SabreEDGE

GPU Servers for Deep Learning

customize

Full Customization

Build the right system for your needs, with the latest GPU, CPU, memory, & storage technology.

software

Hassle Free Software

Ships with the latest AI development software with easy update features included.

Wide GPU Selection

Wide GPU Selection

We stock a large inventory of NVIDIA GPUs. If it’s LLMs, computer vision, or machine learning, we got you covered.

AMD Powered Deep Learning GPU Servers

4GPU SabreEDGE 2U

Starting at

$7,005.21

  • Processor: 1x AMD EPYC 7002/7003 Series
  • Drive Bays: 8x 3.5"/2.5" Hot-Swap
  • Supports: Up to 4x Double-Wide PCI-E 4.0 cards
  • Rack Height: 2U
Configure
4 GPU SabreEDGE 2U

Starting at

$9,542.67

  • Processor: 2x AMD EPYC 7002/7003 Series
  • Drive Bays: 8x 3.5" Hot-swap drives
  • Supports: 4x PCI-E 4.0 x16 Double-Wide cards
  • Rack Height: 2U
Configure
8 GPU SabreEDGE 4U

Starting at

$12,959.25

  • Processor: 2x AMD EPYC 7002/7003 Series
  • Drive Bays: 10x 2.5" Hot-Swap
  • Supports: 8x PCI-E 4.0 x16 Double-Wide cards
  • Rack Height: 4U
Configure

Intel Powered Deep Learning GPU Servers

SabreEDGE-2U-1

Starting at

$9,403.92

  • Processor: 2x 3rd Gen Intel Xeon Scalable
  • Drive Bays: 8x 3.5"/2.5" Hot-Swap (2x NVMe)
  • Supports: Up to 4x Double-Wide cards
  • Rack Height: 2U
Configure
SabreEDGE-4U-12

Starting at

$13,730.70

  • Processor: 2x 3rd Gen Intel Xeon Scalable
  • Drive Bays: 12x 3.5"/2.5" Hot-Swap
  • Supports: Up to 10x Double-Wide cards
  • Rack Height: 4U
Configure

Starting at

$270,653.52

  • Processor: 2x 4th/5th Gen Intel Xeon Scalable family
  • Drive Bays: 19x 2.5" Hot-Swap
  • Supports: 8x NVIDIA H100 SXM5 GPUs
  • Rack Height: 8U
Configure
Tensorflow RN 152 Benchmark

Bring the power of NVIDIA RTX to AI workflows with NVIDIA A100 GPUs. Get up to 800 GB of ultra-fast local memory on desktop workstations to quickly process large datasets and compute-intensive workloads anywhere. Leverage the latest in AI-acceleration tools and graphics technology for stunning data visualization and configure remote access for maximum flexibility.

NVIDIA Hopper H100

The Future of Data Center Computing

During GTC 2022, NVIDIA announced their next generation Tensor Core GPU the NVIDIA H100. Built on the all-new Hopper architecture, the H100 will be their flagship GPU accelerator boasting unprecedented performance gains, and exceptional scalability, and is the future of AI development and exascale computing. Accelerate workloads with confidence and speed with NVIDIA H100.

We work with leading institutions.

Pre-Installed Software for Deep Learning

Tensorflow
PyTorch
keras
caffe
anaconda
Rapids
mxnet
caffe2
cuDNN
NVIDIA DIGITS
NVIDIA CUDA
Docker