HPEC with NVIDIA Ampere and ConnectX-6
- NVIDIA RTX GA104 GPU with 17.66 TFLOPS FP32 peak performance*, 5888 CUDA Cores, 184 Tensor Cores, 46 RT Cores
- NVIDIA ConnectX-6 providing the module with up to 100GbE Ethernet and PCIe Gen4
- 16 GB GDDR6 256-bit memory with up to 512 GB/s
- Module power: 80-170W, configurable
*Peak performance requires the highest power configuration mode.
The VPX3U-A4500E-CX6 HPEC module includes an NVIDIA RTX™ A4500 embedded GPU and a ConnectX SmartNIC. The A4500 embedded GPU provides advanced processing capabilities for high performance embedded computing (HPEC) and artificial intelligence (AI) processing. The ConnectX-6 provides the Ethernet and PCIe connectivity needed to move large datasets efficiently.
The NVIDIA Ampere architecture includes CUDA cores for HPEC, 3rd generation Tensor cores for AI and data science computations, and 2nd generation Ampere Ray Tracing (RT) cores for visually accurate rendering. The Ampere architecture introduces support for PCIe Gen 4, improvements to memory handling, and improvements to warp scheduling, which all optimize the way data is distributed for processing. The Ampere GPU is very power efficient, delivering significantly more performance per watt compared to previous generations.
The NVIDIA ConnectX-6 SmartNIC provides PCIe and Ethernet connectivity. ConnectX-6 is ideal for the high-speed, secure, data transfer capabilities required for data-heavy tasks such as sensor data processing and other C4ISR tasks. The ConnectX-6 also provides support for RDMA over Converged Ethernet (RoCE), enabling the fastest method for transferring data across the network to the GPU.
WOLF-144L 3U VPX Board
This product is currently undergoing development and/or verification by the WOLF engineering team.
Contact WOLF for the latest update on product availabilty or to discuss your MCOTS requirements.
- Ampere GPGPU parallel processing:
- CUDA Toolkit 11, CUDA Compute capability 8.6
- OpenCL™ 3.0, DirectX® 12 Ultimate, OpenGL 4.6, OpenGL ES 3.2, Vulkan™ 1.2
- CUDA-X AI and CUDA-X HPEC libraires
- 184 Tensor Cores (3rd Gen)
- 46 Ray Tracing cores (2nd Gen)
- NVENC (7th Gen) and NVDEC (5th Gen) with up to 8K video encoding and hardware decoding support
CONNECTIVITY / SYSTEM MANAGEMENT
- PCIe Gen4 configurable interface, up to x16
- Daisy Chain and NTB options
- Backplane Ethernet with 10/40/100 GBASE-KR4 and 10GBASE-KR data planes; supports GPUDirect RDMA
- Block-level hardware encryption and the use of dedicated encryption keys per user
- Switching is offloaded from the CPU and run on the ConnectX hardware with NVIDIA ASAP2 technology
- On-board IPMI controller for system management
- Windows and Linux drivers
MECHANICAL / OPEN SYSTEMS ARCHITECTURE
- High level of ruggedization:
- Rugged Conduction cooled
- Operating temperature: Conduction cooled -40° to +85°C
- Vibration (sine wave): 10G peak, 5 - 2000Hz
- Shock: 40G peak for conduction cooled
- Dimensions: 160mm x 100mm x 25.4mm
- Weight (approximately): TBD
- SOSA Aligned options with depopulated P2 or P2B
- Support for the SOSA legacy payload slot profile
Open the product datasheet - includes block diagram, order codes, etc.