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.

 

[AlltheWay LeeWhy]

I don't think Intel can sway me to Raptor Lake with Zen 4 V-cache edition supposed to be releasing by end of this year. But benchmarks first, decisions later.

Isn't Raptor Lake the very last monolithic CPU from Intel for desktop? Meteor Lake begins the chiplet gen. Doesn't using chiplets like AMD has been doing for some time now increase memory latency times?

I remember dropping into the mid 30s in nanoseconds with Coffee Lake, now it's the upper 40s in nanoseconds with Rocket Lake. AMD's mem latencies are way up in the 70ns and 80ns range.

Of course we cannot be certain, but it's a pretty good bet Intel's IMCs will be binned better and DDR5 latencies lower vs AMD for Raptor vs Raphael.

For implementation of DDR5 I'd lean towards Raptor Lake. But I keep telling myself the very moment ROG creates a 2-dimm motherboard for AMD, I'll give it a good hard look.

This kit below is simply DDR4 a b-die kit before I discovered higher mem speeds with Hynix. Same 11600K with the "simply stellar" IMC.

 

[igor_kavinski]

I don't have personal experience with i5-12600K but I read that it can do max DDR4-4100 at Gear 1 ratio. Not sure if RPL will do better. It may not since Intel wants to focus more on DDR5.

 

[Exist50]

A motherboard that follows Intel's guidelines should provide the correct power limits for the CPU though, right? So a 12900 non k should get the stock power limits on a Z motherboard that is following Intel's guidelines? Edit: I guess I am just curious if the same motherboard manufacturers pushing power in Z390 era are doing same things now.

Sure, if they follow the guidelines then it will be more tame. Intel's whole PL2 = PL1 think with Alder Lake K series is basically just giving up and acknowledging that that's something many motherboards we're doing anyway, so they might as well legitimize it. Not sure what AMD's policy is (power limits enforced on board manufacturers or not), but conceptually they could do similar.

 

[igor_kavinski]

But I keep telling myself the very moment ROG creates a 2-dimm motherboard for AMD, I'll give it a good hard look.

They just might. AMD keeps hinting that Zen 4 will have excellent memory overclocking features.

 

[AlltheWay LeeWhy]

They just might. AMD keeps hinting that Zen 4 will have excellent memory overclocking features.

Such an exciting time to be an enthusiast. Processors from both camps should be outstanding.

Wanna know a very cool little factoid about my Rocket to Raptor transition?

There's no need to disassemble or drain the loop to install a new motherboard, CPU, DDR5, PCIe 5.0 SSD, simply lift the EKWB water block out of the way and drop right back on top the new hardware.

How many custom water loops can do that? lol

So much more than a pretty face.

 

[igor_kavinski]

Rocket to Raptor transition

Your patience will be rewarded handsomely

 

[aigomorla]

How many custom water loops can do that? lol

Is this a formal challenge to me?

Spoiler

Lemme introduce you to a real torcher rack... from days when Bling was secondary, and performance was primary...
Symmetry always makes things look so much nicer...


I used to get ESs like this to cherry pick.

That is one of the beds i validated Intel ES's on.
So i take your challange and raise you 2 x480's on an original DangerDen Rack, which btw is the grandfather of all open racks.
Meaning they were first.

Oh and that full cover motherboard block is also the grandfather of all future full board blocks, as i and a friend of mine named BeiFei invented that style as i needed something to keep the board cool as well when pumping insane amounts of voltages to validate said CPU's.

I take your challenge and double down on you sir.

 

[DrMrLordX]

I hope we don't get to 100 pages in this thread speculating/bickering before the actual reviews hit the web

That's what you should expect before the product launches. Any day now, Intel . . .

Doesn't using chiplets like AMD has been doing for some time now increase memory latency times?

tbh all of Intel's DDR4 designs (at least) have gotten worse latency-wise since Comet Lake. Rocket Lake had a naff DDR4 memory controller, and Alder Lake isn't much better. I haven't yet seen DDR4 latencies anywhere near as good on Rocket Lake or Alder Lake as they were on Comet Lake.

Once Intel goes chiplet they'll be 100% DDR5 so nobody will be able to do like-for-like comparisons with Comet Lake. DDR5 increases mem latency regardless.

 

[aigomorla]

Once Intel goes chiplet they'll be 100% DDR5 so nobody will be able to do like-for-like comparisons with Comet Lake. DDR5 increases mem latency regardless.

One of the greatest difficulties is comparing cross bench across memory platforms like that.
We had issues like that during DDR2 -> DDR3 and DDR3 -> DDR4

There would always be several boards which allowed older ram on newer chips, but mostly showed dismal performance gains. (price / performance ratio was horrible).

But they were mostly back in the days where the memory latency and speed did not impact as much outside calculations of SuperPi.
I still remember getting all excited when i broke that sub 10sec superpi... (OG Penryn CPU) it was pretty similar to getting a 10sec 1/4 mile time in a car.

 

[AlltheWay LeeWhy]

That's what you should expect before the product launches. Any day now, Intel . . .

tbh all of Intel's DDR4 designs (at least) have gotten worse latency-wise since Comet Lake. Rocket Lake had a naff DDR4 memory controller, and Alder Lake isn't much better. I haven't yet seen DDR4 latencies anywhere near as good on Rocket Lake or Alder Lake as they were on Comet Lake.

Once Intel goes chiplet they'll be 100% DDR5 so nobody will be able to do like-for-like comparisons with Comet Lake. DDR5 increases mem latency regardless.

Does the term naff equal daft? I don't know UK lingo. lol

I may be the only enthusiast who had a positive experience with Rocket Lake, only purchased the inexpensive i5 11600K for $300, because there was no way the thermally inefficient cores of the 11900k were going to compose my very 1st 8core processor. We all know RKL was only a stop-gap measure by Intel at best, and Silicon Lottery was asking almost $900 for the high-binned 11900Ks. Absolutely no deal, picked up the 11600K for $300 and the money saved, $600USD went towards the Z590 Apex motherboard.

1) The 11600K ran up to 5.5Ghz every single core using an older DDR4 4600 CL18 b-die kit - Gskill F4-4600C18D-16GTRS

2) Picked up an ultimate Samsung b-die kit XMP'd at 4800 CL17 and that kit ran up to 5066 CL17 - Gskill F4-4800C17D-16TRS

3) Bought the Gskill 5333 CL22 DDR4 Hynix kit and it runs up to 5333 CL19 5600 CL20 and an astounding 5866 CL21 - Gskill F4-5333C22D-16GTES

I'm thinking the IMC within my Rocket Lake 11600K i5 is more stellar than naff.

This video is Bing of ROG pushing the Hynix kit to 6133Mhz using ambient air.

 

[igor_kavinski]

I'm thinking the IMC within my Rocket Lake 11600K i5 is more stellar than naff.

Reviewers focus so much time on the high end i7/i9 parts that they miss out the goodness in the i5 parts. I loved the i5-12400. It's been my fastest ST performance experience in real life and I'm excited that RPL/Zen 4 that I might end up getting will be even faster. This should be an awesome year for system upgrades.

 

[DrMrLordX]

Does the term naff equal daft? I don't know UK lingo. lol

Definition of NAFF

lacking in style or good taste : vulgar and unfashionable… See the full definition

www.merriam-webster.com

1) The 11600K ran up to 5.5Ghz every single core using an older DDR4 4600 CL18 b-die kit - Gskill F4-4600C18D-16GTRS

2) Picked up an ultimate Samsung b-die kit XMP'd at 4800 CL17 and that kit ran up to 5066 CL17 - Gskill F4-4800C17D-16TRS

3) Bought the Gskill 5333 CL22 DDR4 Hynix kit and it runs up to 5333 CL19 5600 CL20 and an astounding 5866 CL21 - Gskill F4-5333C22D-16GTES

I'm thinking the IMC within my Rocket Lake 11600K i5 is more stellar than naff.

That's extremely rare. Early Rocket Lake samples (11900k etc.) were struggling to go past DDR4-3733, and were per clock getting much worse latency than Comet Lake.

 

[AlltheWay LeeWhy]

Definition of NAFF

lacking in style or good taste : vulgar and unfashionable… See the full definition

www.merriam-webster.com

That's extremely rare. Early Rocket Lake samples (11900k etc.) were struggling to go past DDR4-3733, and were per clock getting much worse latency than Comet Lake.

I seem to be the odd man out with RKL, must've gotten lucky.

I do know of another enthusiast "there is another" who purchased the same motherboard and same mem kit as I did, only change - he's using an 11900K, and he's able to achieve ALL my results along with the 6133Mhz super mem speed, that I never wanted to attempt, lol. That's too much voltage for my taste and I need this pc to complete work, so I never became brave enough to take the risk.

 

[AlltheWay LeeWhy]

Is this a formal challenge to me?

Spoiler

Lemme introduce you to a real torcher rack... from days when Bling was secondary, and performance was primary...
Symmetry always makes things look so much nicer...
View attachment 63932View attachment 63933

I used to get ESs like this to cherry pick. View attachment 63934

That is one of the beds i validated Intel ES's on.
So i take your challange and raise you 2 x480's on an original DangerDen Rack, which btw is the grandfather of all open racks.
Meaning they were first.

Oh and that full cover motherboard block is also the grandfather of all future full board blocks, as i and a friend of mine named BeiFei invented that style as i needed something to keep the board cool as well when pumping insane amounts of voltages to validate said CPU's.

I take your challenge and double down on you sir.

I LOVE IT!

Made the move to open air chassis sometime in 2015 with a CaseLabs S8 test bench, does anyone remember those?

Then grabbed a 2nd CL test bench, and currently using a Dimastech Mini V2 test bench as 2ndary rig and the Primochill SX Pro test bench as primary rig.

Ah, those builds from the past.

The Batman:

Five rads, yet I never got around to completing the loop, kinda sad...

 

[TheELF]

First, you create this post all about the 7950x in a Raptor lake thread... Not good. And you are assuming that the 7950x is 230w ppt, NO, the socket AM5 is. From your own article:

"AMD would like to issue a correction to the socket power and TDP limits of the upcoming AMD Socket AM5. AMD Socket AM5 supports up to a 170 W TDP with a PPT up to 230 W. TDP*1.35 is the standard calculation for TDP v. PPT for AMD sockets in the "Zen" era, and the new 170 W TDP group is no exception (170*1.35=229.5). "

As for the temps and the rest, I will wait for the official Raptor lake and Zen 4 reviews...

You really do like to troll don't you.

Yes, AMD sends more power to the socket because the CPU DOESN'T need it, yup checks out.
They just send the additional power to the aether to appease the cybergods from the beyond.

Explaining AMD Ryzen Precision Boost Overdrive (PBO), AutoOC, & Benchmarks | GamersNexus

stub With the launch of the Ryzen 3000 series processors, we’ve noticed a distinct confusion among readers and viewers when it comes to the phrases “Precision Boost 2,” “XFR,” “Precision Boost Overdrive,” which is different from Precision Boost, and “AutoOC.” There is also a lot of confusion...

www.gamersnexus.net

Package Power Tracking (“PPT”): The PPT threshold is the allowed socket power consumption permitted across the voltage rails supplying the socket. Applications with high thread counts, and/or “heavy” threads, can encounter PPT limits that can be alleviated with a raised PPT limit.

Sure, if they follow the guidelines then it will be more tame. Intel's whole PL2 = PL1 think with Alder Lake K series is basically just giving up and acknowledging that that's something many motherboards we're doing anyway, so they might as well legitimize it. Not sure what AMD's policy is (power limits enforced on board manufacturers or not), but conceptually they could do similar.

PL1=PL2 was already legitimate in the previous gen it was covered by the warranty and everything, they just tried to make it more clear to the common person that PL1=125 and PL2=241 is the "base power" usage while PL1=PL2 is the "max turbo" usage.

 

[AlltheWay LeeWhy]

delete

 

[AlltheWay LeeWhy]

delete

 

[igor_kavinski]

ASRock's next-gen Z790/H770 motherboard lineup includes DDR4 variants - VideoCardz.com

DDR4 lives! Rejoice, those with expensive kits!

 

[nicalandia]

I did not see this posted anywhere here>



Delays in Raptor Lake-S

 

[deasd]

Oh....rumor time... it seems looks bad

https://twitter.com/x/status/1545218052207177729

edit: wrong thread maybe. sorry. too much branch threads right now....

 

[moinmoin]

Oh....rumor time... it seems looks bad

https://twitter.com/x/status/1545218052207177729

View attachment 64197

Point 4 doesn't seem possible to me on such a short notice. But Intel going mixed node chiplets then delaying is certainly bound to become a juicy mess. Though knowing Intel I'd expect Intel to stockpile "early" dies and pretend everything worked as planned once all chiplets are up to speed and can be launched as complete chips.

 

[Henry swagger]

Intel Core i9-13900K 5.5 GHz Raptor Lake CPU Benchmarked, Crushes The Ryzen 9 5950X & Core i9-12900K In Single & Multi-Threaded Tests

Intel's Core i9-13900K 5.5 GHz Raptor Lake CPU has been benchmarked within Geekbench 5 and obliterates the Core i9-12900K & Ryzen 9 5950X.

wccftech.com

Beast 43% faster than 12900k in multi-core performance before final clocks.. will be even higher in final silicon 😁😃

 

[nicalandia]

Intel Core i9-13900K 5.5 GHz Raptor Lake CPU Benchmarked, Crushes The Ryzen 9 5950X & Core i9-12900K In Single & Multi-Threaded Tests

Intel's Core i9-13900K 5.5 GHz Raptor Lake CPU has been benchmarked within Geekbench 5 and obliterates the Core i9-12900K & Ryzen 9 5950X.

wccftech.com

Beast 43% faster than 12900k in multi-core performance before final clocks.. will be even higher in final silicon 😁😃

Not faster than this one 12900K

 

[igor_kavinski]

View attachment 64325

That has got to be LN2 cooled 6+ GHz.

 

[nicalandia]

That has got to be LN2 cooled 6+ GHz.

Shhh....