AVR-800-X1A Rugged AI Inference – Tesla T4 & Xeon®D-2183IT
Key Features
Processor: Intel® Xeon® Processor D-2183IT
GPU: NVIDIA Tesla T4 GPU
Operation Temperature: -200C ~550C
Complies with MIL-STD-810, MIL-STD-461
Natural Passive Convection/Conduction Cooling
Ingress Protection – IP65
The AVR800-X1A is a rugged AI inference platform from 7STARLAKE, purpose-built to support the NVIDIA® Tesla T4 GPU and Intel® Xeon® Skylake-D processor for high-performance AI computing in demanding environments.
Built on 7STARLAKE’s Open Modular, Scalable Architecture, the AVR800-X1A features an optimized cooling solution for the Tesla T4, ensuring stable and efficient operation even in harsh conditions. It also includes an M.2 NVMe slot for high-speed storage access, enhancing overall system responsiveness. Designed for advanced inference acceleration, this platform excels in voice, video, image processing, and recommendation services. The NVIDIA® Tesla T4 GPU delivers 8.1 TFLOPS (FP32) and 130 TOPS (INT8), enabling real-time AI inference on trained neural network models. With its powerful computing performance, robust expansion capabilities, and ruggedized design, the AVR800-X1A is the perfect solution for demanding edge AI applications that require high performance and reliability in extreme environments.
CPU | Intel® Xeon® Processor D-2183IT (Frequency 2.20GHz, Turbo Boost Frequency up to 3.00GHz) |
---|---|
Memory | 4x DIMMs Up to 256GB ECC RDIMM DDR4-2400MHz |
Storage | 1x M.2 2280 M key NVMe socket (PCle Gen3 x4) for NVMe SSD |
Graphics | NVIDIA Tesla T4 (2560 CUDA, 16GB GDDR6) |
Video | 1x VGA (ASPEED AST2500) / Up to 1920×1200@60Hz 32bps |
Ethernet 10GbE | 2x 10GbE |
USB | 2x USB2.0 |
Power Supply | DC-DC 18 to 36V (300W max) MIL-STD 461 |
Operating temperature | |
Storage Temperature | -40 ~ 85° C |
IP Rating | IP65 |
Environmental | MIL-STD-810G |
EMI/EMC | MIL-STD-461F |
Dimensions (W x D x H) | 400 x 311 x 155mm |
Weight | 15Kg – final weights is dependent on specific configuration |