Contact
Qualcomm Cloud AI 100 Accelerator
hpeSPT-E67D543ED636DF

Qualcomm Cloud AI 100 Accelerator

R 172 350
Available to order
3–7 day delivery

New: 12-month warranty, manufacturer-backed and mirrored locally (longer where the manufacturer's own warranty is longer).

30-day returns on unopened items — you arrange the courier; we pay it on faulty or dead-on-arrival units.

Quantity
1

Qualcomm Cloud AI 100 Accelerator.

Device Type: GPU computing processor - low profile. Bus Type: PCI Express 4.0 x8. Features: Half-height half-length (HHHL), 14 Qualcomm AI Cores, 350 TOPs peak INT8 performance, 175 Tflops peak half-precision (FP16) performance, 126 MB on-die SRAM. Size: 16 GB. Technology: LPDDR4X SDRAM. Bandwidth: 548 GBps. Power Consumption Operational: 75 Watt. Width: 1.811 cm.

Specifications

Size
16 GB

Highlights

  • PCI Express 4.0 x8 low-profile half-height half-length form factor
  • 14 Qualcomm AI Cores with 126 MB on-die SRAM
  • 350 TOPs peak INT8 and 175 Tflops peak FP16 performance
  • 16 GB LPDDR4X SDRAM with 548 GBps bandwidth
  • Energy-efficient 75 Watt operational power consumption

Memory

Size
16 GB
Technology
LPDDR4X SDRAM
Bandwidth
548 GBps

General

Device Type
GPU computing processor - low profile
Bus Type
PCI Express 4.0 x8
Features
Half-height half-length (HHHL), 14 Qualcomm AI Cores, 350 TOPs peak INT8 performance, 175 Tflops peak half-precision (FP16) performance, 126 MB on-die SRAM

Miscellaneous

Power Consumption Operational
75 Watt
Width
1.811 cm
Depth
16.765 cm
Height
6.89 cm
Weight
384 g

Related Products

The Qualcomm Cloud AI 100 Accelerator is a low-profile PCIe 4.0 x8 GPU computing processor designed for data centre AI inference workloads. Built with 14 AI Cores and 16 GB of LPDDR4X memory at 75W power consumption, it is engineered for enterprise server environments requiring high-throughput machine learning performance.

Frequently asked questions

What is the Qualcomm Cloud AI 100 Accelerator used for?

It is a specialised PCIe computing processor engineered for artificial intelligence and machine learning inference tasks in servers. It delivers high-density compute using 14 AI Cores while maintaining a low-profile footprint.

What form factor does this AI accelerator card use?

This card features a half-height half-length (HHHL) low-profile form factor that connects via a standard PCI Express 4.0 x8 interface, making it compatible with compact server chassis.

Does this accelerator include a South African warranty?

Yes, orders placed through Server Hub include a local South African warranty and support, with all pricing shown in rand.