries, IP, middleware, and frameworks are available to dynamically customize accelerated computing using industry-standard design flows.
The Versal ACAP portfolio is based on TSMC’s 7nm process technology ad is claimed to be the first platform to combine software programmability with dynamically configurable domain-specific hardware acceleration. It includes a mix of next-generation scalar engines for embedded compute, adaptable engines for FPGA silicon programmability, and intelligent engines for AI inference and advanced signal processing for advantageous raw performance and performance per Watt capabilities compared to CPU and GPU implementations, adds Xilinx.
The Versal AI Core series delivers the portfolio’s highest compute and lowest latency, for AI inference throughput and performance through the AI Engine. This core is optimised for cloud, networking and autonomous technology and is claimed to offer the highest range of AI and workload acceleration available in the industry.
The Versal Prime series is optimised for connectivity and in-line acceleration to operate across multiple markets and a diverse set of workloads, continues the company.
Both the Versal AI Core series and Versal Prime series include multiple devices, each with dual-core Arm Cortex-A72 application processors, dual-core Arm Cortex-R5F real-time processors, over 2m logic cells of adaptable hardware, and over 3,000 DSP engines optimised for high-precision floating point and low latency. The Versal AI Core devices offer up to 400 AI Engines optimised for AI inference and advanced signal processing workloads.
The Versal portfolio includes four additional series of devices, architected to deliver scalability and AI inference capabilities for diverse applications, from cloud and networking to wireless communications, edge computing, and endpoints.
The Versal AI Core and Versal Prime series will be generally available in the second half of 2019.
