• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

Discussion Intel current and future Lakes & Rapids thread

Page 478 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

mikk

Diamond Member
May 15, 2012
3,242
1,057
136
That's absurd.

This is the reality. It's completely different between Intel and AMD when it comes to unconfirmed informations in basically every forum. Good leaks from Intel must be fantasy and things like that, there is much more scepticism involved. On AMD a huge hype train will grow and this is also a reason why there are more AMD related fakes in the web historically by the way. It's like that for a very long time, no matter if it was the Bulldozer or Zen+ era. I know that most people don't like what I said given that the big majority is pro AMD but it's reality. DrMrLordX above actually is a good example, he can't deal with it being Raja Koduri (possibly) right on ADL-S and comes up with this spam, some people would never admit he was right.
 

Carfax83

Diamond Member
Nov 1, 2010
6,068
870
126
Why would Gracemont be closer to Core specifically in AVX? If its capabilities were that good, it would probably support AVX512, so if anything, I expect AVX performance to be a weakness. And also think people should scale back expectations to Skylake IPC level. Kind of funny how fast that went from nearly unheard of to the low end of speculation.
+1.

I would also like to know what sources @eek2121 is basing that claim on. I mean it would be nice if it's true, but stereotypically vector performance is not a strong suit for low wattage cores that are expected to attend mostly to background tasks or to "assist" the big cores.
 
  • Like
Reactions: uzzi38 and Tlh97

Thunder 57

Golden Member
Aug 19, 2007
1,676
1,724
136
This is the reality. It's completely different between Intel and AMD when it comes to unconfirmed informations in basically every forum. Good leaks from Intel must be fantasy and things like that, there is much more scepticism involved. On AMD a huge hype train will grow and this is also a reason why there are more AMD related fakes in the web historically by the way. It's like that for a very long time, no matter if it was the Bulldozer or Zen+ era. I know that most people don't like what I said given that the big majority is pro AMD but it's reality. DrMrLordX above actually is a good example, he can't deal with it being Raja Koduri (possibly) right on ADL-S and comes up with this spam, some people would never admit he was right.
Raja is a joke, and so is your post. Also, I wasn't aware that alder Lake was out and it was a big success.
 
Last edited:
  • Like
Reactions: Tlh97

mikk

Diamond Member
May 15, 2012
3,242
1,057
136
Raja is a joke, and so is your post. Also, I wasn't aware that alder Lake was out and it was a big success.

Raja is a joke because he works for the devil Intel, right? Your response is the biggest joke, at least try harder next time. Reality can be hard sometimes and it was quite expected.
 

Mopetar

Diamond Member
Jan 31, 2011
6,177
3,006
136
Raja is a joke because he works for the devil Intel, right?
I think it has more to do with his past history of overpromising on projects that he worked on and underdelivering. I can't say I have much love for him either, but that'd be true of any company he works at. I suspect that there are probably some hardcore AMD/Radeon fans that are glad he left.

But I wouldn't worry too much since I don't think Raja had much of anything to do with Sunny Cove or Gracemont. It'd be kind of weird for him to make up BS figures about something he didn't work on at all. Anything he got and talked about came from someone else first.
 
  • Like
Reactions: uzzi38

Thunder 57

Golden Member
Aug 19, 2007
1,676
1,724
136
Raja is a joke because he works for the devil Intel, right? Your response is the biggest joke, at least try harder next time. Reality can be hard sometimes and it was quite expected.
Oh go to hell. I never said Intel was "the devil". Nice projecting. He was a joke at AMD as well. Hell, at least at Intel he didn't say anything too stupid. And what reality? That guy has always been overrated IMO, show me what he has done that is impressive.

Based on post history, if anything you would say AMD is "the devil" long before I said Intel was. Why anyone would want one company in a duopoly to fail is beyond me. We saw what happened when AMD couldn't compete. I don't want to see that again.
 
Last edited:

vstar

Member
May 8, 2019
41
36
61
Why would Gracemont be closer to Core specifically in AVX? If its capabilities were that good, it would probably support AVX512, so if anything, I expect AVX performance to be a weakness. And also think people should scale back expectations to Skylake IPC level. Kind of funny how fast that went from nearly unheard of to the low end of speculation.
I think that the AVX capabilities with Gracemont have been greatly improved.

Here are some sources, but we should know for sure in a few months:


 

A///

Golden Member
Feb 24, 2017
1,023
746
136
Wasn't Raja 'demoted' to mainly the graphics division by Gelsinger? This doesn't read like a promotion.

Koduri, who previously served as Intel’s general manager of architecture, graphics and software, is an innovator in GPU computing technology and will now lead the new Accelerated Computing Systems and Graphics Group. Focusing on high performance computing (HPC) and graphics, the new unit is chartered with delivering HPC and graphics products for customers across client, enterprise and data center markets.
 

DrMrLordX

Lifer
Apr 27, 2000
17,818
6,802
136
As much as a like throw stones at Raja, the situation is different.
At AMD he was "the kind that came back", remember? At Intel he is just another hire.
Raja will always be Raja. T he question is whether or not Gelsinger will tolerate some of his more-outlandish shenanigans. It should be noted that Koduri has nothing to do with the development of Alder Lake.

Wasn't Raja 'demoted' to mainly the graphics division by Gelsinger? This doesn't read like a promotion.
Seems so. Koduri was allegedly jockeying for more power within Intel during the Murthy dust-up. Doesn't seem like he came out all that well by the time Gelsinger replaced Swan. Ah, office politics!
 

coercitiv

Diamond Member
Jan 24, 2014
4,546
6,258
136
AVX2 performance should be better. Why is 512 being mentioned? It's still a fairly niche instruction set.
Probably because Gracemont is curently being "valued" as a heavy hitter in AVX, which is characteristic of big cores meant for desktop/server environments. There's a paradox here, since Gracemont is supposed to be the lean core that takes little space and eats little power. The reasoning is that if you go through the trouble of adding AVX capabilities that rival Sunny Cove, you're no longer making a small core and might as well go all the way. (and lack AVX512 is used here in an attempt to prove to contrary, supposedly means that AVX capabilities in general aren't at SC level)

The counter-argument one might introduce is that GC Gracemont is the new "medium" core, which is supposedly being groomed into replacing Cove cores in many products starting with the next generations. This would probably take an iterative approach, which would see GC GM and it's lineage turn into "jacks of all trades, but masters of none" - strong allrounders rapidly morphing into high performance cores.

I gues you can pick your poison from the above :)

I am more curious how inter core group switching will be handled. Someone mentioned latency on a comments board a few days ago and I haven't found anything discussing such matters.
Latency has been discussed on this thread as well. The bottom line is we know there will be a penalty to pay, but current data isn't helping us much into establishing a baseline for that penalty. The only semi-relevant information we have so far is the one derived from Lakefield, which is problematic since the chip is aimed at very low power operation and may have been tuned for higher overall latency. (inter-core within the small cluster was ~50ns, jumping to the big core took an extra ~10ns).

Anyway, the time for voicing concerns about ADL-S hybrid implementation is over, we have now entered the hype season, for better or worse.
 
Last edited:

IntelUser2000

Elite Member
Oct 14, 2003
7,597
2,458
136
I think that the AVX capabilities with Gracemont have been greatly improved.

Here are some sources, but we should know for sure in a few months:
That doesn't mean AVX is significantly better than Skylake. It's identical to Skylake.

L1I cache is bigger because likely it better feeds the dual cluster decode. Skylake has the uop cache, in addition to the 4-wide decode.

L2 cache is bigger on Gracemont sure, but the reason middle-sized L2 caches exist in CPUs like Skylake is because it fills the gap between the super-fast but tiny L1 caches and slow but large L3 caches.

The Load/Store units are enlarged every time Intel increases vector capabilities, and Skylake's L/S capabilities are 4x Tremont. Even if Gracemont has 4x Tremont, that just makes it equal to Skylake not better.

In nowhere in that chart shows that Gracemont is better than Skylake, nevermind for AVX!

EDIT: specifically, the 11900k scored 69% of what the 5950X according to Anandtech. Golden cover is said to be 20% faster, but let’s say it is only 10% faster…
You are looking at the wrong data. It's 69% of 5900X. To compare to 5950X you have to search for 5950X. It looks like it's 58% of 5950X. That's a long ways to go.

You are also saying that Gracemont and Golden Cove hybrid will work perfectly with no overhead.

I am not saying it can't happen but extremely skeptical considering having two way different cores work together is happening for the first time on Alderlake. In Lakefield it set ST for Sunny Cove and MT for Tremont. And it still had minor issues.

You are saying in the Windows world, being complicated as it is Intel will get it perfectly right in the first time implementation? That with tens of millions of applications, and settings, and Windows being notorious for crappy implementations and bugs will get all working no problems whatsoever.

Fairy tales. Real world Fan Fictions. Why don't we go ask the fairy godmother to bestow blessings on our favorite company?
 
Last edited:

Exist50

Senior member
Aug 18, 2016
329
352
136
Regarding scaling, Cinebench will likely be very low overhead, and close to ideal. They're just parallel render threads. Not much complexity. The real test will be more mixed workloads like gaming and content creation.
 
  • Like
Reactions: uzzi38

A///

Golden Member
Feb 24, 2017
1,023
746
136
Seems so. Koduri was allegedly jockeying for more power within Intel during the Murthy dust-up. Doesn't seem like he came out all that well by the time Gelsinger replaced Swan. Ah, office politics!
That's precisely what I thought. I'm not sure what he was angling for during said dust-up because Murthy was made redundant for deceiving Swan and Co about everything. Gelsinger tapping former Intel employees to come back is interesting enough to warrant a separate thread. Refresh my memory, aside from those all star employees moving on, didn't Murthy fire a few thousand engineers in various points of their career?
 
  • Like
Reactions: Tlh97

Carfax83

Diamond Member
Nov 1, 2010
6,068
870
126
The counter-argument one might introduce is that GC is the new "medium" core, which is supposedly being groomed into replacing Cove cores in many products starting with the next generations.
I'm confused by this sentence. When you say GC I'm assuming you mean Golden Cove right? But Golden Cove is the high performance core so how could it be groomed into replacing Cove cores when it's already a Cove core?

I think you meant to say Gracemont or GM, unless there's something I don't know about.
 
  • Like
Reactions: uzzi38 and Tlh97

Gideon

Golden Member
Nov 27, 2007
1,439
2,889
136
if Gracemont is as good as these leaks suggest and Raptor lake can easily fit 16, Intel is missing a huge opportunity not offering 128 core SKUs (and perhaps also smaller 80 or 64 core ones) for hyperscalers.

It would fit many web workflows much better than big cores and counter Altera's offerrings very well.
 

ASK THE COMMUNITY