vacancies advertise contact news tip The Vault
Win with Seagate - Seagate FireCuda 510 SSD! [x]
facebook rss twitter

Intel announces Xeon Scalable processors with up to 56 cores

by Mark Tyson on 6 August 2019, 15:11

Tags: Intel (NASDAQ:INTC)

Quick Link: HEXUS.net/qaeckc

Add to My Vault: x

Intel has announced its upcoming Xeon Scalable processor family (codename Cooper Lake) will offer up to 56 cores and built-in AI acceleration in a standard socketed CPU. Cooper Lake Xeons are scheduled to arrive in H1 2020 and will be 14nm processors that use the same socket as 10nm Ice Lake Xeons, scheduled to release later the same year.

The current range of Intel Xeons, the second gen Xeon Scalable processors (codename Cascade Lake) were launched back in April this year as BGA packages. They come with 28 physical cores at the most, thus the new Cooper Lake chips will offer up to double the core count. Intel adds that the future Intel Xeon Scalable processors will offer higher memory bandwidth, and higher AI inference and training performance compared to the standard Intel Xeon Platinum 8200 processor. Furthermore, the Cooper Lake chips will deliver this performance in a "lower power envelope than the current Intel Xeon Platinum 9200 processors," says Intel.

Another significant change in the Cooper Lake architecture is the new bfloat16 support added to Intel Deep Learning Boost (Intel DL Boost). According to sources speaking to AnandTech, bfloat16 support will only be present in Cooper Lake and not Ice Lake, and the common socket used by the upcoming processor lines will be dubbed LGA4189.

Intel's VP and GM of data centre marketing, Lisa Spelman, says that its current range of Xeons have "generated a lot of excitement among our customers who are deploying the technology to run their high-performance computing (HPC), advanced analytics, artificial intelligence and high-density infrastructure." Spelman goes on to explain that the upcoming extended 56-core processor offerings will help Intel address "a much broader range of customers who hunger for more processor performance and memory bandwidth".

You can buy Intel Xeon Platinum 9200 processors today as part of a pre-configured systems from select OEMs, including Atos, HPE, Lenovo, Penguin Computing, Megware and authorised Intel resellers. Alternatively, you might just wait to see what AMD will reveal very shortly - details of the Zen 2 core packing 7nm Epyc Rome CPUs, aimed at similar markets and applications.



HEXUS Forums :: 16 Comments

Login with Forum Account

Don't have an account? Register today!
“Standard socket” - does that mean they're going to stop changing a single pin here and there to make us all buy new mobos? Or does it just mean the socket isn't stupidly big?
Did anyone else read “bloat16 support”? In the era of custom inference & training silicon this AI feature really seems beyond too little too late.
philehidiot
“Standard socket” - does that mean they're going to stop changing a single pin here and there to make us all buy new mobos? Or does it just mean the socket isn't stupidly big?

Nope, new socket.

You can see what they're doing, they want people to buy the cooper lake so they can upgrade to ice lake. But the mobile parts really aren't that promising on 10nm at low power, will it scale high power?
philehidiot
“Standard socket” - does that mean they're going to stop changing a single pin here and there to make us all buy new mobos? Or does it just mean the socket isn't stupidly big?

I'd imagine that's their way of saying the consumer-focused socket, rather than server ones.

There's no way they're going to give up on their insistence of making a new socket as often as they can.

EDIT: Scratch that. I've now actually read the article it was mentioned in. I guess it means that the server board you put it in doesn't have to have any particular bells and whistles that may not be available on others.
Tabbykatze
Nope, new socket.

You can see what they're doing, they want people to buy the cooper lake so they can upgrade to ice lake. But the mobile parts really aren't that promising on 10nm at low power, will it scale high power?

I'll be honest, my question was hopeful but mostly just purely sarcastic.