Specifications
MPN
XCVE2202-2MSESFVA784
Device
XCVE2202
Family
AMD Versal AI Edge Series (VE)
Package
SFVA784 (784-ball FCBGA)
Speed Grade
-2
Temperature Grade
E — extended (0 °C to +100 °C Tj)
INT8 (max sparsity)
32 TOPS
INT8 (dense)
16 TOPS
BFLOAT16
8 TFLOPS
AI Engine-ML
24
DSP Engines
324
Application Processing Unit
Dual-core Arm Cortex-A72, 48 KB / 32 KB L1 with parity & ECC, 1 MB L2 with ECC
Real-time Processing Unit
Dual-core Arm Cortex-R5F, 32 KB / 32 KB L1, 256 KB TCM with ECC
On-chip Memory
256 KB with ECC
Connectivity
Ethernet (x2), UART (x2), CAN-FD (x2), USB 2.0 (x1), SPI (x2), I2C (x2)
Toolchain
AMD Vivado / Vitis unified software platform
Vendor
AMD
Overview
The XCVE2202 (ordering code XCVE2202-2MSESFVA784) is an AMD Versal AI Edge series device in a 784-ball FCBGA package. It is built for edge inference where a discrete accelerator card is not viable: 24 AI Engine-ML tiles deliver 32 TOPS TOPS of INT8 throughput with maximum sparsity (16 TOPS TOPS dense) and 8 TFLOPS BFLOAT16, alongside 324 DSP engines and a hardened embedded processing system.
The processing system pairs a dual-core Arm Cortex-A72 application unit with a dual-core Arm Cortex-R5F real-time unit and 256 KB of ECC-protected on-chip memory, giving functional-safety-capable deterministic control next to the inference fabric. Connectivity covers Ethernet, USB 2.0, UART, SPI, I2C and CAN-FD — the fieldbus mix used by automotive, robotics and industrial machine-vision designs. Development uses AMD Vivado and Vitis. We supply the XCVE2202-2MSESFVA784 factory-sealed through distribution channels; contact us for availability and pricing.
Key Benefits
High INT8 throughput per watt for edge inference; hardened AI Engine-ML array leaves the programmable logic free for sensor interfaces; ECC-protected memory and a real-time core for functional-safety designs; single-chip replacement for an accelerator card plus host.
Applications
Automotive ADAS and in-cabin sensing, industrial machine vision and robotics, medical imaging, smart-city video analytics, aerospace and defence edge processing.
Request a Quote — XCVE2202-2MSESFVA784
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →