Qualcomm Cloud AI 100 Ultra

Cloud AI 100 PCIe 2023 7nm

Overview

—

FP32 TFLOPS

128 GB

Memory LPDDR4X

548 GB/s

Memory Bandwidth

150 W

TDP

Performance Metrics

Precision	Peak (dense)	Bits	Type
FP64	—	—	Standard
FP32	—	—	Standard
TF32	—	—	Standard
BF16	—	—	Standard
FP16	—	—	Standard
FP8	—	—	Standard
FP6	—	—	Standard
FP4	—	—	Standard
INT8	870.0	8	Standard

Memory Specifications

Capacity: 128 GB
Type: LPDDR4X
Bandwidth: 548 GB/s
Interface Width: —

Power & Efficiency

TDP: 150 W
Max power (Flopper estimate): ~173 W est.
FP32 dense TFLOPS per W of TDP: —
FP16 dense TFLOPS per W of TDP: —

Flopper estimate: 150 W TDP x 1.15. The vendor publishes no maximum board power for this part. Per W of TDP figures divide dense peak by accelerator TDP. Board power only: excludes host CPUs, networking, cooling and facility overhead.

Hardware Details

Vendor: Qualcomm

Architecture: Cloud AI 100

Process Node: 7nm

Form Factor: PCIe

Launch Year: 2023

Compute Engine

AI Cores: 64

Interconnect

PCIe: Gen4 x16

Enterprise Features

ECC: Yes