P6-SW160-M-I-Double

A double configuration of the P6-SW160-M-I Gen6 GPU expansion platform — maximum GPU density in a single, validated system built on Microchip switching.

Best for: large AI training clusters and HPC GPU pools that need to scale in one step.

PCIe Gen6

Microchip inside

Double configuration

Request a Quote Download Datasheet

How It’s Used

From rack to running workloads.

Step 01

Install the platform

Rack the double expansion system and populate it with your Gen6 GPUs.

Step 02

Connect the hosts

Link each server with a Ustorage Gen6 host card, adding retimers for longer cable runs.

Step 03

Scale your workloads

Assign GPUs to training or inference jobs and grow the pool as demand increases.

Datasheet

Specifications

Values marked [ ] will be provided in the official datasheet.

Interface

PCIe 6.0 (64 GT/s per lane)

Switch silicon

Microchip [part number]

GPU slots

[to be provided]

Host connections

[to be provided]

Form factor / dimensions

[to be provided]

Power supply options

[to be provided]

Operating environment

[to be provided]

Compliance

[to be provided]

Complete the System

Works well with

Switch Board

P6-SW160-M-I-R12

GPU switch board to fan out Gen6 lanes.

View product →

Host Card

P6-SW64-M-H-R10

High-bandwidth host connection to the chassis.

View product →

Retimer Card

P6-CDFP-RT-H

Clean Gen6 signals over longer CDFP cables.

View product →

Resources

Datasheet, images & video