Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads

Page 128 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
T

Tigerick

Senior member
Apr 1, 2022
781
748
106
PPT1.jpg
PPT2.jpg
PPT3.jpg



As Hot Chips 34 starting this week, Intel will unveil technical information of upcoming Meteor Lake (MTL) and Arrow Lake (ARL), new generation platform after Raptor Lake. Both MTL and ARL represent new direction which Intel will move to multiple chiplets and combine as one SoC platform.

MTL also represents new compute tile that based on Intel 4 process which is based on EUV lithography, a first from Intel. Intel expects to ship MTL mobile SoC in 2023.

ARL will come after MTL so Intel should be shipping it in 2024, that is what Intel roadmap is telling us. ARL compute tile will be manufactured by Intel 20A process, a first from Intel to use GAA transistors called RibbonFET.



LNL-MX.png

Intel Core Ultra 100 - Meteor Lake

INTEL-CORE-100-ULTRA-METEOR-LAKE-OFFCIAL-SLIDE-2.jpg

As mentioned by Tomshardware, TSMC will manufacture the I/O, SoC, and GPU tiles. That means Intel will manufacture only the CPU and Foveros tiles. (Notably, Intel calls the I/O tile an 'I/O Expander,' hence the IOE moniker.)



Clockspeed.png
 

Attachments

  • PantherLake.png
    PantherLake.png
    283.5 KB · Views: 24,025
  • LNL.png
    LNL.png
    881.8 KB · Views: 25,516
Last edited:
  • Like
Reactions: LaudaSur, jorden6655, john404 and 15 others
inf64

inf64

Diamond Member
Mar 11, 2011
3,884
4,691
136
eek2121 said:
FYI, Arrow Lake has a significant drop in power consumption. From recent leaks, it is more efficient than Zen 4. PL2 is said to be down to 177W.

If that 5%/15% is over Raptor Lake Refresh, that puts the chip well ahead of the competition in terms of both absolute performance as well as perf/watt.

Intel is also said to be launching Battlemage GPUs.

Will Zen 5 bridge the gap? Probably. I stand by my performance projections. Zen 5 and Arrow Lake will be extremely close in terms of performance. I won’t speculate perf/watt competitiveness, because there have been claims Intel is using TSMC for some parts, and that muddies the waters. I suspect Intel may end up with a slight edge, but we will see.

I will be considering Intel for my next upgrade if these rumors turn out to be true.
Click to expand...
www.computerbase.de

CPU-Benchmark: Prozessor-Vergleich 2025

Welche CPU für Games und Apps kaufen? ComputerBase leistet Kaufberatung mit Intel Core Ultra, AMD Ryzen 9000 und 7000 auf Basis von Tests.
www.computerbase.de www.computerbase.de

Computerbase ,with their very questionable choice for ST benchmarks (no spec, no Geekbanch; using two different versions of the same CB tool), has 13900KS ~11% faster than 7950X(3D). Expreview has it at what I think is correct value of 102.1% for 13900KS and 95.5% for 7950X - this is ~7% advantage for 13900KS. This is basically the same as what the 14900K (6Ghz max boost) should achieve.
Add ~5% on top of ~7% (1.05x1.07) and you get ~12% advantage for Arrow Lake in ST versus 7950X3D.
Gaming wise, Computerbase has 7950X3D at 9% higher average fps and 5% higher frametime values versus 13900KS. Expreview has no gaming data, but 3Dcenter has a nice average breakdown (here). 7950X3D is ~3% faster than 13900KS. If ArrowLake is only 5% faster than 14900KS in gaming, then it might just edge out slightly the 7950X3D.

As for MT performance, Computer base has 7950X at 2% faster (X3D variant is 5% slower ) vs 13900KS. Expreview has 7950X at 2% faster than 13900KS, while X3D variant is ~5% slower.
I assume that 14900K might be ~5% faster in MT than 13900KS. Add ~15% on top of those numbers : 1.05x1.15 and we get 20% faster than 13900KS and around 18% faster than 7950X.

In summary, ArrowLake vs Zen 4:
ST: ~12% faster than 7950X(3D)
MT: ~18% faster than 7950X and ~26% faster than 7950X3D
Gaming: up to few % faster than 7950X3D (if that) and 20% faster than 7950X

So vanilla Zen 5 has a really good chance to leapfrog ArrowLake if AMD manages to get at least 20% higher performance (IPC x clocks) in both ST and MT workloads, versus 7950X. Gaming should be even unless Zen 5 gets even better gaming uplift due to better front end and higher speed memory support. X3D variant of Zen 5 should be clearly ahead of any ArrowLake part, unless intel brings on some sort of stacked cache for it.

Edit: I expect Arrow Lake to have 20% higher IPC versus Raptor Lake while clocking at 5.3Ghz max. That makes it around ~5% faster than 6Ghz Raptor Lake refresh in ST.
For MT, the total gain of ~15% (from the slide) will likely be due to better E cores that have more IPC and clock higher.
 
Last edited:
  • Like
Reactions: lightmanek, Tlh97, Det0x and 2 others
A

Abwx

Lifer
Apr 2, 2011
11,837
4,790
136
inf64 said:
www.computerbase.de

CPU-Benchmark: Prozessor-Vergleich 2025

Welche CPU für Games und Apps kaufen? ComputerBase leistet Kaufberatung mit Intel Core Ultra, AMD Ryzen 9000 und 7000 auf Basis von Tests.
www.computerbase.de www.computerbase.de

Computerbase ,with their very questionable choice for ST benchmarks (no spec, no Geekbanch; using two different versions of the same CB tool), has 13900KS ~11% faster than 7950X(3D).
Click to expand...

They also use Povray wich doesnt enable AVX2 for AMD, otherwise the 7950X would score 18% better in this test, so you get the picture of the actual ST perf for FP, since all three tests are FP...
 
Last edited:
tamz_msc

tamz_msc

Diamond Member
Jan 5, 2017
3,865
3,729
136
DrMrLordX said:
Which is funny ,since that leaked slide we saw awhile back showed 8+16 Arrow Lake-S with a PL2 of 250W.
Click to expand...
Those were simulations/performance projections that IgorsLab leaked. It is very much possible that the v-f curve for whatever node - Intel or TSMC - that ARL uses gives better perf at lower iso-power than Intel 7 which really shines at higher iso-power.

My feeling is that if you could limit both RPL-refresh and ARL at 177 W PL2, ARL would be even faster than what the leaks indicate.
 
  • Like
Reactions: S'renne and Henry swagger
C

CopernicusLazarus

Banned
Oct 14, 2023
23
13
36
inf64 said:
www.computerbase.de

CPU-Benchmark: Prozessor-Vergleich 2025

Welche CPU für Games und Apps kaufen? ComputerBase leistet Kaufberatung mit Intel Core Ultra, AMD Ryzen 9000 und 7000 auf Basis von Tests.
www.computerbase.de www.computerbase.de

Computerbase ,with their very questionable choice for ST benchmarks (no spec, no Geekbanch; using two different versions of the same CB tool), has 13900KS ~11% faster than 7950X(3D). Expreview has it at what I think is correct value of 102.1% for 13900KS and 95.5% for 7950X - this is ~7% advantage for 13900KS. This is basically the same as what the 14900K (6Ghz max boost) should achieve.
Add ~5% on top of ~7% (1.05x1.07) and you get ~12% advantage for Arrow Lake in ST versus 7950X3D.
Gaming wise, Computerbase has 7950X3D at 9% higher average fps and 5% higher frametime values versus 13900KS. Expreview has no gaming data, but 3Dcenter has a nice average breakdown (here). 7950X3D is ~3% faster than 13900KS. If ArrowLake is only 5% faster than 14900KS in gaming, then it might just edge out slightly the 7950X3D.

As for MT performance, Computer base has 7950X at 2% faster (X3D variant is 5% slower ) vs 13900KS. Expreview has 7950X at 2% faster than 13900KS, while X3D variant is ~5% slower.
I assume that 14900K might be ~5% faster in MT than 13900KS. Add ~15% on top of those numbers : 1.05x1.15 and we get 20% faster than 13900KS and around 18% faster than 7950X.

In summary, ArrowLake vs Zen 4:
ST: ~12% faster than 7950X(3D)
MT: ~18% faster than 7950X and ~26% faster than 7950X3D
Gaming: up to few % faster than 7950X3D (if that) and 20% faster than 7950X

So vanilla Zen 5 has a really good chance to leapfrog ArrowLake if AMD manages to get at least 20% higher performance (IPC x clocks) in both ST and MT workloads, versus 7950X. Gaming should be even unless Zen 5 gets even better gaming uplift due to better front end and higher speed memory support. X3D variant of Zen 5 should be clearly ahead of any ArrowLake part, unless intel brings on some sort of stacked cache for it.

Edit: I expect Arrow Lake to have 20% higher IPC versus Raptor Lake while clocking at 5.3Ghz max. That makes it around ~5% faster than 6Ghz Raptor Lake refresh in ST.
For MT, the total gain of ~15% (from the slide) will likely be due to better E cores that have more IPC and clock higher.
Click to expand...
I feel like computerbase is something like the German version of userbenchmark. In multiple tests with different power limits they consistently have the 13900k on top in both MT and ST workloads in efficiency. That seems off?
 
S

SiliconFly

Golden Member
Mar 10, 2023
1,925
1,284
106
CopernicusLazarus said:
I feel like computerbase is something like the German version of userbenchmark. In multiple tests with different power limits they consistently have the 13900k on top in both MT and ST workloads in efficiency. That seems off?
Click to expand...
13900s are already on par with 7950X3D in power efficiency last year itself. 14900s should take it a bit further taking both performance & efficiency crowns.

But honestly, most aren't really excited by a vestigial product like RPL-refresh that shouldn't exist in this day & age (like rocket lake). I'm more looking forward towards MTL & ARL. RPL now in Q4 2023 is a disgrace.
 
I

igor_kavinski

Lifer
Jul 27, 2020
26,030
17,959
146
videocardz.com

MSI Z790 MAX BIOS feature increases Intel CPU throttling temperature to 115°C - VideoCardz.com

MSI solution for overheating Intel CPUs is increased throttle threshold temperature MSI is adopting a new feature for its Z790 MAX motherboards. A screenshot from the ACE MAX motherboard BIOS reveals a new addition to the OC settings, termed “CPU Fast Throttle Threshold.” This functionality is...
videocardz.com videocardz.com

Does the 14900K need the throttle limit to be raised to hit 6 GHz reliably? Me wonders...
 
Hulk

Hulk

Diamond Member
Oct 9, 1999
5,118
3,660
136
The question regarding power efficiency and the 14900K is not if it uses more power than the 13900K but how much power does it use at ISO frequencies on the same workloads. This is a hard question to answer because no one seems to do such testing. For example, try and and find a comparison between Alder Lake and Raptor Lake with the same number of P's and E's at at the same frequency.

Also, testing is complicated by the fact that there is a variation in parts due to binning (quality of silicon).

This leads to the next question regarding the Raptor Refresh, which is are these parts simply higher binned Raptor parts due to increase yields, or another refinement of Intel 7 (aka 10++++++++++++), or a combination of both?

Here's my take.

The Raptor Refresh is indeed a refinement of Intel 7, the last gasp for the hard-starting and long-lived Intel 10nm node... the one that has caused them so many problems but in the end pulled got across the finish line.

Increasing performance of parts down stack from the 13900K is pretty easy if yields/binning is better than a year ago. Increase clocks by 200MHz and add another E cluster to the 13700K.

The 14900K is a little trickier. It looks like this part will be 6GHz ST and 5.7GHz MT, with the E's at 4.4GHz. This IS marginally better than the 13900KS at what should be a markedly lower price.

Bringing a 14900KS to market will be tougher. 6.2GHz for some cherry parts is probably in the cards, but going higher than 5.7GHz all-core is going to be tough without selling the chip with a nuke reactor to power it and the associated nuke cooling to cool it.

I'll probably buy a 14900K if only to test it against the 13900K I used to have. I have a lot of data and can test the 14900K in the exact same system as the 13900K.

Finally, even though I've mainly been an Intel guy from a buying perspective over the years, there is NO comparison when it comes to Zen 4 and Raptor Lake efficiency. Zen 4 is a marvel of efficiency.
 
  • Like
Reactions: lightmanek, Tlh97, igor_kavinski and 1 other person
H

H433x0n

Golden Member
Mar 15, 2023
1,224
1,606
106
Hulk said:
The question regarding power efficiency and the 14900K is not if it uses more power than the 13900K but how much power does it use at ISO frequencies on the same workloads. This is a hard question to answer because no one seems to do such testing. For example, try and and find a comparison between Alder Lake and Raptor Lake with the same number of P's and E's at at the same frequency.

Also, testing is complicated by the fact that there is a variation in parts due to binning (quality of silicon).

This leads to the next question regarding the Raptor Refresh, which is are these parts simply higher binned Raptor parts due to increase yields, or another refinement of Intel 7 (aka 10++++++++++++), or a combination of both?

Here's my take.

The Raptor Refresh is indeed a refinement of Intel 7, the last gasp for the hard-starting and long-lived Intel 10nm node... the one that has caused them so many problems but in the end pulled got across the finish line.

Increasing performance of parts down stack from the 13900K is pretty easy if yields/binning is better than a year ago. Increase clocks by 200MHz and add another E cluster to the 13700K.

The 14900K is a little trickier. It looks like this part will be 6GHz ST and 5.7GHz MT, with the E's at 4.4GHz. This IS marginally better than the 13900KS at what should be a markedly lower price.

Bringing a 14900KS to market will be tougher. 6.2GHz for some cherry parts is probably in the cards, but going higher than 5.7GHz all-core is going to be tough without selling the chip with a nuke reactor to power it and the associated nuke cooling to cool it.

I'll probably buy a 14900K if only to test it against the 13900K I used to have. I have a lot of data and can test the 14900K in the exact same system as the 13900K.

Finally, even though I've mainly been an Intel guy from a buying perspective over the years, there is NO comparison when it comes to Zen 4 and Raptor Lake efficiency. Zen 4 is a marvel of efficiency.
Click to expand...
It's +200mhz at iso power based on the published specs.

Abwx said:
No, that s even worse for RPL because TSMC s process scale badly at higher TDP.
Click to expand...
I know, I mean when you compare both at 125W isopower it will be closer than before since it has a +200mhz base clock.
 
A

Abwx

Lifer
Apr 2, 2011
11,837
4,790
136
H433x0n said:
I know, I mean when you compare both at 125W isopower it will be closer than before since it has a +200mhz base clock.
Click to expand...

Efficency should be measured at same accomplished work wich mean at same throughput.

So far AMD has currently 60% better efficency and from the numbers we got Intel will increase efficency by 52% at best with their next process, so ARL could eventually be comparable as current AMD products on the DT side of thing but not for mobile, indeed Intel was cautious enough to state that their future mobile chip will be THEIR most efficient mobile chip ever but not THE, notice the nuance...
 
  • Haha
Reactions: CopernicusLazarus
C

CopernicusLazarus

Banned
Oct 14, 2023
23
13
36
Hulk said:
The question regarding power efficiency and the 14900K is not if it uses more power than the 13900K but how much power does it use at ISO frequencies on the same workloads. This is a hard question to answer because no one seems to do such testing. For example, try and and find a comparison between Alder Lake and Raptor Lake with the same number of P's and E's at at the same frequency.
Click to expand...
Computerbase does these tests and intel is always the most efficient cpus on their testing. Here is the most recent testing they've done, the 13900k is the most efficient cpu they ever tested at 45w

1697442210306.png
 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom