Versal AI Core Series

Breakthrough Acceleration for 5G, Networking and Cloud

Product Advantages

Versal AI Core series delivers breakthrough AI inference acceleration with AI engines that deliver over 100x greater compute performance than today’s server-class CPUs. The series is designed for a breadth of applications including cloud for dynamic workloads, and network for massive bandwidth, all while delivering advanced safety and security features. AI and data scientists, software, and hardware developers may all take advantage of the high compute density to accelerate the performance of any application.

Value	Deliverables
System Performance	AI Engines for over 100X greater compute vs. today’s server-class CPUs Programmable memory hierarchy enables customization per compute task to avoid memory performance bottlenecks, doubling efficiency over conventional cache-based GPUs Multicore Arm A72-based processing system with 2x single-threaded performance of previous generation A53’s 25Gb/s PCIe Gen4 with CCIX for cache-coherent connectivity to CPU host
BOM Cost Reduction	Single AI Core device replaces more than an entire server rack of high-end CPUs Dynamically swap between workloads to consolidate multiple accelerator devices into one Dynamically reconfigurable in milliseconds to perform anything from image recognition to network acceleration to video encoding to financial analysis to storage acceleration to scientific computing tasks.
Total Power Reduction	Up to 50% lower power vs. prior generation devices (when using new features)
Accelerated Design Productivity	SW tools optimized for deep learning and machine learning. Pre-set interfaces ("shell") with enhanced security, easily integrated into compute infrastructure Compile framework models in minutes