As an Amazon Associate I earn from qualifying purchases from amazon.ca

Increasing workloads for UCS X-Sequence with UCS X-Material Expertise


Final October, I wrote a weblog introducing Cisco UCS X-Material Expertise. Final month we began paying off the promise with our first merchandise. We describe X-Sequence as having the density and effectivity of blade servers with the expandability of rack servers. First, we elevated the quantity of storage – going from two drives on a B-Sequence to 6 drives on a X-Sequence. Now we’re including GPUs to X-Sequence that had been beforehand solely obtainable in rack servers.

How does it work?

The VIC (Digital Interface Card) on the server node connects to the UCS X9416 X-Material Module. The X-Material Module connects to the UCS X440p PCIe node with the GPUs. This elegant, simply upgradable, cable-free answer is simply doable on a mid-plane, free chassis design like UCS X-Sequence.

Cisco UCS X9416 X-Material Module

The primary UCS X-Material Expertise module is PCIe Gen 4 enlargement for the UCS X210c M6 Compute Node. The 2 X9416 X-Material Modules increase the PCIe bus from the server to the UCS X440p PCIe Node. No cables, no fuss, no muss.

Cisco UCS X440p PCIe Node

Increasingly more functions can profit from accelerators like GPUs. Starting from AI/ML to the stalwart VDI, including a number of GPUs to a server can significantly enhance the consumer expertise and utility efficiency.

The Cisco UCS X440p PCIe Node lets you add as much as 4 GPUs to a Cisco UCS X210c Compute Node along with the UCS X9416 X-Material Module.

Cisco UCS X9508 Chassis with Cisco UCS X440p PCIe nodes

Totally different workloads require several types of GPUs. Cisco initially helps:

  • As much as two Nvidia A100 Tensor Core GPUs
  • As much as two Nvidia A16 GPUs
  • As much as two Nvidia A40 GPUs
  • As much as 4 Nvidia T4 Tensor Core GPUs

The modular design of UCS X-Sequence lets you decouple the CPU / GPU refresh cycles. GPU suppliers like Nvidia, AMD, and Intel launch their merchandise at a special cadence than CPUs. In case your utility efficiency is delicate to GPU efficiency, with the ability to merely slide in an UCS X440p PCIe node with the most recent GPUs permits you prolong the funding of all the opposite answer elements (chassis, servers, IFMs, & PSUs) offering a greater general TCO.

Cisco Intersight

Administration has at all times been a UCS superpower. Having the ability to handle each part for X-Sequence in a single app is paramount. From stock to firmware updates, you handle all of the UCS X-Material Expertise elements with the identical course of you handle the server and Clever Material Modules.

Run any app

Your necessities for contemporary, accelerated workloads shouldn’t dictate your server kind issue. They need to seamlessly combine right into a system that additionally runs all of your conventional functions from core infrastructure, to database, to VSI. A single system, managed from the cloud, spanning all of your workloads not solely simplifies your setting, however will help you concentrate on enterprise wants, not determining what distinctive {hardware} is required for a selected utility.

What’s subsequent?

Within the subsequent two weeks I’ll cowl the 5th Gen Material and GPUs on the X210c M6 node itself.

 


Sources

Discover the entire UCS X-Sequence blogs

 

Share:

We will be happy to hear your thoughts

Leave a reply

flyviolette
Logo
Enable registration in settings - general
Compare items
  • Total (0)
Compare
0
Shopping cart