Discussion Intel current and future Lakes & Rapids thread

Page 569 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

epsilon84

Golden Member
Aug 29, 2010
1,142
927
136
Probably. Intel is already capable of being efficient, but they wanted impressive numbers so pulled out the stops and pushed the i9 to the limits. Only this time it's worth that power use and if you look at the gaming results it doesn't even need all of that power.

If they would have reigned in the power they still could have beat AMD, but it wouldn't be by double digits in a lot of the places they win and that's probably a worse headline than the huge power draw that they chips can pull down. They already have that kind of reputation right now anyways and people buying Intel don't seem to care too much anyways.

Basically this. Even at 14nm, Intel was fairly efficient if not for the near 5GHz boost clocks and the exponential climb in power that came with that.

Optimal performance/watt for Intel 10nm (sorry, Intel 7 ) is still around the ~4GHz mark it seems. Run a 12900K at 4.0 on the P cores (with an appropriate undervolt) and it would be comparable or better in performance/watt to a 5950X, but it would lose a lot more MT benchmarks than it wins. It would still beat a 5900X, but no one ever remembers the silver medalist!

I'm sure it was a calculated decision to cop the flak for the high power draw to have the 'fastest desktop CPU in the world'. That's the headline grabbing feat 'Intel defeats AMD' not 'it takes 250W to do so'
 

DrMrLordX

Lifer
Apr 27, 2000
21,791
11,133
136
Raptor Lake-S's IGP is the exact same as Alder Lake-S's. It doesn't get the Gen13 display upgrade that Alder Lake-P gets. Guessing this means Raptor Lake-P is the same as Alder Lake-P too.

Did you mean to say that Raptor Lake-S will have the same iGPU as Alder Lake-P?
 

Mopetar

Diamond Member
Jan 31, 2011
8,000
6,434
136
Basically this. Even at 14nm, Intel was fairly efficient if not for the near 5GHz boost clocks and the exponential climb in power that came with that.

Intel did run into a bit of a problem where they were stuck on 14nm for such a long time that they couldn't get any uplift from a new node whereas AMD was able to get a boost moving from GF to TSMC which helped out a lot on top of the other design changes they made while iterating on Zen. Intel had further problems because their designs were a bit stagnant until now and it seems like Zen caught them off guard by being able to occupy a few different product niches and outcompete Intel there right out of the gate.

But even if AMD had remained well behind Intel, Intel still would have had to compete with themselves and the lack of any big changes on either their process node or within the design of their CPUs meant that they had to keep pushing the power further to be able to reach the targeted numbers over their own previous chips. Even though AMD has regained some market share, Intel's largest competitor is still Intel themselves and plenty of consumers will gladly still buy an older Intel CPU that gives them 85% of the performance for a much smaller fraction of the price.

At least with their new design Intel has something that can compete better with AMD's high core count chips. This is probably the most excited/interested I've been about an Intel launch in quite a while. Their chips have always been quite efficient if the clocks were reigned in a bit and these are no different. I can understand why they pushed them as far as they did because at the end of the day the numbers matter most and some testers will test at lower power budgets as well so they can have numbers to show performance and numbers to show efficiency.
 

coercitiv

Diamond Member
Jan 24, 2014
6,377
12,764
136
But even if AMD had remained well behind Intel, Intel still would have had to compete with themselves and the lack of any big changes on either their process node or within the design of their CPUs meant that they had to keep pushing the power further to be able to reach the targeted numbers over their own previous chips. Even though AMD has regained some market share, Intel's largest competitor is still Intel themselves and plenty of consumers will gladly still buy an older Intel CPU that gives them 85% of the performance for a much smaller fraction of the price.
In theory you are right, but in hindsight you aren't. The global events in the last few years led to a situation in which Intel was able to sell everything they produced, no matter what they produced. Intel also grew into a veritable master of controlled performance increments, that's how we ended up with quad-cores on the consumer desktop even in 2017, a decade after we got quad desktop CPUs. Sure they had plans to increase core count with Cannon Lake, but it would still have been a controlled & incremental approach, aimed at maximizing revenue.

The power push was aimed at coping with AMD, were it not for their competitive products they could have continued the trend of pushing ST performance on a limited number of cores while getting slightly more MT performance via increased core count and node optimization. The big power numbers would have been reserved for HEDT, the platform we never speak about here anymore.
 

LightningZ71

Golden Member
Mar 10, 2017
1,658
1,939
136
Personally, I'm still of the opinion that there wasn't a whole lot that Intel changed with respect to their CPU lineup that was influenced by AMD's performance resurgence. I can point to two products that likely never would have existed without AMD pushing Intel: The 10 core Comet lake products and Rocket Lake. I suspect that we likely wouldn't have seen 8 core 9th gen parts either. It's also possible that Kady Lake-G might not have seen the light of day without AMD pushing APU performance as much as they did, though, I believe that the timings for it required that Intel have begun work on it before details about Raven Ridge were first released.

Intel's seeming malaise in that era was largely due to continually stubbing their two with their 10nm process. They wanted to use that process to start pushing higher core counts as it allowed sufficient circuit density to allow them to produce their desired number of die per wafer with those products. Instead, they continually pushed more and more products on 14nm with higher core counts, which butchered their dpw numbers.
 

tomatosummit

Member
Mar 21, 2019
184
177
116
Personally, I'm still of the opinion that there wasn't a whole lot that Intel changed with respect to their CPU lineup that was influenced by AMD's performance resurgence. I can point to two products that likely never would have existed without AMD pushing Intel: The 10 core Comet lake products and Rocket Lake. I suspect that we likely wouldn't have seen 8 core 9th gen parts either. It's also possible that Kady Lake-G might not have seen the light of day without AMD pushing APU performance as much as they did, though, I believe that the timings for it required that Intel have begun work on it before details about Raven Ridge were first released.

Intel's seeming malaise in that era was largely due to continually stubbing their two with their 10nm process. They wanted to use that process to start pushing higher core counts as it allowed sufficient circuit density to allow them to produce their desired number of die per wafer with those products. Instead, they continually pushed more and more products on 14nm with higher core counts, which butchered their dpw numbers.

Everything intel was releasing was changed by zen performance.
No matter what people might say about zen1 performance and the minor issues it had, it really did disrupt the market.
Intel desktop cpus go from 4cores to 6,8,10 and so overclocked that power draw has doubled. With 8th gen intel has started releasing the K series 6months early to recover mind share and 9th gen introduced the i9 top tier branding that amd happily followed with r9.
Laptop cpus have seen a similar trend, 2/4cores for 7 generations and through to 10gen it was now 4/8cores and 5ghz.
Xeon scalable has seen the same power draw increase and the largest dies sizes have been spread through much more of the product stack, you would see various xeon gold5000 cpus with more cache than the LCC dies could support for example.

All the above would hugely increase the silicon required for products (therefore manufacture costs) and was often coupled with a reduction in asking price in the more mainstream skus.
11th gen is a red headed step child. Tigerlake H and rocketlake are probable best left forgotten although they seemed to be the end of 10nm problems and intel's internal problems. (Fab head leaving and being able to develop architecture more dynamicly)
I would argue that if skylake consumer was 8 cores at launch (4 in small laptops), even if it's low clocked, ryzen1 would have made far less impact, and there was no good reason intel withheld that. Multiple die shrinks since nehlam with very little to show for it for consumers and amd started very strong and zen2 catapulted them ahead.

Sapphire rapids and alder lake seem to be the first real bespoke responses to zen (plus four years).
Alder is ropey, as overall performance isn't that much better for the increased costs through new ddr5 platform and heat/power to get what's advertised. I do anticipate the 2+8 cpu as it's the most interesting part, desktop should have been a single sapphire chiplet for a 15core 250w golden cove monster.
Sapphire, from what has been leaked or revealed, also doesn't show that much promise, hopefully it's a decent competitor to rome and milan but too little too late too hot. Again most interesting thing about it is not performance but the chiplet design, especially how it's very similar to naples epyc cpus but with modern high speed interconnects and also how future on package memory will compare to stacked L3.
 

LightningZ71

Golden Member
Mar 10, 2017
1,658
1,939
136
Oh, I didn't specify, but I was only referring to Desktop. In Mobile, while Raven Ridge didn't move the needle much on single core performance, it absolutely did push Intel to release cost competitive 4 core products (though, they had been releasing quad core laptops for some time in their Q lineup, just at very high prices) and also expand their iGPU in Ice Lake over their 14nm designs. Renoir was a big change in the market. Affordable and power efficient 6 and 8 cores in mobile that were performance leaders in at least some tasks was an absolute game changer. Servers, the success of EPYC speaks for itself.

I don't see Threadripper as being as big of a gain as it is sometimes portrayed. I can often by second hand server parts for less money that can outperform it in MT tasks and reasonably match it in I/O. Professionally, it has a niche, but, it's not a game changer.

But, on desktop, the change to their actual parts strategy wasn't that big. AMD definitely kept Intel's ASPs on 10th and 11th gen parts under control, and certainly forced them to be more aggressive about power usage for performance.
 

tomatosummit

Member
Mar 21, 2019
184
177
116
Oh, I didn't specify, but I was only referring to Desktop. In Mobile, while Raven Ridge didn't move the needle much on single core performance, it absolutely did push Intel to release cost competitive 4 core products (though, they had been releasing quad core laptops for some time in their Q lineup, just at very high prices) and also expand their iGPU in Ice Lake over their 14nm designs. Renoir was a big change in the market. Affordable and power efficient 6 and 8 cores in mobile that were performance leaders in at least some tasks was an absolute game changer. Servers, the success of EPYC speaks for itself.

I don't see Threadripper as being as big of a gain as it is sometimes portrayed. I can often by second hand server parts for less money that can outperform it in MT tasks and reasonably match it in I/O. Professionally, it has a niche, but, it's not a game changer.

But, on desktop, the change to their actual parts strategy wasn't that big. AMD definitely kept Intel's ASPs on 10th and 11th gen parts under control, and certainly forced them to be more aggressive about power usage for performance.

But by 10th and 11th gen the ball was already rolling in favour of amd, in the diy desktop market intel had all but lost mind share. It had spill over to oem and mainstream as well where i5s were now 6cores and then 12 threads in 10gen.
Intel's desktop plans were exposed from their initial hedt reveals or leaks. Skylake-x was originally only 6-10 cores using only the lcc die until threadripper made a mockery of intel and the the hcc based 12-18 core hedt cpus were released late.
I don't believe for a second intel would have been releasing 5ghz 10core cpus without competition, maybe a 6core to tide the masses over for three generations while 10nm is fixed and without competition 10nm could have launched without needing nonsense clock speeds on the desktop. Adding two cores to the current design was an easy fix but you can tell it wasn't anywhere past the drawing board because it came a year after ryzen, then it dragged out for far too long with 10nm failures.
Cannon lake was the original plan and who knows what the full plans for that were, all we saw was the two core laptop cpu, for desktop I think we can only guess.

But I will note threadripper 1 and 2 deserve every ounce of praise they get. Completely ruined the concept of hedt for the time. After intel was dead in that segment zen2 TR comes along and raises the prices back up.
 
Reactions: Tlh97

mikk

Diamond Member
May 15, 2012
4,168
2,205
136

interesting that Meteor Lake could be just for laptops while Arrow Lake on TSMC 3nm will be 14th gen desktop for Intel.


This is no surprise for me considering how close Arrow is to Meteor and given that Intel preferred mobile on a new process in the past (Broadwell-U, Cannonlake-U, Tigerlake-U) and given that Intel is low on EUV initially. Makes perfect sense.

Arrow Lake 1 year after Rapor Lake implies to me this is the desktop successor for Raptor Lake and no Meteor Lake for desktop as expected, at least not in 2023.

The biggest clue was this: http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...-rapids-thread.2509080/page-395#post-40470481
 

nicalandia

Diamond Member
Jan 10, 2019
3,331
5,282
136
Can any one guess how would that compared to NVIDIA and AMD future releases? at least in feature wise

 

moinmoin

Diamond Member
Jun 1, 2017
4,993
7,763
136
interesting that Meteor Lake could be just for laptops while Arrow Lake on TSMC 3nm will be 14th gen desktop for Intel.
Going by that list, if true this doesn't show confidence in the "Intel 4" node. Also new designs in Lion Cove and Skymont would first appear on a competing foundry by around a year.
 

SAAA

Senior member
May 14, 2014
541
126
116
But by 10th and 11th gen the ball was already rolling in favour of amd, in the diy desktop market intel had all but lost mind share. It had spill over to oem and mainstream as well where i5s were now 6cores and then 12 threads in 10gen.
Intel's desktop plans were exposed from their initial hedt reveals or leaks. Skylake-x was originally only 6-10 cores using only the lcc die until threadripper made a mockery of intel and the the hcc based 12-18 core hedt cpus were released late.
I don't believe for a second intel would have been releasing 5ghz 10core cpus without competition, maybe a 6core to tide the masses over for three generations while 10nm is fixed and without competition 10nm could have launched without needing nonsense clock speeds on the desktop. Adding two cores to the current design was an easy fix but you can tell it wasn't anywhere past the drawing board because it came a year after ryzen, then it dragged out for far too long with 10nm failures.
Cannon lake was the original plan and who knows what the full plans for that were, all we saw was the two core laptop cpu, for desktop I think we can only guess.

But I will note threadripper 1 and 2 deserve every ounce of praise they get. Completely ruined the concept of hedt for the time. After intel was dead in that segment zen2 TR comes along and raises the prices back up.

Plan was to release 8 cores cannonlake, there were even some leaked sample CPUs a while back. They would have had 8 core skylake + IPC in 2016-2017 had all gone well with 10 nm, regardless of AMD, but probably very pricey at the highest end. Not that pricey to be honest as 6 haswell cores weren't that expensive at $ 400, probably current I9 range. As for frequency probably 4GHz only, aiming to rise that with 10 nm improvements along with >IPC (tiger after Icelake was already in sight).

I do agree pushing for 5 GHz and more cores every year even on 14 nm was in response to AMD's Zen, even so what other option was there? New architectures would have been cool but we know how that ended up with Rocket lake... they weren't ready to backport designs.
 

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
Great test with same clocks and same DDR4 highly tuned memory.
WIth proper memory tuning ADL seems to have no trouble beating previous in gaming. Quite some effect is due to more cache, but architecturally the advantages are definately there.


The previuos pages are also fun, esp. Linpack, SuperPI where Rocket Lake has no trouble beating ADL due to faster memory subsystem. To the defence of ADL - the memory was capped at what RKL is capable off, so probably possible to extract more performance by going up with mem, but this same clock, same uncore clock, same memory setup highlights how horribly incompetent ADL memory subsystem is.

Some great content from Igor lately. Love the detail and dedication to enthusiasts.
 

DrMrLordX

Lifer
Apr 27, 2000
21,791
11,133
136
interesting that Meteor Lake could be just for laptops while Arrow Lake on TSMC 3nm will be 14th gen desktop for Intel.

No surprises there, that's kind of what I expected when Arrow Lake started showing up on speculative roadmaps for Intel, though I had thought that Arrow Lake might be another 10ESF CPU. Intel isn't getting that much N3, maybe 20kwpm, and it isn't clear how long TSMC will be under contract to supply wafers. Combined that puts Intel's projected volume of 7nm/Intel 4 + TSMC N3 at ~40 kwpm. Intel has to share that between Meteor Lake, Granite Rapids, Arrow Lake, and various Xe/Arc products. Not counting everything else they produce! And we haven't even addressed potential yield issues on Intel 4.

Yes, they're going to a chiplet strategy and will shoehorn 10ESF/Intel 7 in there where they can, but they can't/shouldn't use that node for any dice that are performance-critical. TSMC N3 and Intel 7nm/Intel 4 should both be significantly more dense and performant than 10ESF/Intel 7.
 

Bouowmx

Golden Member
Nov 13, 2016
1,139
550
146

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
Just typing this for memory
Alder Lake Gear1 up to 4200 MT/s

The point of Igors investigation was to do Apples to Apples. That meant sticking to what RKL supported max for Gear1. They also dialed in same secondaries and tertiaries.
So what was left, was pure comparison of chips at same clock speeds, same uncore speed, same memory speed and same timings.

ADL can surely go higher on Gear1, but its memory subsystem is dissapointing. I had theory, that extra latency is purely due to L3 cache being larger combined with lower default uncore clocks, but that is not the case in Igor's testing. It is just bad engineering.
 

andermans

Member
Sep 11, 2020
151
153
76

dullard

Elite Member
May 21, 2001
25,185
3,609
126
Kinda curious how that would align with https://videocardz.com/newz/intel-arrow-lake-p-gpu-rumored-to-feature-320-execution-units . Is one of those leaks busted already? (edit: even though they technically don't directly contradict each other they don't make that much sense together?)
I'm purely speculating here, but if true, it does make a lot of sense of the rumored timelines. With Raptor Lake 4Q2022 and Arrow Lake 4Q2023, that left a measly 12 months for both Raptor Lake and Meteor Lake to be around. That was quite a short lifetime for those chips. Meaning that Intel might be doing parallel plans for backup. Meteor Lake and Arrow Lake might be two forks of similar products. If Intel 4 yield is low, or if Redwood Cove/Crestmont does not perform well, then Meteor Lake would be a niche smaller-area mobile product until Arrow Lake can take over. Alternatively if TSMC N3 (which is currently in at-risk production) struggles, then Intel has Meteor Lake available to ramp up as they get more EUV instruments.

The lack of definition about the number of E-cores in Meteor Lake is also interesting. If Intel 4 yield is low, then having only 16 E cores will be easier to produce but wouldn't be very exciting as a multithreaded desktop chip in 2023. But if yield is high, then they could go up to 24 E cores.

Essentially, is this Intel's way of preventing another delay like they had with 10 nm?
 
Last edited:

Ajay

Lifer
Jan 8, 2001
16,075
8,103
136
Essentially, is this Intel's way of preventing another delay like they had with 10 nm?
Using TSMC N3 certainly is. It gives them some breathing room to step up their process technology without losing ground on the product side. Intel's orders for High NA EUV equipment from ASML can't come along fast enough (installations are expected to enter volume production lines some time in 2025, last I read).
 

Mopetar

Diamond Member
Jan 31, 2011
8,000
6,434
136
In theory you are right, but in hindsight you aren't. The global events in the last few years led to a situation in which Intel was able to sell everything they produced, no matter what they produced. Intel also grew into a veritable master of controlled performance increments, that's how we ended up with quad-cores on the consumer desktop even in 2017, a decade after we got quad desktop CPUs. Sure they had plans to increase core count with Cannon Lake, but it would still have been a controlled & incremental approach, aimed at maximizing revenue.

The global situation of the past two years have nothing to do with Intel's actions in 2017 and being able to sell everything isn't unique to Intel because even the sub-$100 Athlon parts AMD has released are now selling for well over $100.

Prior to Zen, Intel simply had no real competition outside of the low-end consumer market and after Bulldozer I don't think anyone can blame Intel for not expecting that AMD would have anything that could compete. But AMD somehow managed to find a few different high-margin niches with Zen, but built on that with Zen 2 and Zen 3 and were able to find more spots to crawl into.

Intel would probably still be selling 4-core x700K CPUs today if there weren't competition from AMD because there's no pressure for them to change. That's just the reality of any market that's dominated by a single player and if AMD somehow managed to supplant Intel, they'd do the same, which is pretty evident by the price increases in Zen 3 when they finally managed to overtake Intel.
 
Reactions: Tlh97

jpiniero

Lifer
Oct 1, 2010
14,823
5,441
136
Intel would probably still be selling 4-core x700K CPUs today if there weren't competition from AMD because there's no pressure for them to change.

Again, Intel was going to 8 cores with Cannonlake. Intel still needs to convince people to buy new even in a monopoly situation.

Even now it's still not entirely clear how much impact AMD is having on Intel's desktop volume when it mostly looks like it's based upon OEM demand which AMD is not really playing in at this point.
 

maddie

Diamond Member
Jul 18, 2010
4,787
4,771
136
Again, Intel was going to 8 cores with Cannonlake. Intel still needs to convince people to buy new even in a monopoly situation.

Even now it's still not entirely clear how much impact AMD is having on Intel's desktop volume when it mostly looks like it's based upon OEM demand which AMD is not really playing in at this point.
Sure, and charged a huge premium most probably.

You act as if computers are forever and that software demands stay constant.

What is the turnover for companies and most computers owners? 4 yrs, 5 yrs, 6yrs, 7yrs? If you dominate the system you will have 25%, 20%, 17%, 14% of the total installed base worldwide buying each year. That's a lot of revenue.

So no, Intel needing to convince people to buy new as an argument to suggest speedy innovation would ensue is flatly false. Incremental innovations worked for years when they dominated sales fully. How many here and in the wider press were excited at the small improvements each generation? People were upgrading for 10% and sounding happy to have the opportunity.

If anyone actually believes that Intel would not have chased ever higher margins if the opportunity had presented itself, then please contact me. I have an unused new bridge for sale.

AMD's impact is what any thinking person would have. The future is not as secure as we would like and we have to start reacting now. AMD, with Zen, represented that threat. Zen 2 and Zen 3 has confirmed it.

I tired of these naive arguments pretending that all would have been the same sans Zen. How many of the younger viewers see these generational improvements as what the older ones accepted as the norm, except for that period of stagnation or slow advancement.

IPC improvement is HARD, that's why you should be satisfied. Remember that argument.

Now, 20%+ per generation is accepted as a new normal. Why is that?

These attempts to forgive an unforgivable implementation of monopolistic policy frankly piss me off.
 

dullard

Elite Member
May 21, 2001
25,185
3,609
126
Intel would probably still be selling 4-core x700K CPUs today if there weren't competition from AMD because there's no pressure for them to change.
Why do you ignore the 6 core i7 desktop chips (i7 970 and i7 980) that Intel sold in 2010 and 2011 when AMD wasn't much of competition? These aren't even the Extreme Edition chips.



Review: https://www.anandtech.com/show/3833/intels-core-i7-970-reviewed-slightly-more-affordable-6core/3

The fact is that duopolies don't really compel companies to innovate much more than monopolies do. You'd be correct in true competition situations. But not in a duopoly and especially not in a duopoly when the two companies can distinguish their chips with each having a strong point that wins.
 
Last edited:
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |