Monday, July 17th 2023

Intel "Arrow Lake-S" Desktop Processor Projected 6%-21% Faster than "Raptor Lake-S"

Intel's future-generation "Arrow Lake-S" desktop processor is already being sampled internally, and to some of the company's closest industry partners, and some of the first performance projections of the processor, comparing it with the current "Raptor Lake-S" (Core i9-13900K), have surfaced, and upcoming "Raptor Lake Refresh" desktop processor (probably the i9-14900K), have surfaced. First, while the "Raptor Lake Refresh" family sees core-count increases across the board for Core i3, Core i5, and Core i7 brand extensions, the 14th Gen Core i9 series is widely expected to be a damp squib compared to the current i9-13900 series, and it shows in the performance projection graphs, where the supposed-i9-14900K is barely 0% to 3% faster, probably on account of slightly higher clock speeds (100-300 MHz).

The "Arrow Lake-S" processor in these graphs has a core-configuration of 8P+16E. Since this is a projection, it does not reflect the final core-configuration of "Arrow Lake-S," but is a guideline on what performance increase to expect versus "Raptor Lake," assuming the same core-configuration and power limits. All said and done, "Arrow Lake-S" is projected to offer a performance increase ranging between 6% in the worst case, to 21% in the best-case benchmark, compared to the current i9-13900K, assuming an identical core-config and power-limits. The CPU benchmarks in the projection span the SPECrate2017 suite, CrossMark, SYSmark 25, WebXPRT 4, Chrome Speedometer 2.1, and Geekbench 5.4.5 ST and MT.
One area where "Arrow Lake-S" is expected to offer a performance leap is with its integrated graphics. Based on the Xe-LPG graphics architecture (DirectX 12 Ultimate capable), and armed with 8 Xe cores (128 EU), the iGPU is projected to offer a massive 240% graphics performance uplift over the current Xe-LP based iGPU of the "Raptor Lake-S" that has 32 EU.

With Intel expected to call 2023 a wrap with the "Raptor Lake Refresh" series planned for Q4-2023 on the existing LGA1700 platform, all eyes are on what Intel does in 2024. The company's subsequent desktop platform will introduce the new Socket LGA1851, and require a new motherboard. It's unclear if the platform will debut with a "Meteor Lake-S" as the microarchitecture's compute tile tops out at a 6P+16E core-count. "Arrow Lake-S" surfaced on leaked roadmap slides with a mid-2024 mass-production commencement timeline, which should put product launches some time in the second half of 2024.
Source: Igor's Lab
Add your own comment

75 Comments on Intel "Arrow Lake-S" Desktop Processor Projected 6%-21% Faster than "Raptor Lake-S"

#51
fevgatos
M440you are comparing 13th gen tuned, limited to 65 watt (where it matches 1ccd ryzen 7600,7700 in efficiency watt-to-watt, it's not better. it's still way less efficient that 2ccd 7900,7950) to a 1 ccd ryzen series stock (7700x stock where it's less efficient and gains only 5% to 7700).
I'm not comparing anything, the name amd chose for their products means they want to compare. You know, R5 vs i5, R7 vs i7?

And no, you are wrong, the 13700k at same watts as 7700 / 7800x 3d smashes in to pieces in every workload. A 13700k at 65w is faster than the 7700x at 125, lol

I mean you have a 7700x, post your score on whatever benchmark you want and we can compare :D
Posted on Reply
#52
TumbleGeorge
fevgatosYes

14 - 13 - 12

How many numbers do you see?
I'm still only seeing 12 and 13 if you're asking me about store deals. OK, now put the percentage difference in single-core performance between the generations you wrote.
Posted on Reply
#53
fevgatos
TumbleGeorgeI'm still only seeing 12 and 13 if you're asking me about store deals. OK, now put the percentage difference in single-core performance between the generations you wrote.
Well I said soon to be 3 gens old. Give it 2 - 3 months.


Can't you see it's problematic that an i7 12700 (as tested by hwunboxed) beats the latest and greatest Zen 4 R7 in efficiency?
Posted on Reply
#54
TumbleGeorge
fevgatosCan't you see it's problematic that an i7 12700 (as tested by hwunboxed) beats the latest and greatest Zen 4 R7 in efficiency?
Try to be more persuasive. Here, in this topic.
Posted on Reply
#55
SchumannFrequency
fevgatosAnd no, you are wrong, the 13700k at same watts as 7700 / 7800x 3d smashes in to pieces in every workload.
Especially consumer computers usually idle more than that they run heavy workloads.

Posted on Reply
#56
Unregistered
fevgatosI'm not comparing anything, the name amd chose for their products means they want to compare. You know, R5 vs i5, R7 vs i7?

And no, you are wrong, the 13700k at same watts as 7700 / 7800x 3d smashes in to pieces in every workload. A 13700k at 65w is faster than the 7700x at 125, lol

I mean you have a 7700x, post your score on whatever benchmark you want and we can compare :D
#57
fevgatos
M440
Thank you good sir. Let's see

The 13700k at 88w is faster than the 7700x at 142w!!! The 13600k at 125w is faster than the 7700x at 142w

And let's keep in mind, these are not all core workloads. If they were, the difference would be even bigger.

Also the 13900k at 45w is the most efficient cpu on the entire list.

That Intel inefficiency
Posted on Reply
#58
Unregistered
fevgatosI mean you have a 7700x, post your score on whatever benchmark you want and we can compare :D
Ok, challenge accepted.

What CPU at what watt will you be using?

Ill be using Ryzen 7700x@85watt
#59
fevgatos
M440Ok, challenge accepted.

What CPU at what watt will you be using?

Ill be using Ryzen 7700x@85watt
I'll match the wattage to make it fair. 12900k, which is a worse version of 13700k.
Posted on Reply
#60
Unregistered
fevgatosI'll match the wattage to make it fair. 12900k, which is a worse version of 13700k.
Could you please or somebody with 13th gen run limited to 85 watt Geekbench 6.1.0 avx2 with hwinfo screenshot.

65 watt is interesting too
#61
fevgatos
M440Could you please or somebody with 13th gen run limited to 85 watts with hwinfo screenshot, Geekbench 6.1.0 avx2.

65 watt is interesting too
Isn't it better to use benchmate since it records scores and wattage?

Funny though, you went directly to avx. You know exactly what's up, that the 7700x doesn't stand a chance in hell in normal workloads. Rofl, I like you man, that was sweet. But if you already know, why don't you straight up admit it?

You should use ycruncher with avx instead of Geekbench though, I think it does much better on zen 4
Posted on Reply
#62
TumbleGeorge
Hmmm, in which topic do you measure your results? There are suitable ones. Or do you do it in a private conversation?
Posted on Reply
#63
fevgatos
TumbleGeorgeHmmm, in which topic do you measure your results? There are suitable ones. Or do you do it in a private conversation?
Im still waiting for him to post his results :P
Posted on Reply
#64
Unregistered
fevgatosIsn't it better to use benchmate since it records scores and wattage?

Funny though, you went directly to avx. You know exactly what's up, that the 7700x doesn't stand a chance in hell in normal workloads. Rofl, I like you man, that was sweet. But if you already know, why don't you straight up admit it?

You should use ycruncher with avx instead of Geekbench though, I think it does much better on zen 4
You might be confusing avx2 and avx512

This will test mixed single core and mult in somewhat real daily tasks. Interesting if severe wattage limitation have influence on that. I can not find such results on the web. Every other test you have up there

My results are in the system specs
#65
fevgatos
M440My results are in the system specs
That's at 85w ppt?
Posted on Reply
#67
fevgatos
M440yes
I can match the MT at 70w, ill post results soon with power draw and all, both at 85w and unlimited
Posted on Reply
#68
chrcoluk
State of this thread. :(

For reference I dropped 55w on power limit vs what my chip runs at without restraint and it lowered my cinebench productivity score by a whopping 3%, so any reviews commenting on performance per watt that should be taken into account for a K processor and Z board combo which has fully unlocked voltage and power controls. I know my ASRock board e.g. the default isnt a 241/241 or 253/253 and definitely isnt a 4096/4096. It was 241/125. Which isnt far off TPU's recommended setting when they did PL testing on a 12900k.

Also at the time I made my purchase the Intel platform was considerably cheaper in capital outlay required due to not needing DDR5 and cheaper boards. So there is many variables here, but this doesnt need to be some kind of shots taken at each other match, I think both CPU's and platforms are great. Just with different pros and cons to each other.
Posted on Reply
#69
fevgatos
chrcolukI think both CPU's and platforms are great. Just with different pros and cons to each other.
ΥES, but efficiency unlike popular belief isn't one of AMD's pros. For MT that is, for games the 3d chips are good
M440yes
Here is mine, 85w power limit, but I get the same score at any wattage honestly. Hwinfo peaked at 81w but it averaged at 12w, it doesn't really pull that much power, usually it stayed below 25w during the whole run. Can we test a properly threaded benchmark that actually maxes out the power limit?

browser.geekbench.com/v6/cpu/1970278
Posted on Reply
#70
Unregistered
fevgatosΥES, but efficiency unlike popular belief isn't one of AMD's pros. For MT that is, for games the 3d chips are good


Here is mine, 85w power limit, but I get the same score at any wattage honestly. Hwinfo peaked at 81w but it averaged at 12w, it doesn't really pull that much power, usually it stayed below 25w during the whole run. Can we test a properly threaded benchmark that actually maxes out the power limit?

browser.geekbench.com/v6/cpu/1970278
It's ok, it should be a perfectly usable and responsive system in daily tasks being limited to 85watt.

Still - matching the 8 core ryzen does not mean its even close to 2ccd products or any 8p+8e lga1700 part was ever priced on par with 8 core am5.

www-computerbase-de.translate.goog/2022-10/intel-core-i9-13900k-i7-13700-i5-13600k-test/2/?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=pl&_x_tr_pto=wapp

Posted on Edit | Reply
#71
fevgatos
M440Still - matching the 8 core ryzen does not mean its even close to 2ccd products or any 8p+8e lga1700 part was ever priced on par with 8 core am5.
Well, the 7700x did launch at similar price to the 13700k. The 7800x 3d DEFINITELY launched at same / higher price than the 13700k. And yes, the 2-ccds Ryzens are indeed more efficient than their Intel counterparts - but at same wattages the difference isn't that big. In fact it's usually around 10% with some exceptions that AMD does really well like Vray. You limit let's say a 13700k and a 7900x to 100w,you won't notice a performance difference.

People act like zen 4 is so far ahead which is not true at all, especially when talking about the single ccd zen 4 that are behind
Posted on Reply
#72
persondb
john_The iGPU performance gains might force AMD to start integrating Infinity Cache in it's iGPUs.
That was compared to a 13900K which is only a 32EU SKU, so 2.5x the timespy score wouldn't be much actually.
Posted on Reply
#73
RandallFlagg
Honestly not as impressive as I thought it would be, given this is supposed to be both a tick and a tock, the first consumer CPU on the new Intel 4 node. I was expecting something similar to the bump Alder Lake had vs 10th gen and Rocket Lake with all the talk about tiles and new arch new nodes. Was kind of hoping this would bring in enough of a bump I could replace my rig with a big laptop next year and not lose performance, but doesn't look that way.

Not sure I believe this leak either though. Benchmarks can be iffy enough with many variables without throwing in words like 'projected' performance.
Posted on Reply
#74
ToxicTaZ
I think I would be happy with My new upcoming Intel 14900KS for a while, I'm upgrading from Intel 9900KS.....I used to buy into first gen CPUs and do a second upgrade on the same board....not no more.... I'll wait until Intel 18A (5nm++) would be on new generation Intel Socket V2 (LGA 2551) PCIe 6.0 Ready. Beast Lake?.... Wait and see..

Cheers
Posted on Reply
#75
TheoneandonlyMrK
I'm only reading the Headline on this one, and that's some beautiful bullshit blurb 6/21% faster , great , narrowed that right down there.

They might as well have said.

"It's definitely faster"
Posted on Reply
Add your own comment
Oct 31st, 2024 19:55 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts