Neousys GPU Computing Systems
Supporting up to Intel® Xeon® E and 9th/ 8th-Gen Core™, Neousys edge AI computing solutions offer unparalleled performances with true wide-temperature operation to ensure CPU/ GPU do not thermal-throttle under harsh conditions. With an array of ruggedized embedded GPU solutions that feature NVIDIA® Tesla, Quadro®, Jetson Xavier, RTX 30 series consumer-grade graphics cards and Google TPU, Neousys GPU computing solutions are ideal and can be found in medical imaging, video analysis, deep learning machine vision, autonomous machines and more.
Complete GPU Support
Ranging from RTX 30 series, Tesla, Quadro, Jetson™ Xavier to Google TPU for high-performance or power-efficient applications.
Patented Thermal Design
Offering better heat distribution and dissipation for optimal performance to prevent CPU/ GPU from throttling.
Ignition Power Control
Built-in ignition control to safely shutdown and startup the system.
Multi-GPUs via Single Wide-range DC Input
Accepting a wide range DC input from 8 to 48V, and requires only a single source of power input to sustain operation for dual high end RTX 30 series GPU cards.
Adaptive GPU Bracket
The patented adaptive GPU bracket ensures installed graphics cards are always secured in position to withstand shock and vibration.
Rich I/Os with Screw-lock Machanism
Available with an abundance of I/Os and screw-lock mechanism for reinforced connections.
Patented Damping Bracket
The patented damping bracket effectively absorbs shock and vibration up to 3Grms for reliable and stable operations.
PCIe/ PCI add-on slots allow for connectivity or functionality expansion.
NVIDIA® Tesla T4
NVIDIA® Quadro P2200
NVIDIA® Jetson AGX Xavier™
NVIDIA® RTX 30 Series
Mainstream and Google TPU
Supporting up to dual NVIDIA® RTX30 Series to produce up to dual 59 TFLOPs in FP32, the platform features patents proven by Baidu Apollo that offer reliability & accelerated inference capabilities.
Factory AI Automation
Supporting up to two high-end GPUs, it can simultaneously process video analytics, complex computing algorithms and neural network models on field-side for real-time video analytics.
Being ultra-compact and extremely power-efficient, Google TPUs and Jetson Xavier delivers workstation graphics processing power to AGVs for longer battery operation time.
Featuring extreme-ruggedness, unique fanless thermal design, IP67 waterproof with Tesla T4 or Quadro P2200 inference accelerator, they are capable of operating in volatile environmental conditions.
Medical Imaging & IVA
As the most compact NVIDIA® Tesla qualified server, it supports up to dual Tesla T4s for a maximum of 260 TOPS in INT8 computations and thrives at complex parallel workloads.
Utilizing TensorFLow framework, edge TPU is ideal for high volume 8-bit precision computations and when coupled with GMSL cameras, it offers real-time facial recognition technology on the go.