SR800-D21A20
Military IP65 Xeon D GPU Server ® D-2183IT & Nvidia MXM A2000
- IP65 MIlitary Xeon-D GPU Server
- Intel® Xeon® D-2183IT (2.0GHz, 16 cores, 32 threads)
- NVIDIA Ampere MXM A2000 Tensor Core GPU Integrated
- DDR4-3200 MHz Up to 512GB RDIMM ECC
- 2 x 2TB MLC support H/W AES Secure Erase Key
- MIL-STD-810 Temperature, Shock, Vibration
- MIL-STD 461 EMI/EMC Certified
- Technical Profile
- Specifications
- Order Information
- Download
Edge AI Inference, NVIDIA A2000 GPU & INTEL XEON D-2183IT
The SR800-D21A20 is a ruggedized AI inference platform designed specifically for advanced inference acceleration applications such as voice, video, image, and recommendation services. This platform is powered by the NVIDIA A2000 MXM is a high-performance mobile graphics solution that is based on the NVIDIA Ampere architecture. It features 2560 CUDA cores, 20 RT cores, and 80 Tensor cores, as well as 4GB of GDDR6 memory.
The A2000 offers blazing-fast processing speeds and superior responsiveness, enabling you to tackle complex workloads with ease. Experience seamless rendering, real-time simulation, and efficient data processing like never before.
The SR800-D21A20 utilizes 7STARLAKE's Open Modular, Scalable Architecture and provides an optimized cooling solution for the NVIDIA A2000 GPU, seamlessly integrates into your existing workflow, empowering professionals across industries such as architecture, engineering, media, and entertainment. With compatibility across popular software applications and support for NVIDIA RTX™ technology, you can accelerate your creative process, reduce rendering times, and streamline your work. The A2000 is your gateway to efficient and effortless productivity.
Overall, the SR800-D21A20 is an ideal solution for customers looking for a ruggedized AI inference platform that can handle a variety of edge computing applications with ease.
I/O Placement
Dimensions
| System | ||
|---|---|---|
|
Processor |
Intel® Xeon® Processor D-2183IT (Frequency 2.0GHz, Turbo Boost Frequency up to 3.0GHz), 16 Core, 32 Thread Support, 22MB Smart Cache |
|
|
Memory type |
4 x Up to 512GB LRDIMM/256GB RDIMM, 4CH DDR4 3200MHz |
|
| Chipset | SoC, integrated with CPU | |
| GPU | ||
| NVIDIA | NVIDIA A2000 MXM GPU | |
| TFLOPS | 8.25 | |
| CUDA Cores | 2560 | |
| Memory | 4 GB GOOR6, 192 GB/sec | |
| Graphics Output | ||
| 1 x VGA | Aspeed AST2500 | |
| Resolution | Up to 1920x1200@60Hz 32bpp | |
| Storage | ||
| HDD/SDD | 2 x 2.5" 2TB MLC SSD (Easy Swappable) | |
| Front and Sise I/O | ||
| X1(DC-IN) | 1 x Amphenol TV07RW-13-04P (4PIN) | |
| X2(10GbE) | 1 x Amphenol FSIMPOFTV70ZNN | |
| X3(1GbE) | 1 x Amphenol RJFTV62A1N | |
| X4(USB3.0) | 1 x AmphenolUSB3FTV7AZNF312 | |
| VGA | D-sub15 connector with waterproof cap | |
| SSD Tray | 2 x Dual 2.5" HDD/SSD Easy Swap Tray Dedicated LED | |
| Dedicated LED | 2 x Red/Green LEDs ( SSD) | |
| Secure Erase Button | 1 x AES hardware erase button | |
| Power Requirement | ||
| Power Input | DC-DC 18 to 36V (300W max) MIL-STD 461 | |
| Applications, Operating System | ||
| Applications | C4ISR, Commercial and Military Platforms Requiring Compliance to MIL-STD-810 Process Control, where Harsh Temperature, Shock, Vibration, Altitude, Dust and EMI Conditions | |
| Operating System | Windows 10 64Bit, Windows Server 2019 64bit, Windows 2016 64bit, Hyper-V Server 2016 R2, Ubuntu16.04.3 LTS/17.10/18.04.1LTS, Fedora 25/26, RedHat Linux EL 6.8/6.9/7.3/7.4/7.6, VMware ESXi 6.5u1 ,Vmware ESXi 6.7U2 | |
| Physical | ||
| Dimension | 432x 138 x281 mm (D x H x W) | |
| Weight | 15Kg (33.06Ibs) | |
| Chassis | Aluminum Alloy, Corrosion Resistant | |
| Finish | Anodic aluminum oxide | |
| Cooling | Natural Passive Convection/Conduction Cooling. No Moving Parts Ingress Protection | |
| Ingress Protechtion | IP65 | |
| Environmental | ||
| Operating Test MIL-STD-810 | ||
| Low air pressure | Method 500.5 | Operation/Air Carriage 4572m(15.000 ft) |
| Procedure 2 | ||
| Low Temperature | Method 502.5 | -20°C, 4 hours, ±3°C |
| Procedure 2 | ||
| High Temperature | Method 501.5 | +55°C, 4 hours, ±3°C |
| Procedure 2 | ||
| Humidity | Method 507.5 |
85%-95% RH without condensation, 24 hours/ cycle, conduct 10 cycle |
| Vibration | Method 514.6 | 5-500Hz, Vertical 7.7Grms, 40mins x 3axis |
| Category 24 | ||
| Shock | Method 516.6 | 20 Grms, 11ms, 3 axes |
| Non-Operating Test MIL-STD-810 | ||
| Low Temperature | Method 502.5 |
-33°C, 4 hours, change rate:≦20°C/ Hour -15°C, 72hours (By request) |
| High Temperature | Method 501.5 | +71°C, 4 hours, change rate:≦20°C/ Hour |
| Procedure 1 | +68°C, 240 hours (By request) | |
| Vibration | Method 514.6 | 5-500Hz, Vertical 7.7Grms, 40mins x 3axis |
| Shock | Method 516.6 | 20 Grms, 11ms, 3 axes |
| Salt Fog | Method 509.7 | Salt Spray (50±5)g/L |
| MIL-STD-461 | ||
| Conducted Emissions | CE102 basic curve | 10kHz – 30MHz |
| Power Leads | ||
| Conducted Emissions | RE102-4 | 1.5HMz-30MHz-5GHz |
| Electric Field | ||
| Radiated Susceptibility | RS103 | 1.5 MHz – 3GHz, 50 V/m equal for all frequencies |
| 2MHz – 80MHz, 50 V/m equal for all frequencies | ||
| Electric Field | 80MHz – 3GHz, 50 V/m equal for all frequencies | |
| 3GHz – 5GHz, 50 V/m equal for all frequencies | ||
| Electrostatic Discharge | EN 61000-4-2 | Air DISCHARGE: 8 Kv, Contact discharge : 6kV |
| Electromagnetic compatibility | EN 61000-4-4 | Signal and DC Net: 1 kV |
| Electromagnetic compatibility | EN 61000-4-5 | Lead vs. ground potential 1Kv, ignal und DC Net: 1 kV |
| Radio disturbance | EN55022 | Class A |
| Electromagnetic compatibility | EN61000-4-3 | 10V/m |
| Electromagnetic compatibility | EN61000-4-5 | Lead vs. ground potential 1Kv, ignal und DC Net: 0.5 kV |
| MIL-STD-1275 Specifications | ||
| Steady State | 20V~33V | |
| Steady Low | 20V~33V | |
| Steady High | 18V/500ms | |
SR800-D21A20
Military MXM-GPU System GPGPU AI Inference Computer with Intel® Xeon D-2183IT Processor, NVIDIA A2000, DDR4-3200 MHz 128GB RDIMM ECC, 2x2TB MLC support hardware erase for Fast & Mass Storage, IP65 rating, MIL-STD-D38999 Connectors, 18~36V DC-IN, Extreme Rugged operating temperature -20~+60°C
SR800-D21A20-8C
Military MXM-GPU System GPGPU AI Inference Computer with Intel® Xeon D-2146NT Processor, NVIDIA A2000, DDR4-3200 MHz 128GB RDIMM ECC, 2x2TB MLC support hardware erase for Fast & Mass Storage, IP65 rating, MIL-STD-D38999 Connectors, 18~36V DC-IN, Extreme Rugged operating temperature -20~+60°C
| Attachment | Size |
|---|---|
| 7Starlake_SR800-D21_Test Report-CE RE102_20240408_all.pdf (1.87 MB) | 1.87 MB |
| SR800-D21 INTERFACE CONTROL DOCUMENT.pdf (1.52 MB) | 1.52 MB |
| SR800-D21 System test report _20230825.pdf (2.35 MB) | 2.35 MB |