Computational and Graphics Accelerators for Servers


Technical Specifications

Peak Double Precision Performance

  • 78.02 GFlops

Peak Single Precision Performance

  • 1.25 TFlops

Number of accelerators per card

  • 1


  • 512 STREAM processors (8 compute units)

Memory size per board

  • 2GB

Memory bandwidth for board

  • 48 GB/s

Accelerator applications

  • General-purpose digital rendering applications

Architecture features

  • Supports AMD STREAM technology, a set of GPU hardware and software features designed specifically to address high-performance workloads and workflows. The WX2100 Double Rate Geometry Boost allows the GPU to process geometry data at a rate of twice per clock cycle. The GPU on the WX2100 cards features four graphics engines that allow it to process four primitives per clock cycle. AMD Radeon Pro Series cards are optimized for OpenCL


  • Compatible with HPE ProLiant MicroServer Gen10

Minimum dimensions ( H x W x D)

  • 2.03 x 30.99 x 11.43 cm


  • 0.73 kg