Skip to main content

HPC in Military System

HPC

 

GPGPU AI Inference Optimized Platform: Support MXM-GPU & PEG-GPU

Nowadays, the demand of AI inference is increasing. AI inference focuses on analysis and prediction in the field, and the technology can be used in many advanced applications. GPGPU is different from the normal GPUs; GPGPU can perform computations typically conducted by a Central Processing Unit (CPU). Harnessing the power of GPUs for general purposes can be accomplished via parallel computing platforms, which allow software developers and engineers to access graphics cards and write programs that enable GPUs to manage any task that can be parallelized.

 

7STARLAKE provides GPGPU platform to offer the best solution for AI inference, the platform supports MXM-GPU and PEG-GPU to deliver high speed, low latency, more complex performance.

MXM stands for Mobile PCI Express Module, it is an interconnect standard for GPUs. 7STARLAKE’s MXM modules feature the most scalability and upgradability for GPGPU computing performance and rigorous data processing capabilities. Our MXM modules are the compact, thinnest graphics module solution, delivering the latest and cutting-edge GPU benefits for embedded systems.

PEG is short for PCIe Express Graphic cards. PCI Express is a common high-speed interface on board for graphic cards. 7STARLAKE’s PEG featuring NVIDIA embedded GPUs and connect with PCI express that is easy for integration in embedded market segments.

7STARLAKE’s GPGPU AI inference optimized platform is the ideal solution for performance demanding systems.

 

Design to certified by MIL-STD 810, Vibration (514.5), Shock (516.5)

When it comes to true ruggedness, MIL-STD-810 standard is considered the upmost principle. 7STARLAKE’s 1U and 2U Server is designed to meet MIL-STD 810 to withstand vibration and shock. With using the shock tester, it can analyze the product ability to tolerate the strong external stress and the degradation of structural weaknesses to quickly obtain the data of structure strength, anti-shock appearance and anti-drop. To assure the server can withstand detonating explosives in water and assessing the resulting high-impact mechanical shock. Mechanical Shock test is conducted to ensure that equipment can withstand drops encountered during handling, transportation, and normal use. The test is performed with both the system operating/non.

 

 


MXM-GPU Computer

Model Vehicle 2 3
AV700 SR700 AV600-TX2 AV600-T AV600 AV600-TX2-DO AV600-T-DO AV600-DO
IP Rating IP65 IP65 IP65 IP66 IP66 IP65 IP66 IP66
Rack-Mount - - - - - - - -
MXM-GPU -
PCIe/104 GPU - - - - - - - -
GPU 35W -
GPU 110W - - - -
Xeon 25W
Xeon 45W -
Xeon 65W - - - - - - - -
Xeon 100W - - - - - - - -
Swappable SSD - - - - - -
NVMe Drive - - - - - - - -
AES (H/W) Key - - - - - Option Option Option
10GbE (NIC) - - - - - Option Option Option
Swappable COMS - - - - - Option Option Option
MIL-STD 461/1275
MIL-STD 461/704 - - - - -
MIL-STD 461/1399 - -

 

 

PCIe/104 GPU Server 

Rackmount GPU Server

Model 2 3
SR800 SR800-X2 AV800 THOR100-X4-D7 THOR100-X11EH THOR200-D15EG
IP Rating IP65 IP65 IP65 IP65 IP65 IP65
Rack-Mount - - -
MXM-GPU - - - -
PCIe/104 GPU - - - - -
GPU 35W - - -
GPU 110W - - - - -
Xeon 25W
Xeon 45W -
Xeon 65W -
Xeon 100W - - - - -
Swappable SSD Option -

NVMe Drive

- -
AES (H/W) Key Option - Option Option
10GbE (NIC) - Option Option
Swappable COMS Option - Option Option
MIL-STD- 461/1275
MIL-STD- 461/704 - - -
MIL-STD-461/1399