Tuesday, August 6th 2019

Next-generation Intel Xeon Scalable Processors to Deliver Breakthrough Platform Performance with up to 56 Processor Cores

Intel today announced its future Intel Xeon Scalable processor family (codename Cooper Lake) will offer customers up to 56 processor cores per socket and built-in AI training acceleration in a standard, socketed CPU as part of its mainline Intel Xeon Scalable platforms, with availability in the first half of 2020. The breakthrough platform performance delivered within the high-core-count Cooper Lake processors will leverage the capabilities built into the Intel Xeon Platinum 9200 series, which today is gaining momentum among the world's most demanding HPC customers, including HLRN, Advania, 4Paradigm, and others.

"The Intel Xeon Platinum 9200 series that we introduced as part of our 2nd Generation Intel Xeon Scalable processor family 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. Extended 56-core processor offerings into our mainline Intel Xeon Scalable platforms enables us to serve a much broader range of customers who hunger for more processor performance and memory bandwidth."
-Lisa Spelman, vice president and general manager of Data Center Marketing, Intel Corporation
The future Intel Xeon Scalable processors (codename Cooper Lake) will deliver twice the processor core count (up to 56 cores), higher memory bandwidth, and higher AI inference and training performance compared to the standard Intel Xeon Platinum 8200 processor. The future 56-core Cooper Lake processor is expected to deliver a lower power envelope than the current Intel Xeon Platinum 9200 processors. Cooper Lake will be the first x86 processor to deliver built-in high-performance AI training acceleration capabilities through new bfloat16 support added to Intel Deep Learning Boost (Intel DL Boost ). Cooper Lake will have platform compatibility with the upcoming 10nm Ice Lake processor.

For more than 20 years, Intel Xeon processors have delivered the platform and performance leadership that gives data center and enterprise customers the flexibility to pick the right solution for their computing needs. Next-generation Intel Xeon Scalable processors (Cooper Lake) build off Intel's uninterrupted server processor track record by delivering leadership performance for customers' real-world workloads and business application needs.

Intel Xeon Platinum 9200 processors are available for purchase today as part of a pre-configured systems from select OEMs, including Atos, HPE, Lenovo, Penguin Computing, Megware and authorized Intel resellers. Learn more information about the Intel Xeon Platinum 9200 processors.
Add your own comment

56 Comments on Next-generation Intel Xeon Scalable Processors to Deliver Breakthrough Platform Performance with up to 56 Processor Cores

#1
svan71
56 is less than 64
Posted on Reply
#2
windwhirl
svan71, post: 4093069, member: 152013"
56 is less than 64
Not every bit of performance depends on core counts.
Posted on Reply
#3
svan71
In this scenario cores are king.
Posted on Reply
#4
Crackong
windwhirl, post: 4093071, member: 175818"
Not every bit of performance depends on core counts.
Yes also power consumption is a huge factor in server parts.
64 cores 225W vs 56 cores 400W , this is a no-brainer .
Posted on Reply
#5
champsilva
svan71, post: 4093078, member: 152013"
In this scenario cores are king.
Not always.
Posted on Reply
#6
Steevo
champsilva, post: 4093083, member: 164164"
Not always.
It is when leaks of the single socket 60 core part show it outperforming dual socket setups.
Posted on Reply
#7
HwGeek
Brace Yourself: More Intel's slides are coming!.
So sad to see this, they can only keep spamming the HW sites with more and more "New HW" slides.
Posted on Reply
#8
Mephis
Steevo, post: 4093094, member: 19251"
It is when leaks of the single socket 60 core part show it outperforming dual socket setups.
Links please?
Posted on Reply
#10
Patriot
Go actually talk to the vendors... they say no one wants these... due to the 400w tdp.

There are going to be no mainstream servers with these, just a handful of HPC only boxes that will have to be watercooled...
Penguin computing doesn't even list these (one of the hpc vendors in the list) Cray said they were going to make a box but have not shown it yet, that is why HPE is in the list, they recently acquired Cray, but there will be no HPE Proliants with them in them, only select Apollo servers for Megascale customers for Supercomputers.

This is Damage control before Epyc launch.
https://www.servethehome.com/amd-epyc-rome-namd-intel-xeon-computex-2019/
https://www.servethehome.com/intel-xeon-platinum-9200-series-lacks-mainstream-support/
https://www.servethehome.com/2020-cooper-lake-socketed-with-56-cores-and-bfloat16/
Posted on Reply
#11
techmagnet
champsilva, post: 4093083, member: 164164"
Not always.
56-core $10000 ..64-cores $7000 yeah no brainer.
Posted on Reply
#12
Mephis
londiste, post: 4093123, member: 169790"
https://www.tomshardware.com/news/amd-epyc-7742-vs-intel-xeon-benchmarks,40089.htmltl;dr
Assuming the leak is correct, it does in some tests. Dual socket setup being 2x28 cores vs 64.
Are we really surprised that a single socket 64 core chip beat a dual socket 28 core system in some tests? Also there were tests where a single 28 core chip beat dual socket 64 core chips. So, don't be offended if I wait for reviews by reliable parties (Anandtech, etc) versus anonymous "leaked" benches. And I'm willing to bet that both chips trade benches when that happens.
Posted on Reply
#13
Patriot
Mephis, post: 4093130, member: 186806"
Are we really surprised that a single socket 64 core chip beat a dual socket 28 core system in some tests? Also there were tests where a single 28 core chip beat dual socket 64 core chips. So, don't be offended if I wait for reviews by reliable parties (Anandtech, etc) versus anonymous "leaked" benches. And I'm willing to bet that both chips trade benches when that happens.
Give it a day. Intel is in damage control for Rome's imminent launch.

https://www.servethehome.com/amd-epyc-rome-namd-intel-xeon-computex-2019/
Posted on Reply
#14
londiste
Patriot, post: 4093132, member: 77367"
Give it a day. Intel is in damage control for Rome's imminent launch.
The thing is, Intel did that already in April with the Xeon 9200 series launch. One of the links in your post says (and given the early 2020 ETA it is likely) this current announcement is Cooper Lake. Whether Intel can improve considerably on Cascade Lake-AP when it comes to power consumption, we will have to wait and see. New platform and 8-channel memory should be the main improvement there.

On the other side of the fence, Rome is kind of out there (in supercomputers for practically a year) but still not officially released.
Posted on Reply
#15
john_
Intel is trying to convince it's customers that huge powerful chips are just around the corner, so they don't go with AMD.
As for those CPUs, probably the AI part could be interenting and something where Intel could differenciate itself from AMD. As for the cores, at 14nm,.... well, HUGE PSU...
Posted on Reply
#17
Patriot
londiste, post: 4093135, member: 169790"
The thing is, Intel did that already in April with the Xeon 9200 series launch. One of the links in your post says (and given the early 2020 ETA it is likely) this current announcement is Cooper Lake. Whether Intel can improve considerably on Cascade Lake-AP when it comes to power consumption, we will have to wait and see. New platform and 8-channel memory should be the main improvement there.

On the other side of the fence, Rome is kind of out there (in supercomputers for practically a year) but still not officially released.
This is damage control for imminent public release, STH, anand reviews incoming.
This is about cooperlake because the industry Balked at the paper launch of 9200 bga chips, these are socketed...and "less power" but not 10nm, but will be upgradeable to 10nm icelake.... in a few years.
This is very much a please don't buy epyc, we promise we are not done competing, and I am certain they are not done, but they are certainly in one of the worst positions they have ever been in on the server market.

What they do have is a clock speed advantage because of 14nm, so they will retain competitiveness in certain workloads, however they cannot flex that too strongly or they will have performance degradations with icelake 10nm. as we have seen their 10nm+ having 600-1000mhz clock reductions on the icelake laptop chips erasing the 18% ipc gains.
Posted on Reply
#18
Steevo
Mephis, post: 4093130, member: 186806"
Are we really surprised that a single socket 64 core chip beat a dual socket 28 core system in some tests? Also there were tests where a single 28 core chip beat dual socket 64 core chips. So, don't be offended if I wait for reviews by reliable parties (Anandtech, etc) versus anonymous "leaked" benches. And I'm willing to bet that both chips trade benches when that happens.
On tests where memory bandwidth matters, yes I am impressed, it means a dual socket 120 core would wipe the floor with the other brand and do it for less money, with lower TDP. All things that matter to the end user, and the person who has to pay for the server. It shows AMD has done hard work to increase the mask the cache gives them over latency and points us towards a CPU that will have great performance in the long run, not a short life before being replaced with a new and more expensive setup.
Posted on Reply
#19
Crackong
Mephis, post: 4093130, member: 186806"
Are we really surprised that a single socket 64 core chip beat a dual socket 28 core system in some tests?
No,
We are surprised a $7800 chip beating 2x $10000 chip
And a 225W chip beating 2x 205W chip.
Posted on Reply
#20
Mephis
Steevo, post: 4093163, member: 19251"
On tests where memory bandwidth matters, yes I am impressed, it means a dual socket 120 core would wipe the floor with the other brand and do it for less money, with lower TDP. All things that matter to the end user, and the person who has to pay for the server. It shows AMD has done hard work to increase the mask the cache gives them over latency and points us towards a CPU that will have great performance in the long run, not a short life before being replaced with a new and more expensive setup.
I don't think anyone can deny that AMD has made huge progress with their recent designs. I guess my biggest point of contention is the notion that Intel is doomed. There is no doubt their backs are up against a corner, but they have been there before and we all know what the result was. I personally am pulling for both companies (Nvidia too), because customers win when there is strong competition.

Crackong, post: 4093164, member: 185495"
No,
We are surprised a $7800 chip beating 2x $10000 chip
And a 225W chip beating 2x 205W chip.
Agreed that is impressive. Just want to point out that the price isn't really that big of deal when you compare it to total cost of ownership. Things like RAM, Software and Service dwarf the cost of the CPUs.
Posted on Reply
#21
HwGeek
Remember that this 225W chip beats "2*205W TDP" chips, in real test Intel uses near 600W for dual 8280.
Posted on Reply
#22
londiste
Patriot, post: 4093162, member: 77367"
This is damage control for imminent public release, STH, anand reviews incoming.
Oh, you are right. Epyc Rome should be released tomorrow. Kind of forgot about that with the constant rumored release dates over this year :)
Posted on Reply
#23
HwGeek
LOL- Intel to it's loyal customers after EPYC ROME Launch:
Posted on Reply
#24
Crustybeaver
Mephis, post: 4093171, member: 186806"
I personally am pulling for both companies (Nvidia too), because customers win when there is strong competition.
What do you mean by pulling for Nvidia? AMD is the inferior product line :confused:
Posted on Reply
#25
nemesis.ie
Re-read what he said, it sounded like he wants ALL of the vendors to have good products so we the consumer win, which is the sane way to look at it.
Posted on Reply
Add your own comment