Discussion Intel current and future Lakes & Rapids thread

[Exist50]

The 56C/112T 8480+ will have a base clock of 2.00 Ghz and boost of 3.00 Ghz. The Ice Lake 8380 with 40C/80T have a base of 2.30 and max boost of 3.4 Ghz. That's a fact.

We've seen Golden Cove in Alder Lake, and we know it clocks far higher than Sunny Cove across a wide range of power levels. That is a fact. So why are you so quick to throw out everything we actually know?

For example, why are you so certain that 3GHz is a single core boost speed (as the Ice Lake number is), and not all core?

 

[nicalandia]

We've seen Golden Cove in Alder Lake, and we know it clocks far higher than Sunny Cove across a wide range of power levels. That is a fact. So why are you so quick to throw out everything we actually know?

You know better than that my friend. Client ≠ Server. Server CPUs lots of other things to worry about as far as power consumption goes, Ever Larger Ring Interconnect, 8 Channel Memory. Sapphire Rapids is quite the complex CPU, so complex that Intel has had to delay it for validation purposes on multiple times is we don't even know when they will finally release it.

For example, why are you so certain that 3GHz is a single core boost speed (as the Ice Lake number is), and not all core?

YuuKi_AnS has already released the Release SKU list ahead of launch. And that 8480+ base speed is identical to the release product.

 

[coercitiv]

YuuKi_AnS has already released the Release SKU list ahead of launch. And that 8480+ base speed is identical to the release product.

Is this the Release SKU list?

 

[nicalandia]

Is this the Release SKU list?

View attachment 65114

No, this is the list


From: Yuuki_AnS @ yuuki_ans

◇ Xeon Platinum 8400 (Sapphire Rapids-SP / Intel 7 / LGA4766 / 8ch DDR5)
　 8490H 60-core / 120-thread 1.90-2.90GHz TDP350W
8480+ 56-core / 112-thread 2.00-3.00GHz TDP350W
　 8471N 52-core / 104-thread 1.80-2.80GHz TDP300W
　 8470Q 52-core / 104-thread 2.00-3.00GHz TDP350W
　 8470N 52-core / 104-thread 1.70-2.70GHz TDP300W
　 8470 52-core / 104-thread 2.00-3.00GHz TDP350W 8468V
　 48- core / 96-thread 2.40-2.90GHz TDP330W
　 8468H 48-core / 96-thread 2.10-3.00GHz TDP330W
　 8468 48-core / 96-thread 2.10-3.00GHz TDP350W
　 8461V 48-core / 96-thread 2.20-2.80GHz TDP300W
　 8460Y + 40-core / 80-thread 2.00-2.80GHz TDP300W
　 8460H 40-core / 80-thread 2.20-3.10GHz TDP350W
　 8458P 44-core / 88-thread 2.70-3.20GHz TDP350W
　 8454H 32-core / 64-thread 2.10-2.70GHz TDP270W
　 8452Y 36-core / 72-thread 2.00-2.80GHz TDP300W
　 8450H 28-core / 56-thread 2.00-2.60GHz TDP250W
　 8444H 16-core / 32-thread 2.00-2.80 GHz TDP320W

◇ Xeon Gold 6400 (Sapphire Rapids-SP / Intel 7 / LGA4766 / 8ch DDR5)
　 6454Y + 32-core / 64-thread 2.60-3.80GHz TDP270W
　 6454S 32-core / 64-thread 2.20-2.80GHz TDP270W
　 6448Y 32-core /64-thread 2.20-3.30GHz TDP225W
　 6448H 32-core / 64-thread 2.20-3.82GHz TDP225W
　 6444Y 16-core / 32-thread 3.50-4.10GHz TDP270W
　 6442Y 24-core / 48-thread 2.60-3.00GHz TDP225W
　 6441V 44 -core / 88-thread 2.10-2.60GHz TDP270W
　 6438Y + 32-core / 64-thread 1.90-2.10-3.00GHz TDP205W
　 6438N 32-core / 64-thread 2.00-3.00GHz TDP205W
　 6438M 32-core / 64-thread 2.30-3.10GHz TDP205W
　 6434H 8-core / 16-thread 4.00-4.10GHz TDP205W
　 6434 8-core / 16-thread 3.90-4.20GHz TDP270W
　 6430 32-core / 64-thread 1.90-3.00 GHz TDP270W
　 6428N 32-core / 64-thread 1.80-2.70GHz TDP185W
　 6426Y 16-core / 32-thread 2.60-3.50GHz TDP185W
　 6421N 32-core / 64-thread 1.80-2.70GHz TDP185W
　 6418H 24-core / 48-thread 2.00 -3.00GHz TDP185W
　 6416H 18-core / 36-thread 2.20-3.00GHz TDP165W
　 6414U 32-core / 64-thread 2.00-2.60GHz TDP250W

◇ Xeon Gold 5400 (Sapphire Rapids-SP / Intel 7 / LGA4766 / 8ch DDR5)
　 5420+ 28-core / 56-thread 1.90-2.10GHz TDP205W
　 5418Y 24-core / 48-thread 2.10-2.90GHz TDP185W
　 5418N 24-core / 48-thread 2.00-2.80GHz TDP165W
　 5416S 16-core / 32-thread 2.10-2.90GHz TDP150W
　 5415+ 8-core / 16-thread 2.90-3.70GHz TDP150W
　 5411N 24-core / 48-thread 2.00-2.80GHz TDP165W

◇ Xeon Silver 4400 (Sapphire Rapids-SP / Intel 7 / LGA4766 / 8ch DDR5)
　 4416+ 20-core / 40-thread 2.10-3.00GHz TDP165W
　 4410T 12-core / 24-thread 2.10-3.00GHz TDP145W
　 4410T 10-core / 20-thread 2.90-3.10GHz TDP150W

◇ Xeon Bronze 3400 (Sapphire Rapids-SP / Intel 7 / LGA4766 / 8ch DDR5)
　 3408U 8-core / 16-thread 1.80-1.90GHz TDP125W

北森瓦版 - 【怪情報】“Sapphire Rapids-SP”のラインナップ？

北森瓦版 - Northwood Blog (Author : 北森八雲. Since July 10, 2006.)

northwood.blog.fc2.com

 

[coercitiv]

No, this is the list

That's a list posted by @yuuki_ans on July 8, and I just provided you with an updated list from July 17, by the same leaker.

The least you can do is acknowledge that and check out the info.

 

[nicalandia]

That's a list posted by @yuuki_ans on July 8, and I just provided you with an updated list from July 17, by the same leaker.

The least you can do is acknowledge that and check out the info.

View attachment 65119

How is what you posted even comparable as far as full list goes? You just posted a screenshot of the test he perform on two Sapphire Rapids(one QS sample and one early sample) and an early sample of Genoa(The first post I did not acknowledge was a twitter post about the test and this one is a single picture of the test bed)

It's the same source by the way(YuuKi_AnS) but the earlier source I provided is actually in plain text and not a Huge Picture that may or may not be easily readable.

The point is the 8480+ That was benched on GB is either a very late E3 QS sample or a full release CPU due to them having the same base clock of 2.00 Ghz.

 

[coercitiv]

How is what you posted is even comparable as far as full list goes?

Are you having difficulties reading the 3.8Ghz boost speed for Xeon 8480+?

 

[nicalandia]

Are you having difficulties reading the 3.8Ghz boost speed for Xeon 8480+?

Not interested on Boost Clock, You Quoted me asking if what you posted was the release SKU, I pointed and posted the full list. Base clock is where most Server CPUs spend their time.

I'ts 2.0 Ghz for that 8480+ wheather it's a QS or Final release sample.

 

[Markfw]

Are you having difficulties reading the 3.8Ghz boost speed for Xeon 8480+?

Boost speed is for like ONE core, correct ? When is a server EVER going to have one core out of 56 ????

 

[coercitiv]

Not interested on Boost Clock

You specifically mentioned a boost of 3Ghz for 8480+ and a boost of 3.4Ghz for Ice Lake, and then ignored @Exist50 question on whether you were certain this was a single core boost in the case of SPR.

If you have no interest in boost clocks that's OK with me, so don't mention them as part of your argument, makes following the discussion that much easier!

Boost speed is for like ONE core, correct ? When is a server EVER going to have one core out of 56 ????

I suggest you take the time and read the replies, this is not about boost speed per se, more about @nicalandia using partial leaks as FACTS.

 

[nicalandia]

You specifically mentioned a boost of 3Ghz for 8480+ and a boost of 3.4Ghz for Ice Lake, and then ignored @Exist50 question on whether you were certain this was a single core boost in the case of SPR.

I was thinking more about the base clocks since they are what matters most. I mean a max boost are really irrelevant for this segment.

 

[Markfw]

You specifically mentioned a boost of 3Ghz for 8480+ and a boost of 3.4Ghz for Ice Lake, and then ignored @Exist50 question on whether you were certain this was a single core boost in the case of SPR.

If you have no interest in boost clocks that's OK with me, so don't mention them as part of your argument, makes following the discussion that much easier!


I suggest you take the time and read the replies, this is not about boost speed per se, more about @nicalandia using partial leaks as FACTS.

I did read the replies. I am just not getting into that. Its not released yet, so they are still leaks IMO.

 

[coercitiv]

I was thinking more about the boost clocks since they are what matters most.

You probably meant base clocks... it's late here, maybe it's late on your side as well. Let's call it a night, SPR isn't going anywhere.

 

[nicalandia]

I did read the replies. I am just not getting into that. Its not released yet, so they are still leaks IMO.

Don't pay attention to those folks, since when Max Boost have been set in stone by Intel or AMD? I mean the 5950X is rated at 4.9 and it everyone knows they boost past 5 Ghz all of the time. The final product will have a max boost of 3 Ghz wheather the CPU boost past that it's something entirely different. Base boost is where the work is done.

 

[Exist50]

You know better than that my friend. Client ≠ Server.

I'm comparing to another server chip. Coincidentally, one also plagued by delays, but on a process with far worse performance.

YuuKi_AnS has already released the Release SKU list ahead of launch. And that 8480+ base speed is identical to the release product.

You're not answering the question. Why specifically do you think 3GHz is a single core boost number instead of the far more reasonable all core? As @coercitiv points out, you're even ignoring that very leaker's numbers.

 

[Markfw]

Don't pay attention to those folks, since when Max Boost have been set in stone by Intel or AMD? I mean the 5950X is rated at 4.9 and it everyone knows they boost past 5 Ghz all of the time. The final product will have a max boost of 3 Ghz wheather the CPU boost past that it's something entirely different. Base boost is where the work is done.

Since I have several EPYC, I have never seen the boost clock come into play, and I am not a datacenter.

 

[nicalandia]

You're not answering the question. Why specifically do you think 3GHz is a single core boost number instead of the far more reasonable all core? As @coercitiv points out, you're even ignoring that very leaker's numbers.

Max Boost for Release CPU was leaked to be 3.00, but as I mentioned before, max boost set my Intel or AMD is not set in stone. the QS sample was clearly boosting to 3.5 Ghz in Cinebench. Again it's a QS E3 sample do you believe that the release 8480+ CPUs will have a much higher max boost clocks? Does it make any difference since the base boost is what the CPU will need to manage within the TDP?

 

[Exist50]

Max Boost for Release CPU was leaked to be 3.00

That very leaker says 3.8 for max boost, not 3.0. It's silly to keep on doubling down on something your source disagrees with, and doubly so when it defies common sense.

 

[nicalandia]

That very leaker says 3.8 for max boost, not 3.0. It's silly to keep on doubling down on something your source disagrees with, and doubly so when it defies common sense.

It's my bad, I said 3.5 But YuuKi_AnS posted 3.8 Ghz single core boost on one of the tests.

Still that Geekbench 5 is actually not a bad score for an All Core speed of 2.00, its withing what can be expected for that many core Golden Cove.

 

[mikk]

There is the reason for the Sapphire Rapids delay:

https://twitter.com/x/status/1552778451756929025

It won't come out this year, this is a big fail.


Intel 4 and MTL sounds promising so far.

https://twitter.com/x/status/1552777744718651392

 

[nicalandia]

There is the reason for the Sapphire Rapids delay:

https://twitter.com/x/status/1552778451756929025

It won't come out this year, this is a big fail.

https://twitter.com/x/status/1552777744718651392

I expect some SKUS to be released at the end of this year to save face, but not the entire stack.

 

[Exist50]

Intel 4 and MTL sounds promising so far.

Intel 4 might be fine, but I'll believe MTL is healthy when I see it in reviewers' hands.

 

[DrMrLordX]

Let's call it a night, SPR isn't going anywhere.

More truth than you might have intended.

It won't come out this year, this is a big fail.

Depends on how many select customers are happy with the early silicon they're getting. Lots of units can be sent to hyperscalars via ODMs that will fly under our radars until they're reflected in earnings reports somehow. The real question is: what is the bug, and who is affected?

 

[coercitiv]

More truth than you might have intended.

The pun was intended, if that's what you meant.

Lots of units can be sent to hyperscalars via ODMs that will fly under our radars until they're reflected in earnings reports somehow.

That's just it, in the Q&A Intel admitted the bulk of SPR contribution will be in 2023.

We said it's ramping later, main SKUs will contribute mostly next year than this year. We see opportunity in client ASP increases passing on inflation. Our competitiveness is not as strong on DCAI so can't increase pricing. Mostly ramp next year.

 

[moinmoin]

There is the reason for the Sapphire Rapids delay:

https://twitter.com/x/status/1552778451756929025

Charlie indicated it's security related (and that the delay was the right decision). Makes me wonder what security if there are server use cases were it doesn't matter for those select customers.