Question Raptor Lake - Official Thread

[Hulk]

Since we already have the first Raptor Lake leak I'm thinking it should have it's own thread.
What do we know so far?
From Anandtech's Intel Process Roadmap articles from July:

Built on Intel 7 with upgraded FinFET
10-15% PPW (performance-per-watt)
Last non-tiled consumer CPU as Meteor Lake will be tiled

I'm guessing this will be a minor update to ADL with just a few microarchitecture changes to the cores. The larger change will be the new process refinement allowing 8+16 at the top of the stack.

Will it work with current z690 motherboards? If yes then that could be a major selling point for people to move to ADL rather than wait.

 

[Exist50]

Not sure if it's already been posted, but looks like there's a bug in XTU that explains HUB's anomalous power scaling results.

https://www.reddit.com/r/hardware/comments/y9c8yj

 

[controlflow]

I'm not sure why some people are struggling to accept the results from TechTesters.
TechPowerUp shows a pretty clear lead for 13th gen in gaming too and that was on a 3080 (though at 720p).

 

[controlflow]

Not sure if it's already been posted, but looks like there's a bug in XTU that explains HUB's anomalous power scaling results.

https://www.reddit.com/r/hardware/comments/y9c8yj

Was just about to post this.

They are going to need to redo that video IMO. That performance/power curve is completely messed up. They differ from the other reviewers by 30% at some of the lower power limits. Massively understated the performance/watt here.

 

[reb0rn]

There goes your credibility. Back atcha buddy.

Do you not understand GPU limited scenario, I know you do, you just want to push confusion here... there is zero reason to test CPU in gaming test with limited games by GPU, most reviews done it, and when you are show few that took time to do proper test some of you try to discredit it as none other used RTX 4090 as they all should have

I'm not sure why some people are struggling to accept the results from TechTesters.
TechPowerUp shows a pretty clear lead for 13th gen in gaming too and that was on a 3080 (though at 720p).

View attachment 69568

You know why, it`s hard for most to see they are wrong in theirs preaching for months
But just imagine if they used RTX 4090 lead would be defo up to 20% which is PURE DOMINATION in gaming as KING for intel where even 7x 3D version will not fix, and they know it

 

[reb0rn]

del

 

[Markfw]

Do you not understand GPU limited scenario, I know you do, you just want to push confusion here... there is zero reason to test CPU in gaming test with limited games by GPU, most reviews done it, and when you are show few that took time to do proper test some of you try to discredit it as none other used RTX 4090 as they all should have


You know why, it`s hard for most to see they are wrong in theirs preaching for months
But just imagine if they used RTX 4090 lead would be defo up to 20% which is PURE DOMINATION in gaming as KING for intel where even 7x 3D version will not fix, and they know it

So, you say we won't accept ONE sites results ? Here is is what Anandtech had to say :

"The first thing to note in our 1440p testing is that in Civ VI, and throughout, we've seen dominance from AMD's Zen 4 core here. I've retested numerous times to confirm, and they are correct. It's also worth noting that again, the Ryzen 7 5800X3D performs well in some of the titles, especially Red Dead Redemption 2, Grand Theft Auto V, and in Borderlands 3. If a title can utilize all of that 3D V-Cache, then the 5800X3D excels, even against the latest and great Zen 4 and Raptor Lake chips. "

 

[Hitman928]

If you listened the review, you would see it was said for Spiderman there could be a bug as the game is new and really hammer CPU... but you have how many 23 games?????? (there could be some settings shifting some ray trace load to CPU or GPU issue or not as its new game and i guess lots bugs), but it can also mean game is optimized for intel or that bottleneck bag just work in it

but yes try to paint issue with one buggy game to discredit it as none of eview sites had time or wanted to do test with RTX 4090 which only point of what pile of %%#%$## they are

She says that she suspects that there might be a bug in the game optimization but then continues to use the results as if nothing is wrong and does not mention at all that the same CPU was showing a much higher fps in their last benchmark. Additionally, Spiderman was not the only game that shows this behavior, it was just the most egregious. Cyberpunk also mysteriously decreases in performance on the 7900x while the 12900k increases in performance with the faster GPU. Same story with Far Cry 6. Then there are a couple of other cases where the 7900x was originally a little ahead of the 12900k but then in the new benchmarks it is behind. It's just hard to trust someone's results when they seem to be inconsistent with their own previous testing and there's no acknowledgement of this at all.

 

[reb0rn]

So, you say we won't accept ONE sites results ? Here is is what Anandtech had to say :

"The first thing to note in our 1440p testing is that in Civ VI, and throughout, we've seen dominance from AMD's Zen 4 core here. I've retested numerous times to confirm, and they are correct. It's also worth noting that again, the Ryzen 7 5800X3D performs well in some of the titles, especially Red Dead Redemption 2, Grand Theft Auto V, and in Borderlands 3. If a title can utilize all of that 3D V-Cache, then the 5800X3D excels, even against the latest and great Zen 4 and Raptor Lake chips. "

I do not care who say it, if someone test CPU gaming test for new CPU in 1440p and in the same time do not use fastest GPU available (the RTX4090 as at least 50% faster than GPU they used) so games would not be GPU limied.....GOOD BYE..... I give zero respect for that part of review I mean the games test as its worthless!

 

[Exist50]

So, you say we won't accept ONE sites results ? Here is is what Anandtech had to say :

"The first thing to note in our 1440p testing is that in Civ VI, and throughout, we've seen dominance from AMD's Zen 4 core here. I've retested numerous times to confirm, and they are correct. It's also worth noting that again, the Ryzen 7 5800X3D performs well in some of the titles, especially Red Dead Redemption 2, Grand Theft Auto V, and in Borderlands 3. If a title can utilize all of that 3D V-Cache, then the 5800X3D excels, even against the latest and great Zen 4 and Raptor Lake chips. "

Of all things to quote, FPS (instead of turn time) in Civ certainly seems like an odd metric. Anandtech is sorely in need of updating their benchmarking methodology, but I fear that's a lost cause at this point. But they did test more than just Civ...

 

[reb0rn]

She says that she suspects that there might be a bug in the game optimization but then continues to use the results as if nothing is wrong and does not mention at all that the same CPU was showing a much higher fps in their last benchmark. Additionally, Spiderman was not the only game that shows this behavior, it was just the most egregious. Cyberpunk also mysteriously decreases in performance on the 7900x while the 12900k increases in performance with the faster GPU. Same story with Far Cry 6. Then there are a couple of other cases where the 7900x was originally a little ahead of the 12900k but then in the new benchmarks it is behind. It's just hard to trust someone's results when they seem to be inconsistent with their own previous testing and there's no acknowledgement of this at all.

Again you have better test with PROPER GPU as rtx4090 to test CPU gaming that is not limited by GPU????? you have up 2 more sources that used even old 3080 with 10% intel lead in 720P and test that tested just few games due time also shw same
But still you want to FUD here for test of 23 games for one test on crappy game that has bags, LOL

 

[Hitman928]

I'm not sure why some people are struggling to accept the results from TechTesters.
TechPowerUp shows a pretty clear lead for 13th gen in gaming too and that was on a 3080 (though at 720p).

View attachment 69568

TPU's results are fine, they show a little bit larger of a lead than HWU on average, it just comes down to game choices when you're only testing around a dozen games. Hardware Unboxed has a couple of "AMD friendly" and "Intel friendly" games in their lineup whereas TPU's games has a couple "Intel friendly" games but not really any ones for AMD. For sure, RPL is the gaming "king" right now (as I predicted it would be after the Zen 4 reviews) but it will probably be a bit short lived if the 3d cache version(s) of Zen 4 are releasing early next year as rumored.

 

[reb0rn]

15-20% is a lot to beat , and saying it will be is PURE huge speculation
and when you have cheap 13600K that is fast today and AMD x3d might be same or not why wait

for years kid talk and mostly buy only ryzen for gaming as there was huge FUD its so better then intel, well not anymore FUD stop now

 

[Hitman928]

Again you have better test with PROPER GPU as rtx4090 to test CPU gaming that is not limited by GPU????? you have up 2 more sources that used even old 3080 with 10% intel lead in 720P and test that tested just few games due time also shw same
But still you want to FUD here for test of 23 games for one test on crappy game that has bags, LOL

Again, it wasn't just 1 "crappy game", I gave multiple examples where their results were not consistent with their own testing. Additionally, Tech Testers didn't benchmark at 720p, only 1080p, which is over twice as many pixels as 720p. The 4080 is significantly slower than 2x of a 3080, so testing a 3080 and 720p, at worst, should be just as much of a CPU bottleneck as testing a 4080 at 1080p.

 

[controlflow]

TPU's results are fine, they show a little bit larger of a lead than HWU on average, it just comes down to game choices when you're only testing around a dozen games. Hardware Unboxed has a couple of "AMD friendly" and "Intel friendly" games in their lineup whereas TPU's games has a couple "Intel friendly" games but not really any ones for AMD. For sure, RPL is the gaming "king" right now (as I predicted it would be after the Zen 4 reviews) but it will probably be a bit short lived if the 3d cache version(s) of Zen 4 are releasing early next year as rumored.

HardwareUnboxed set powerlimits on their 13rd gen CPUs using XTU. Allegedly there is a bug that causes the CPU frequency to drop below where it should be when XTU is used as opposed to the BIOS. Even when using a PL1/PL2 of 4000+ watts the 13900k from HWU was outperformed by 13900ks from other sites running at 253w limits set in the BIOS. I will wait until HWU revisits their data before considering their review. They acknowledged this issue on Twitter themselves.

 

[Hitman928]

15-20% is a lot to beat , and saying it will be is PURE huge speculation
and when you have cheap 13600K that is fast today and AMD x3d might be same or not why wait

First, even before RPL launched I predicted it would take the gaming crown and have acknowledged as much in this thread. Second, expecting 3d cache to overtake RPL in gaming is speculation which is why I said, "probably". You are free to disagree and of course we should wait for actual reviews to see its performance. Third, I really don't care who wins what, I'm just trying to have an honest conversation with reliable data. Unfortunately, there's tons of bad data out there today, so if I see something that doesn't add up, I'm going to call it out. If you have a problem with that, I guess you can put me on ignore.

 

[Hitman928]

HardwareUnboxed set powerlimits on their 13rd gen CPUs using XTU. Allegedly there is a bug that causes the CPU frequency to drop below where it should be when XTU is used as opposed to the BIOS. Even when using a PL1/PL2 of 4000+ watts the 13900k from HWU was outperformed by 13900ks from other sites running at 253w limits set in the BIOS. I will wait until HWU revisits their data before considering their review. They acknowledged this issue on Twitter themselves.

Yes, it makes their power scaling graph invalid which clears up a lot of confusion between what other sites were reporting. However, for their standard tests, as far as I am aware, they didn't use XTU and just ran the system with the out of the box settings so everything except the power scaling tests should still be valid.

 

[Markfw]

Of all things to quote, FPS (instead of turn time) in Civ certainly seems like an odd metric. Anandtech is sorely in need of updating their benchmarking methodology, but I fear that's a lost cause at this point. But they did test more than just Civ...

I was quoting the site, not that specific benchmark. All I see in general, is that there is no clear gaming winner here, and reborn seems to think they win everything by "20-30%" which is just WRONG.

 

[reb0rn]

@Hitman928
probably sure but even if it happen it will be 2% faster than 13600k, and we can expect AMD can make cheap 3D chp again, with the fabs they pay for production not as intel that can print 100000% more i5 CPU then AMD print X3D

maybe for show of just to please their fan base so make it cheap, with negative profit so they can still FUD games ryzen is best CPU for it and mostly it was not

 

[pakotlar]

Your submission was successful.

internet points earned

 

[OneEng2]

My thoughts on this are as follows:

• While PC desktop gaming benchmarks are interesting, and the 13900K certainly competes well, it seems that within a quarter, the X3D version of the Zen4 processors would be a better bet
• Intel's pricing is very competitive to AMD's right now; however, I question the supply of Intel's new chips compared to AMD's new Zen4 lineup. Actual pricing in the industry may be quite different. We will soon see. At any rate, Intel with its monolithic die is most certainly going to have more trouble with yields compared to AMD's chiplet architecture. Again, time will tell. I would say that AMD will be lowering pricing soon considering the overall system cost of the AM5 platform.
• While the power draw of the new Intel processors is worth some raised eyebrows in the PC desktop, the power scaling of the new Intel architecture could be a real problem for Intel in the much more prolific laptop market where the majority of chips will be sold anyway. AMD's Zen4 architecture appears to scale quite well with power compared to the Intel 13th gen chips (twice as well in fact). This could be a real serious problem for Intel overall.

Additionally, I would like to point out to those that talk about Intel having a processor on 10nm that is competing with AMD on a 5nm node, that isn't quite the whole story. Intel's transistor density on Intel 7 is around 100mil/mm^2. TSMC's 5nm process gives a transistor density of around 173mil/mm^2. It is true that AMD has transistor budget on their side; however, it is clear that Intel has created a quite large die on Intel 7 while AMD's compute die is still relatively small on TSMC 5nm. My point is that Intel has sacrificed die size on a premium node and power efficiency in order to meet (and exceed) the performance of AMD's Zen 4. Can Intel make money with this philosophy? Certainly their profit margins are going to be cut severely compared to the past.

Time will tell for sure. I am very curious to see how the mobile offerings come out. This is where the big money is. Additionally, AMD's advantage with their chiplet design is eliminated with their mobile processors as they also use a monolithic die. Having said this, imagine how power efficient Zen 4 would be if all the logic was on 5nm?

 

[Panino Manino]

Anand's review omitted HEVC in the encoding benchmarks.
Any reason?

 

[gdansk]

I expected Raptor Lake to win in gaming. Especially in bang for the buck. But AMD will have vcache to compete there.
What I didn't expect is that because of AMD's unwillingness to add cores, they also lost in some MT benchmarks. E-core spam works

 

[Heartbreaker]

GN review of the 13600K. Looking very good.

 

[Carfax83]

So, you say we won't accept ONE sites results ? Here is is what Anandtech had to say :

"The first thing to note in our 1440p testing is that in Civ VI, and throughout, we've seen dominance from AMD's Zen 4 core here. I've retested numerous times to confirm, and they are correct. It's also worth noting that again, the Ryzen 7 5800X3D performs well in some of the titles, especially Red Dead Redemption 2, Grand Theft Auto V, and in Borderlands 3. If a title can utilize all of that 3D V-Cache, then the 5800X3D excels, even against the latest and great Zen 4 and Raptor Lake chips. "

Anandtech's testing quality has been abysmal in gaming for a long time now and nothing has changed in the 13th gen review.

They used a 6950xt for the GPU, which is almost like a joke. It's no wonder their gaming benches are full of anomalies because they are GPU bottlenecked to hell.

If any 13th gen review didn't use an RTX 4090 for the gaming benchmarks, it should be discounted.

 

[alexruiz]

Anandtech's testing quality has been abysmal in gaming for a long time now and nothing has changed in the 13th gen review.

They used a 6950xt for the GPU, which is almost like a joke. It's no wonder their gaming benches are full of anomalies because they are GPU bottlenecked to hell.

If any 13th gen review didn't use an RTX 4090 for the gaming benchmarks, it should be discounted.

That "joke" of GPU was the second fastest available video card just a few days ago.
It is still the 3rd fastest video card on the market.

The reviewers who have a clue on how to setup a Ryzen system got very close results between the platforms, even when using the RTX 4090
Those who are getting major discrepancies are as competent as you on setting up a Ryzen system...