NVIDIA Ampere, 5888 CUDA Cores, Video I/O: SDI, DP, Analog
- NVIDIA RTX™ GA104 GPU with 5888 CUDA Cores, 184 Tensor Cores, 46 RT Cores
- 16 GB GDDR6 256-bit memory with up to 512 GB/s
- WOLF FGX for video format conversion
- Support for 3G-SDI, DP/HDMI/DVI, RGB/CVBS formats
- Module power: TBD-190W, configurable
Overview
The VPX3U-A4500E-IO module includes an NVIDIA RTX™ A4500 embedded GPU and a WOLF FGX in a rugged 3U VPX module. The NVIDIA Ampere architecture includes CUDA cores for parallel processing, Tensor cores for dedicated AI accelerated compute, and Ray Tracing cores for superior rendering speeds. The WOLF FGX provides video conversion for formats that are not native to the GPU, such as SDI and analog formats.
The NVIDIA Ampere architecture has introduced many significant improvements to the performance and efficiency of the GPU, with more flexible CUDA FP32/INT core use, more efficient third generation Tensor cores, and second generation RT cores. The Ampere GPU fabrication uses an 8nm manufacturing processing providing significant power improvements which, along with other Ampere architecture improvements, can provide up to 154 GFLOPS/W, providing almost twice the performance per slot compared to the previous Turing generation’s 86 GFLOPS/W or the Pascal generation’s 62 GFLOPS/W.
Unlocking the best performance requires the best cooling capability. WOLF’s advanced cooling technology is designed to move heat using a low weight, high efficiency path from the GPU die to the wedgelocks.
WOLF-1440 3U VPX Board
WOLF modules come with factory configuration options to meet your requirements.
Contact WOLF to discuss MCOTS and custom design services if more extensive modifications are required.
GPU and FGX FEATURES
- Video inputs/outputs:
- 2x 3G-SDI inputs
- 2x DisplayPort 1.4 outputs
- DisplayPort outputs can be configured to output HDMI/DVI
- 2x converted video outputs: 3G-SDI, CVBS or RGB
- 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
- 184 Tensor Cores (3rd Gen), 68 (dense) / 136 (sparse) Tensor TFLOPS
- 46 Ray Tracing cores (2nd Gen)
- Error Correcting Code (ECC) memory
- NVENC (7th Gen) and NVDEC (5th Gen) with up to 8K video encoding and hardware decoding support
CONNECTIVITY / SYSTEM MANAGEMENT
- On-board IPMI controller for system management
- NVIDIA GPUDirect RDMA support
- Configurable PCIe Gen4 switch, includes Daisy Chain options
- Windows and Linux drivers
- NVIDIA Ampere driver support requires one of the following host CPUs: Intel E, S/H/H35 Class, AMD H/HS Class
MECHANICAL / OPEN SYSTEMS ARCHITECTURE
- High level of ruggedization:
- Rugged Conduction cooled
- Operating temperature: -40° to +85°C
- Vibration (sine wave): 10G peak, 5 - 2000Hz
- Shock: 40G peak
- Dimensions: 160mm x 100mm x 25.4mm
- Weight (approximately): TBD
- ANSI/VITA 48, 65 (VPX-REDI, OpenVPX)
- SOSA Aligned support for 14.2.3 legacy profile
Datasheet
Open the product datasheet - includes block diagram, order codes, etc.