Condor GR4-A4500

Condor GR4 3U VPX Card

Video Capture & GPGPU Card

NVIDIA Ampere™ Architecture:
RTX A4500 GPU

High-Performance Embedded Computing

AI inferencing, deep learning, and dedicated H.265 (HVEC) Codec engines

AI & Deep Learning

5,888 CUDA cores (3rd gen), 184 Tensor Cores (3rd gen), and 46 RT Cores (2nd gen)

Condor GR4-A4500

3U VPX Video Capture & GPGPU Card with NVIDIA Ampere A4500 GPU

The Condor GR4-A4500 is a ruggedized OpenVPX 3U form factor embedded video capture and GPGPU card leveraging the cutting-edge NVIDIA Ampere architecture and the NVIDIA RTX™ platform. This 3U VPX card is designed with the NVIDIA Ampere A4500 GPU featuring 16 GB of GDDR6 graphics memory equipped with Error Correction Code (ECC), 46 RT Cores, 184 Tensor Cores, and 5,888 CUDA® cores. Distinguished by its single-slot design, the Condor GR4-A4500 is a high performance GPGPU solution tailored for AI-accelerated tasks including raw video and sensor processing, encoding, decoding, and display.

3U VPX GPGPU Solution with Multi-Sensor Processing Capabilities

The Condor GR4-A4500 offers high-performance embedded computing (HPEC) capabilities to C5ISR mission processing applications such as low-latency GPGPU processing, real-time ray tracing, deep learning (DL), and AI inferencing. This versatile solution features support for PCI Express Gen 4 for increased data transfer speeds, incorporates dedicated H.265/H.264 encode and decode engines, and supports NVIDIA GPUDirect Remote Direct Memory Access (RDMA) for streamlined data transfer operations. The Condor GR4-A4500 supports four 3G-SDI/HD-SDI inputs, four 3G-SDI/HD-SDI outputs, and one DisplayPort™ output.

Condor GR4-A4500 3U VPX Block Diagram

Product Specifications

Interface3U VPX Form Factor
1” Pitch (Conduction Cooled)
PCIe Gen 4 (x4 or x8 support)
Graphics ProcessorNVIDIA RTX A4500 GPU (Ampere Architecture)
DirectX 12, OpenGL 4.5, and Vulkan 1.2
Graphics Memory16 GB GDDR6 with Error Correction Code 256-bit Memory Interface
512 GB/s Memory Bandwidth
GPGPU Capabilities5,888 CUDA Cores. 184 Tensor Cores. 46 RT Cores.
Up to 17.66 TFLOPS FP32 Single Floating Point Performance
Supports CUDA, CUDA-X, OpenCLTM and Shader Model 5.1
H.265 (HEVC) / H.264 (MPEG4/AVC) Hardware Encode & Decode
NVIDIA GPUDirect® RDMA, NVENC, NVDEC
Video InputsFour 3G-SDI/HD-SDI
Video OutputsFour 3G-SDI/HD-SDI One DisplayPort
Software & Platform SupportWindows or Linux on x86
Operating Temperature (MIL-STD-810)-40°C to 85°C (Rugged Conduction Cooled)
Power Consumption3.3V, 5V, and 12V 50 - 90W
Vibration (MIL-STD-810)0.1 g²/Hz
Shock (MIL-STD-810)40 g
Humidity (MIL-STD-810)95% Without Condensation

Joint All-Domain Command & Control (JADC2)

Download our capabilities overview to read how EIZO is uniquely positioned to help implement JADC2 with broad capability for supporting mission-critical data.