Inspur AI: Hardware Platforms

Inspur AI offers platforms for both training and inferencing, cutting-edge systems designed to deliver optimal performance, acceleration, expansion density, and agility for various AI client needs.

TRAINING

High Density: AGX-2

8x GPUs, NVLink, PCIe

agx2-image-card
See More
High Density: AGX-5

16x GPUs, NVSwitch enabled

See More
High Expansion: NF5280M5

8x Tesla P4 GPUs

See More
High Expansion: NF5468M5

16x GPUs, NVSwitch enabled

See More
Resource Pooling: GX4

Expansion Box with NVLink, PCIe

See More
Resource Pooling: SR-AI

Expansion Box with PCIe Fabric Interconnect

See More

INFERENCE

High Expansion: NF5280M5

8x Tesla P4 GPUs

See More
High Expansion: NF5468M5-V

16x Tesla P4 GPUs
COMING SOON

Acceleration: F10A

FPGA Accelerator, Altera-based, 1.366 TFlops
COMING SOON

See More
Acceleration: S10

FPGA Accelerator
COMING SOON

Return to AI & Deep Learning page: