Question Raptor Lake - Official Thread

Hulk · Dec 5, 2021

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.

DrMrLordX · Oct 17, 2023

dullard said:
Meteor Lake is where things start to get interesting.

Not on desktop, it isn't.

Thunder 57 said:
Give us Arrow Lake Intel. Although rumors on that look pretty bleak as well.

That's a year from now, if not (slightly) longer (Dec 2024 anyone?). Also that's probably going to be more of an efficiency jump than anything else over Raptor Lake. Lower PL2 values, but not much more performance. +5% ST, +15-20% MT? Maybe? That would be fine today, but next year? It will have new competition.

blckgrffn · Oct 17, 2023

I think those that invested in 12700K/12900K should feel just about 0 FOMO. They've likely had (essentially) the best Intel could offer for years now.

That's my optimists take

I built several 12400/12600K/12700K rigs and they were all "fine". I am the kind of guy who manually sets PL1 & PL2 limits even.

I am slightly shocked at how little this "generation" brings to the table, but so long as it brings prices down at Microcenter that's a win in my book!

nicalandia · Oct 17, 2023

inf64 said:
I can post these as well

14900K has:
~6% higher MT performance vs 7950X3D with an "amazing" 2.5x (!!!) higher power draw.
~2% lower MT performance vs 7950X with slightly less "amazing" 1.6-1.8x higher power draw.

It's a hot piece of turd, nothing more, nothing less. The best chip out of the bunch is 14700K,

Well Said. AMD 7800X3D Still King of Gaming and 5800X3D Is such a Chad of a CPU.

dullard · Oct 17, 2023

DrMrLordX said:
Not on desktop, it isn't.

Depends on your definition of desktop. I think Meteor Lake as HTPC or SFF will do quite well. The company that I work for only buys SFF desktop computers now (not that I have any say in it). My HTPC is less than 2 years old, so I'm not really in the market for one now. But what the rumors are saying would make Meteor lake into a great HTPC.

DrMrLordX said:
Also that's probably going to be more of an efficiency jump than anything else over Raptor Lake. Lower PL2 values, but not much more performance. +5% ST, +15-20% MT? Maybe?

And if you look at all the posts the last two days they are almost exclusively talking about poor efficiency. You make it sound like a terrible thing if Intel will fix the heat and power issues.

Hulk · Oct 17, 2023

jpiniero said:
The base clock is just a "guarantee" that it will be able to sustain a clock. On the i9, it's 3.2 Ghz for the P cores so anything above that really isn't a lie.

True. It means there is wild variability in these parts silicon lottery-wise. I'm wondering if I should get the 14900K or the 13900KS as they are both the same price and the KS is guaranteed high quality silicon.

igor_kavinski · Oct 17, 2023

Hulk said:
I'm wondering if I should get the 14900K or the 13900KS as they are both the same price and the KS is guaranteed high quality silicon.

The AI OC feature of the new XTU works with 14900K only, AFAIK.

EDIT: Also, I think only 14th gen allows per core thermal limits to be set in BIOS. So you could set your best core to boost till 118C, for example.

igor_kavinski · Oct 17, 2023

MSI preparing "Beyond 6GHz" BIOS settings up to 6.3GHz for upcoming Intel 14th Gen Core CPUs - VideoCardz.com

MSI Z790 MEG ACE MAX up to 6.3 GHz with Core i9-14900K? A screenshot of MSI’s newest Z790 ACE MAX motherboard reveals the company’s plans for easy overclocking options for new CPUs. The MAX series refresh of the Z790 motherboard lineup introduces various changes to design, functionality, and...

videocardz.com

My guess is that Intel will AGAIN partner with Cooler Master for a peltier chiller AIO that will let the 14900K work at 6.3 GHz ST/5.8 GHz all-core clocks.

igor_kavinski · Oct 17, 2023

G.Skill Shows Off DDR5-8600 RAM for Raptor Lake Refresh CPUs

54% higher than the official max supported frequency on Raptor Lake Refresh

www.tomshardware.com

One more reason in favor of 14900K.

I'm very sorry but I'm in full "Intel Shiny Turd" marketing mode

Abwx · Oct 17, 2023

H433x0n said:
That's wild. This would suggest that the base clocks are just a straight up lie / false.

Base clocks are for PL1 = 125W.

Other thing is that Computerbase measured single core power in CB 2024 peaking at 102W before getting back to 55W, that s a hell of peak power for just one core.

Intel Core i9-14900K, i7-14700K & i5-14600K im Test

Zum Abschluss einer Ära gibt es noch einmal ein „Refresh“: Intel Core i9-14900K, Core i7-14700K & Core i5-14600K im ausführlichen Test.

www.computerbase.de

H433x0n · Oct 17, 2023

Abwx said:
Base clocks are for PL1 = 125W.

Other thing is that Computerbase measured single core power in CB 2024 peaking at 102W before getting back to 55W, that s a hell of peak power for just one core.

Intel Core i9-14900K, i7-14700K & i5-14600K im Test

Zum Abschluss einer Ära gibt es noch einmal ein „Refresh“: Intel Core i9-14900K, Core i7-14700K & Core i5-14600K im ausführlichen Test.

www.computerbase.de

I don’t think that spike means much. It’s probably uncore spooling up in anticipation of a big multicore load that never takes place.

As far as base clocks, it’s supposed to mean minimum frequency at maximum workload at a set power. You would assume a +200mhz increase in base clocks would translate to something but I suppose not.

Hulk · Oct 17, 2023

H433x0n said:
As far as base clocks, it’s supposed to mean minimum frequency at maximum workload at a set power. You would assume a +200mhz increase in base clocks would translate to something but I suppose not.

I've been pondering the same thing.

People over at overclock.net are going nuts trying to figure out if the 14900K is actually a better bet than the 13900KS. Posting SP ratings, VID's, and other data relating to Vcore under load, power, frequency...

Lots of fun theories going around that thread. "SP ratings don't mean anything" "SP ratings mean everything" "All of the good SP rated parts come at launch" And on and on!

Overclocking "Raptor Lake refresh" -...

www.overclock.net

H433x0n · Oct 17, 2023

Hulk said:
I've been pondering the same thing.

People over at overclock.net are going nuts trying to figure out if the 14900K is actually a better bet than the 13900KS. Posting SP ratings, VID's, and other data relating to Vcore under load, power, frequency...

Lots of fun theories going around that thread. "SP ratings don't mean anything" "SP ratings mean everything" "All of the good SP rated parts come at launch" And on and on!

Overclocking "Raptor Lake refresh" -...

www.overclock.net

It seems like for 14900K, all we got was a test of everybody’s coolers. The only person that got results that matched my expectations was Jayz2Cents but he’s pretty unreliable so I don’t want to read too much into it. The only thing that remains constant from all reviewers is that when limited to 253W RPL-R will run 8-10* C cooler.

For the 14600K there’s a clear trend where it has both higher frequency and less power consumption. I don’t think I’ve seen a review where the 14600K consumes more power than a 13600K.

For 14700K when compared to 13700K it also appears to have higher P core frequency when both are running at 253W. Even ComputerBase’s numbers which IMO are trash tier for this release show a slight frequency bump for 13700K at isopower.

Abwx · Oct 17, 2023

H433x0n said:
I don’t think that spike means much. It’s probably uncore spooling up in anticipation of a big multicore load that never takes place.

As far as base clocks, it’s supposed to mean minimum frequency at maximum workload at a set power. You would assume a +200mhz increase in base clocks would translate to something but I suppose not.

Uncore does not use that much power, 45W here, to feed datas to a single core, not even to feed all cores, that seem to be the necessary power to reach 6GHz on ST because on the CB 2024 ST test the core hoover at 5.5-5.6GHz and 55W.

H433x0n · Oct 17, 2023

Abwx said:
Uncore does not use that much power, 45W here, to feed datas to a single core, not even to feed all cores, that seem to be the necessary power to reach 6GHz on ST because on the CB 2024 ST test the core hoover at 5.5-5.6GHz and 55W.

What? The data shows 55W at 6.0ghz.

Hitman928 · Oct 17, 2023

H433x0n said:
It seems like for 14900K, all we got was a test of everybody’s coolers. The only person that got results that matched my expectations was Jayz2Cents but he’s pretty unreliable so I don’t want to read too much into it. The only thing that remains constant from all reviewers is that when limited to 253W RPL-R will run 8-10* C cooler.

For the 14600K there’s a clear trend where it has both higher frequency and less power consumption. I don’t think I’ve seen a review where the 14600K consumes more power than a 13600K.

For 14700K when compared to 13700K it also appears to have higher P core frequency when both are running at 253W. Even ComputerBase’s numbers which IMO are trash tier for this release show a slight frequency bump for 13700K at isopower.

Computerbase has the 14600k using more power but TPU has it using less, so, *shrugs*.

Hulk · Oct 17, 2023

H433x0n said:
It seems like for 14900K, all we got was a test of everybody’s coolers. The only person that got results that matched my expectations was Jayz2Cents but he’s pretty unreliable so I don’t want to read too much into it. The only thing that remains constant from all reviewers is that when limited to 253W RPL-R will run 8-10* C cooler.

For the 14600K there’s a clear trend where it has both higher frequency and less power consumption. I don’t think I’ve seen a review where the 14600K consumes more power than a 13600K.

For 14700K when compared to 13700K it also appears to have higher P core frequency when both are running at 253W. Even ComputerBase’s numbers which IMO are trash tier for this release show a slight frequency bump for 13700K at isopower.

Nice summary. Thanks.

If the 14900K is running cooler than the 13900K and both are always thermally limited then the 14900K should be showing consistently better performance as it would be running at a higher frequency, right?

Abwx · Oct 17, 2023

H433x0n said:
What? The data shows 55W at 6.0ghz.

You are right, it was for the 14700K, but still, the 14900K hoover at 5.7-6.0 and the peak is at 117W, so the average frequency is 5.85GHz for 55W average power.

igor_kavinski · Oct 17, 2023

From Techspot: https://www.techspot.com/review/2749-intel-core-14th-gen-cpus/

Look at the power consumption of 14600K vs. 13600K and 14700K vs. 13700K.

3.48% and 3.78% less power consumed by the newer P-cores.

I think this is evidence that 14th gen is better binned silicon.

As for 14900K consuming more, maybe they got a bad sample or TVB is messing up the power numbers or since they let the mobo manage the voltage, the 14900K is getting overvolted?

H433x0n · Oct 17, 2023

igor_kavinski said:
From Techspot: https://www.techspot.com/review/2749-intel-core-14th-gen-cpus/

View attachment 87337
Look at the power consumption of 14600K vs. 13600K and 14700K vs. 13700K.

3.48% and 3.78% less power consumed by the newer P-cores.

I think this is evidence that 14th gen is better binned silicon.

As for 14900K consuming more, maybe they got a bad sample or TVB is messing up the power numbers or since they let the mobo manage the voltage, the 14900K is getting overvolted?

It’s same p cores. I’ve got no idea why the performance isn’t the same across both RPL & RPL-R.

Hulk said:
Nice summary. Thanks.

If the 14900K is running cooler than the 13900K and both are always thermally limited then the 14900K should be showing consistently better performance as it would be running at a higher frequency, right?

You’d think so. I honestly can’t make sense of any of it.

GodisanAtheist · Oct 18, 2023

It's so weird to see the same Intel that brought us Sandy Bridge and Skylake, the Intel with it's boot on AMD's neck through all the Bulldozer years, struggling so badly now.

Granted they're not getting their stuff pushed in like AMD was, but having to basically double the power consumption of your competitor to still lose or at best draw even is... not a great look.

gdansk · Oct 18, 2023

Would the disappointed people have been happier if it was called e.g. 13950K?
Or is it a product that shouldn't exist which such a low performance difference?

coercitiv · Oct 18, 2023

gdansk said:
Would the disappointed people have been happier if it was called e.g. 13950K?
Or is it a product that shouldn't exist which such a low performance difference?

The name doesn't matter. It's a meh refresh but still an overall better offering. The people in the $350+ market are also getting a decent MT boost. If I had any use for a bucket of more cores I would definitely upgrade from 12700K to 14700K.

The "not so great news" from the 14th gen in my view is the 14600K, no movement there in terms of value. But I'm not going to complain because 13600K was a big jump already (the champ of 13th gen), had they put 8 P cores in the 14600K it would have killed 14700+ sales completely. I certainly would have immediately bought a 14600K with 8+8 config though.

JoeRambo · Oct 18, 2023

igor_kavinski said:
Look at the power consumption of 14600K vs. 13600K and 14700K vs. 13700K.

3.48% and 3.78% less power consumed by the newer P-cores.

What i hate about tests like these is reviewers that have their cranial cavity empty. What does that test have to do with CPU power consumption?
Did it occur to those people that 14900K will have HIGHER fps than 14600K (due to more L3) and to generate those FPS, 4090 needs to work harder. The power increase for 4090 is not linear either.
They even have a graph there where 14900K has like 10% more FPS, and it is safe to say when talking about 400W system consumption that 10% increase will have large impact.

For all we know, 14900K might have consumed least and got swamped by 4090.

psolord · Oct 18, 2023

I am thinking getting a 14600kf to upgrade the 12400f, which in turn will go on a secondary system. However I have a DDR4 motherboard, which is quite good really. Will I be losing much of the potential upgrade, with DDR4-4400 ram?

Also do I really care if I don't get the 100%? I mean all I want, is stable RT performance at 60fps, which the 12400f cannot really do. For raster it's great, for RT not so much.

igor_kavinski · Oct 18, 2023

psolord said:
Will I be losing much of the potential upgrade, with DDR4-4400 ram?

Assuming you are able to hit DDR4-4400 with your CPU and mobo, it might be a better option than DDR5 due to DDR4's lower latency: https://www.overclock.net/threads/d...ance-with-13900k-and-rtx-4090.1802973/page-10

Question Raptor Lake - Official Thread

Diamond Member

Lifer

Diamond Member

Diamond Member

Elite Member

Diamond Member

Lifer

Lifer

Lifer

Lifer

Senior member

Diamond Member

Senior member

Lifer

Senior member

Diamond Member

Diamond Member

Lifer

Lifer

Senior member

Diamond Member

Platinum Member

Diamond Member

Golden Member

Golden Member

Lifer