Discussion Intel current and future Lakes & Rapids thread

[Abwx]

So that's "peak" power. And how much power does a 5800x have to itself during a CBR23 run, 142w? The fact that we're even comparing a laptop to a desktop is all you need to know about the performance of tigerlake. It's as if AMD has no laptop processors.

I was to post some numbers to answer to Timmah, so you just popped at the right time, and to be frank i had no idea other than the 5800X being pushed hard by AMD, so here a few numbers in Cinebench measured at Computerbase and wich can be compared to TGL.

Delta at the main is about 150W for the 5800X but that s a DT part that clock higher than the mobile variant since it score around 15800.
The mobile variant when pushed at about 100W score 13900.

Mobile review including previous Intel offerings and 4C TGL :

AMD Ryzen 9 5980HS & 7 5800H „Cezanne“ im Test: Testergebnisse und Benchmarks

AMD Ryzen 5000 Mobile im Test: Testergebnisse und Benchmarks / Die Testmodelle im Überblick

www.computerbase.de

And DT comparison including RKL starting with scores (no TGL included) :

Intel Core i9-11900K & i5-11600K „Rocket Lake-S“ im Test: Benchmarks in Anwendungen

Core i9-11900K und i5-11600K im Test: Benchmarks in Anwendungen / Multi-Core-Szenarien / Single-Core-Szenarien

www.computerbase.de

And powers :

Intel Core i9-11900K & i5-11600K „Rocket Lake-S“ im Test: Leistungsaufnahme und Overclocking

Core i9-11900K und i5-11600K im Test: Leistungsaufnahme und Overclocking / Leistungsaufnahme in Anwendungen / Leistungsaufnahme in Spielen

www.computerbase.de

And to make things complete since you once asked how much power TGL can use in single core i selected the thing here :

AMD Ryzen 9 5980HS & 7 5800H „Cezanne“ im Test: Testergebnisse und Benchmarks

AMD Ryzen 5000 Mobile im Test: Testergebnisse und Benchmarks / Die Testmodelle im Überblick

www.computerbase.de

Peak is at 40W and then it hoover at around 20W, so CB is more demanding than CPUZ bench wich was used by a member.

 

[eek2121]

Full system power at 172W for 30 seconds with idle power at 42W, this amount to 130W delta, that s 100W CPU power once you account for the losses in the AC adaptator and in the MB VRMs, assuming CPU power at idle is 0W...

Besides, duration of the 103W "peak" in HVinfo is 30 seconds according to the measurement at the main, how is that difficult to grasp.?.

I was to post some numbers to answer to Timmah, so you just popped at the right time, and to be frank i had no idea other than the 5800X being pushed hard by AMD, so here a few numbers in Cinebench measured at Computerbase and wich can be compared to TGL.

Delta at the main is about 150W for the 5800X but that s a DT part that clock higher than the mobile variant since it score around 15800.
The mobile variant when pushed at about 100W score 13900.

Mobile review including previous Intel offerings and 4C TGL :

AMD Ryzen 9 5980HS & 7 5800H „Cezanne“ im Test: Testergebnisse und Benchmarks

AMD Ryzen 5000 Mobile im Test: Testergebnisse und Benchmarks / Die Testmodelle im Überblick

www.computerbase.de

And DT comparison including RKL starting with scores (no TGL included) :

Intel Core i9-11900K & i5-11600K „Rocket Lake-S“ im Test: Benchmarks in Anwendungen

Core i9-11900K und i5-11600K im Test: Benchmarks in Anwendungen / Multi-Core-Szenarien / Single-Core-Szenarien

www.computerbase.de

And powers :

Intel Core i9-11900K & i5-11600K „Rocket Lake-S“ im Test: Leistungsaufnahme und Overclocking

Core i9-11900K und i5-11600K im Test: Leistungsaufnahme und Overclocking / Leistungsaufnahme in Anwendungen / Leistungsaufnahme in Spielen

www.computerbase.de

And to make things complete since you once asked how much power TGL can use in single core i selected the thing here :

AMD Ryzen 9 5980HS & 7 5800H „Cezanne“ im Test: Testergebnisse und Benchmarks

AMD Ryzen 5000 Mobile im Test: Testergebnisse und Benchmarks / Die Testmodelle im Überblick

www.computerbase.de

Peak is at 40W and then it hoover at around 20W, so CB is more demanding than CPUZ bench wich was used by a member.

Except you are the one that is incorrect. You are incorrect because you are misinterpreting the results. Anandtech already reviewed Tiger Lake and I linked their page on power consumption numbers above. I also pointed out that on NotebookCheck’s page they show Prime95 stress test as consuming 45-65 watts with a peak of 103 watts.

Prime95 is the most demanding load you can place on a CPU.

Furthermore, there is absolutely nothing wrong with a CPU briefly using more power. Both AMD and Intel chips do it. The 35W Ryzen 5980HS, for example, consumes up to 56W before settling down to 35W.

The 11800h (a 45-65W part depending on config) settles down to 45W long term.

 

[Abwx]

Except you are the one that is incorrect. You are incorrect because you are misinterpreting the results. Anandtech already reviewed Tiger Lake and I linked their page on power consumption numbers above. I also pointed out that on NotebookCheck’s page they show Prime95 stress test as consuming 45-65 watts with a peak of 103 watts.

You are deluding yourself, as if HVinfo say how much time it stay at 100W, to know how much time you should look at the measurement at the wall wich point 30s, you are assuming that HVinfo peak duration is according to your desires.

Live with it rather than posting redundantly useless posts.

Edit :

Once and for all, at NBC it score 5000pts in CB R20 at about 100W and at Anandtech here the score :

AnandTech Forums: Technology, Hardware, Software, and Deals

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

www.anandtech.com

https://www.notebookcheck.net/Razer-Blade-15-Advanced-Model-11th-Gen-Intel-2021-Review-Small-Changes-Big-CPU-Gains.548802.0.html

 

[eek2121]

You are deluding yourself, as if HVinfo say how much time it stay at 100W, to know how much time you should look at the measurement at the wall wich point 30s, you are assuming that HVinfo peak duration is according to your desires.

Live with it rather than posting redundantly useless posts.

HWInfo shows the peak, minimum, and average values. You can track these values over time by logging them to a file. The only one that is "deluding himself' is you. By your accounting, my 5900HX is a power hog because it peaks out at 65W.

If you look at the average value in HWInfo, it will tell you exactly how much power is consumed over the longer term. That is why it is the 'average'. Feel free to stop making things up in order to fit your narrative at any time. The rest of us will get back to discussing future Intel architectures in an Intel thread.

If you have actual evidence the chip is using over 100W on a consistent, ongoing basis, please feel free to provide this evidence. What you have posted above does NOT show this, it shows the opposite.

EDIT: The 5980hs uses 60% more power at peak. So does the 11980HK. The difference between them? The 11980HK is a 65W part and the 5980hs is a 35W part. My 5980HX can hit 65W without batting an eye. I haven't monitored it long term, but it would not surprise me if it hit 72W. What is my point? Looking at peak values tells you nothing.

 

[Abwx]

See above, i edited my post, so you just further made a fool of yourself since you didnt even bother to check the perfs at AT...

 

[DrMrLordX]

And this has 8 Golden Cove cores? How much gets 5950X and 32C Zen2 TR?

No, that's 8 Willow Cove cores. (11800H).

 

[eek2121]

See above, i edited my post, so you just further made a fool of yourself since you didnt even bother to check the perfs at AT...

The chip doesn’t use 100w sustained. Get lost. If you can’t contribute factual information, you really should not be posting here. If you have proof that AnandTech and other reviewers are wrong and Tiger Lake is actually consuming 100w, post it here and someone will take it seriously.

 

[DrMrLordX]

The 11800h (a 45-65W part depending on config) settles down to 45W long term.

It's not going to produce a CBR23 MT score of 12700 @ 45W though.

 

[Abwx]

The chip doesn’t use 100w sustained. Get lost. If you can’t contribute factual information, you really should not be posting here. If you have proof that AnandTech and other reviewers are wrong and Tiger Lake is actually consuming 100w, post it here and someone will take it seriously.

That s factual information, it s just that you are technologically speaking completely illiterate, so even if you have the truth in front of your eyes you ll still rely on self deception..

You brought AT to the discussion and you didnt notice the 3800 score on Cinebench R20, but you re still insisting that the chip made a score of 5000 at NBC (30% more..) using the same power as in AT s review..

Here the score/power curve in Cinebench R23 measured by Techspot that i already posted :

https://static.techspot.com/articles-info/2262/bench/27.png

And the article :

Intel Core i7-11800H Review: Tiger Lake H45 Put to the Test

The Core i7-11800H is a substantial upgrade for Intel, bringing Tiger Lake designs up to 8 cores, suitable for gaming and mobile workstation systems which compete against...

www.techspot.com

 

[DrMrLordX]

AT's test was mostly conducted with modified power settings since the testbed they got from Intel had problems with throttling and other crap. It had a 5s power draw of 75-86W followed by a long-term of 45W. Regrettably, AT did not run CBR23 on that testbed.

 

[mikk]

ADL-S QS clock speed and CB R20 ES2/QS rumor:

Intel Core i9-12900K "Alder Lake-S" Qualification Sample allegedly offers 5.3 GHz turbo clock - VideoCardz.com

Intel Core i9-12900K with 3.9/5.3 GHz clock speeds Qualification samples of Intel Alder Lake-S processors are already available for a few weeks now. It appears that the first information on these samples is slowly beginning to leak. According to a post on NGA.cn forums, a qualification sample of...

videocardz.com

 

[coercitiv]

Hardware Unboxed updated the power/performance scaling for TGL with data for a better bin. Unfortunately the test laptop could only sustain CPU power until around 65W.




I find it amazing how we still have this thread polluted with gratuitous conflict over TGL power scaling (from both sides of the isle, mind you) when we already have good quality data on the subject.

 

[jpiniero]

ADL-S QS clock speed and CB R20 ES2/QS rumor:

Intel Core i9-12900K "Alder Lake-S" Qualification Sample allegedly offers 5.3 GHz turbo clock - VideoCardz.com

Intel Core i9-12900K with 3.9/5.3 GHz clock speeds Qualification samples of Intel Alder Lake-S processors are already available for a few weeks now. It appears that the first information on these samples is slowly beginning to leak. According to a post on NGA.cn forums, a qualification sample of...

videocardz.com

I'd be surprised if the i7 K ended up being 8+4 and i5 K being 6+4. Score is good albeit I figure you are getting R20 11k at 250 W and including some benefit to DDR5.

 

[Exist50]

ADL-S QS clock speed and CB R20 ES2/QS rumor:

Intel Core i9-12900K "Alder Lake-S" Qualification Sample allegedly offers 5.3 GHz turbo clock - VideoCardz.com

Intel Core i9-12900K with 3.9/5.3 GHz clock speeds Qualification samples of Intel Alder Lake-S processors are already available for a few weeks now. It appears that the first information on these samples is slowly beginning to leak. According to a post on NGA.cn forums, a qualification sample of...

videocardz.com

One would hope that up to 3.9GHz on the efficiency cores (to say nothing of unconstrained speed) will debunk the earlier nonsense in this thread about frequency scaling being a myth or somehow not applying to Atom.

 

[Abwx]

Hardware Unboxed updated the power/performance scaling for TGL with data for a better bin. Unfortunately the test laptop could only sustain CPU power until around 65W.

View attachment 47055


I find it amazing how we still have this thread polluted with gratuitous conflict over TGL power scaling (from both sides of the isle, mind you) when we already have good quality data on the subject.

And i find it even more amusing that you are posting the very curve i posted to back my sayings, so dunno how you can get to the conclusion that "both etc", unless that s a way of yours to be "fair"...

Page 474 - Discussion - Intel current and future Lakes & Rapids thread

Page 474 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

I'd be surprised if the i7 K ended up being 8+4 and i5 K being 6+4. Score is good albeit I figure you are getting R20 11k at 250 W and including some benefit to DDR5.

If the process is the same as TGL then frequency scaling vs power will be the same, at least on a lower part of the curve, so scaling from 250 to say 125W will reduce frequency by 25%.

 

[Hougy]

ADL-S QS clock speed and CB R20 ES2/QS rumor:

Intel Core i9-12900K "Alder Lake-S" Qualification Sample allegedly offers 5.3 GHz turbo clock - VideoCardz.com

Intel Core i9-12900K with 3.9/5.3 GHz clock speeds Qualification samples of Intel Alder Lake-S processors are already available for a few weeks now. It appears that the first information on these samples is slowly beginning to leak. According to a post on NGA.cn forums, a qualification sample of...

videocardz.com

What's the probable single threaded score based on this rumor?

 

[mikk]

I'd be surprised if the i7 K ended up being 8+4 and i5 K being 6+4. Score is good albeit I figure you are getting R20 11k at 250 W and including some benefit to DDR5.

The i7 being 8+4 and i5 being 6+4 is matching the info from MLID last month. Memory speed in Cinebench is pretty much irrelevant, there won't be big gains going from DDR4-3200 to DDR5-4800 in this particular benchmark. The projected MT QS score is questionable because we don't have MT clock speeds from ES2 to QS. However if the ES2 score of 9300 is legit I believe a score over 10000 is a realistic target.

 

[JoeRambo]

Score is good albeit I figure you are getting R20 11k at 250 W and including some benefit to DDR5.

From my experience memory tuning does very little to CB scores, so i doubt DDR5 helps much if any? Overall I doubt anyone cares about CB scoring at all, it is firmly AMD turf since ZEN1 hit the market. Heck, i'd take Futuremark cpu test or GB5 anyday over CB results anyday.
I think what people are most looking forward to are day to day usage tests like gaming, browsing, compression, those two guys on internet that are into rendering are already equipped with 64C Threadrippers.

Being strong in CB unfortunately tells very little overall, while having say Tomb Raider CPU render scoring reveals strong CPUs.

 

[coercitiv]

And i find it even more amusing that you are posting the very curve i posted to back my sayings, so dunno how you can get to the conclusion that "both etc", unless that s a way of yours to be "fair"...

Get over yourself, this debate was started long before you had your say on the subject. The curve was also posted previously.

 

[jpiniero]

The i7 being 8+4 and i5 being 6+4 is matching the info from MLID last month.

I am thinking like how Rocket Lake i7 was 8/12 at the QS stage and ended up 8/16. Now it wouldn't surprise me if the lower TDP models did have the small cores cut.

 

[RTX2080]

ehhhh, that Alderlake QS leak's source seems not that reliable according to some people responsed.....
(if you ask me, my impression about that forum is there were so many 'Zen2 OC@4.8Ghz' rumor floating around there when it was 2018 and me as a victim was also cheated by them)

 

[eek2121]

ehhhh, that Alderlake QS leak's source seems not that reliable according to some people responsed.....
(if you ask me, my impression about that forum is there were so many 'Zen2 OC@4.8Ghz' rumor floating around there when it was 2018 and me as a victim was also cheated by them)

If Intel is to be believed, Tiger lake boosts to 5.3 GHz. It does not surprise me that Alder Lake, with much more power/thermal headroom, would do the same. It also explains why Intel set PL2 to 228W.

 

[RTX2080]

If Intel is to be believed, Tiger lake boosts to 5.3 GHz. It does not surprise me that Alder Lake, with much more power/thermal headroom, would do the same. It also explains why Intel set PL2 to 228W.

I'm not saying it won't reach 5.3Ghz. Just take a grain of salt of that 'leak'.

 

[uzzi38]

If Intel is to be believed, Tiger lake boosts to 5.3 GHz. It does not surprise me that Alder Lake, with much more power/thermal headroom, would do the same. It also explains why Intel set PL2 to 228W.

Those two points are entirely unrelated. PL2 being set to 228W has nothing to do with 5.3GHz ST boosts.

 

[eek2121]

Those two points are entirely unrelated. PL2 being set to 228W has nothing to do with 5.3GHz ST boosts.

Did anyone say ST boosts for Alder Lake? Intel needs the extra headroom to push clocks up. ST boosts don't matter because the chip isn't going to use that much power. Multicore boosts, however...