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.

 

[igor_kavinski]

nothing that has to do with the CPU perf but with the software being exclusively, and highly, optimised for one uarch.

That's nothing new. Intel devotes a lot of resources to ensure that software is well optimized for their CPUs, even from third party vendors of popular software. That's the benefit they get from having the lion's share of the market (not talking about enthusiasts but in general, office PCs and typical households with no little to no IT knowledge).

Software developers will most likely be given an Intel corporate PC and that's what they will write their software on and try to optimize for. I don't suppose there are many software publishers that go to great lengths to ensure that their software runs well on AMD CPUs. It's also not bad in a sense. It forces AMD engineers to design their architecture based on the typical instruction mix of existing software rather than relying on developers to optimize for their CPUs which may or may not happen. I think AMD is doing really well performance wise (especially in AVX-512) for a company that a large percentage of the public may not even be aware of.

 

[TheELF]

It s not that there s only a low thread count, it s that for some reason Photoshop seems to be exclusively optimised for Intel, FI a 12600K roughly match a 7950X in this app.

Computerbase tests with Photoshop show that there s something that is cache dependant since a 5800X3D perform better than the 10% higher clocked 5950X, rest is specific optimisation for Intel to explain such a discrepancy.

The difference between the 13900K and the 7950X is at least 60% better for the former, wich is telling that Photoshop is all but an equally optimised soft for both uarch, actually that s optimisation for one and litteraly unoptimisation for the other...

Core i9-13900K, i7-13700K & i5-13600K: Gaming-Könige im Test: Benchmarks in Apps

Intel Raptor Lake im Test: Benchmarks in Apps / So wurde getestet / Leistung in Single-Core-Lasten (klassisch)

www.computerbase.de

Same tests from Computerbase show the 8C 7700X at 1-2% better IPC on MT than a 8 + 0 12900K/13700K, so 70% advantage for a 13900K vs a 7950X is nothing that has to do with the CPU perf but with the software being exclusively, and highly, optimised for one uarch.

What numbers are you referring to?!
Because the computerbase bench you link to doesn't even include photoshop, it does have photoscan but the numbers don't match what you are saying.

The numbers do match to adobe premiere, same company different app.
But it's not 60% better for intel, it's 37% lower time for intel, or it does the task in ~60% of the time.
You always use the bigger number as the base for figuring out percentages.

 

[Abwx]

What numbers are you referring to?!
Because the computerbase bench you link to doesn't even include photoshop, it does have photoscan but the numbers don't match what you are saying.

The numbers do match to adobe premiere, same company different app.
But it's not 60% better for intel, it's 37% lower time for intel, or it does the task in ~60% of the time.
You always use the bigger number as the base for figuring out percentages.

37% lower time means 58.7% faster...

Adobe Premiere is listed at Computerbase and it use Photoshop among others, the numbers are such that it s impossible that it s the CPU pure perfs that make this difference,.

Either it s exclusively optimised for Intel or the Intel CPUs also use the GPU to accelerate the process, Computerbase state that s it s software only, wich is contradicted by the huge numbers difference.

Core i9-13900K, i7-13700K & i5-13600K: Gaming-Könige im Test: Benchmarks in Apps

Intel Raptor Lake im Test: Benchmarks in Apps / So wurde getestet / Leistung in Single-Core-Lasten (klassisch)

www.computerbase.de

 

[Hulk]

This is why many pages ago I wrote that ultimately when CPU's are as closely matched as the ones we are choosing today it comes down to which performs best at the applications most important to you in a given price range.

Eventually, when blasting through Cinebench gets old you actually have to get some work done!

 

[A///]

photoshop is speed sensitive and has been for as long as i can remember. the speed differential between the raptorlake 13900k and the ryzen 7950 on all core and single or dual core boost is minimal, but it may be enough to sway results in the faster in speed chip. photoshop can use hardware acceleration but it isn't across the board. with each major revision adobe adds new tasks that can be accelerated through hardware on the chip or a dedicated card.

the intel iris xe on 11th gen operates outside thouse bound walls and can improve performance even with hardware acceleration turned off. until someone does a deep dive on raptorlake no one will know what intel has included, even though the uhd 770 is present on alderlake, too.

 

[A///]

Eventually, when blasting through Cinebench gets old you actually have to get some work done!

cb has its merits but anyone who's working with maxon software will be using a gpu based third party render tool because cpu is too inefficient even if gpu rendering uses more power. there's a vf curve to this but you can watt limit yoour gpu and still get brilliant performance. cpu rendering has its place in video imo, the caveat being a gpu render on 3d models is easier to look at frame by frame versus a real life video frame by frame that benefits from a higher quality rendering which would be software. redshift and octane are the two i know of for cinema4d. there is arnold out there but it's for autodek and uses both gpu and cpu at the same time.

 

[Hulk]

cb has its merits but anyone who's working with maxon software will be using a gpu based third party render tool because cpu is too inefficient even if gpu rendering uses more power. there's a vf curve to this but you can watt limit yoour gpu and still get brilliant performance. cpu rendering has its place in video imo, the caveat being a gpu render on 3d models is easier to look at frame by frame versus a real life video frame by frame that benefits from a higher quality rendering which would be software. redshift and octane are the two i know of for cinema4d. there is arnold out there but it's for autodek and uses both gpu and cpu at the same time.

Of course. CB is simply a widely used benchmarking tool. It's portable, easy to use, and has no readily accessible tweaks. All of these factors have made it useful for cpu-to-cpu comparisons.

I didn't mean to imply that if you are using CB you could buy a CPU based on that performance. Of course anyone using GPU accelerated software would be able to discern that!

 

[A///]

Of course. CB is simply a widely used benchmarking tool. It's portable, easy to use, and has no readily accessible tweaks. All of these factors have made it useful for cpu-to-cpu comparisons.

I didn't mean to imply that if you are using CB you could buy a CPU based on that performance. Of course anyone using GPU accelerated software would be able to discern that!

it's a little complex than that. iirc i remember being told something at work about r23 being developed with integer or floating point maths being the dominant scores assessor and any processor not focusing on that will be graded lower. I work in the arm world and as such the software means nothing to me outside of this hobby. we use other benchmarking software to evaluate what's being made. what I was told sounded like bs but I couldn't tell you if it was or wasn't. the last time I used cb was a decade ago.

 

[DrMrLordX]

CB is almost entirely floating point for what it's worth. That's why it's an interesting benchmark.

 

[Hulk]

it's a little complex than that. iirc i remember being told something at work about r23 being developed with integer or floating point maths being the dominant scores assessor and any processor not focusing on that will be graded lower. I work in the arm world and as such the software means nothing to me outside of this hobby. we use other benchmarking software to evaluate what's being made. what I was told sounded like bs but I couldn't tell you if it was or wasn't. the last time I used cb was a decade ago.

CB is primarily relevant for CPU comparisons because it is ubiquitous.

It's not complicated at all. That's it.

 

[MarkPost]

37% lower time means 58.7% faster...

Adobe Premiere is listed at Computerbase and it use Photoshop among others, the numbers are such that it s impossible that it s the CPU pure perfs that make this difference,.

Either it s exclusively optimised for Intel or the Intel CPUs also use the GPU to accelerate the process, Computerbase state that s it s software only, wich is contradicted by the huge numbers difference.

Core i9-13900K, i7-13700K & i5-13600K: Gaming-Könige im Test: Benchmarks in Apps

Intel Raptor Lake im Test: Benchmarks in Apps / So wurde getestet / Leistung in Single-Core-Lasten (klassisch)

www.computerbase.de

I have no doubt the answer is the bolded part: Intel CPUs are using GPU acceleration in that Computerbase Premiere benchmark (AMD CPUs doesnt). Really crappy testing from Computerbase if you ask me.

I've created a quick Premiere project (with clip posted in Handbrake thread) with some random video effects, and encoded to HEVC, making sure GPU acceleration is off in both systems, as can be seen in screenshots (anyways my 12900 is KF, so no iGPU and my 7950X is configured with iGPU disabled in BIOS):

12900KF oced @5.0 P-cores/4.0 E-cores

7950X @stock

So a stock 7950X is in fact faster (~14%) than an oced 12900KF

btw HEVC encoding with Premiere only uses 16 threads as much

 

[TheELF]

I have no doubt the answer is the bolded part: Intel CPUs are using GPU acceleration in that Computerbase Premiere benchmark (AMD CPUs doesnt). Really crappy testing from Computerbase if you ask me.

I've created a quick Premiere project (with clip posted in Handbrake thread) with some random video effects, and encoded to HEVC, making sure GPU acceleration is off in both systems, as can be seen in screenshots (anyways my 12900 is KF, so no iGPU and my 7950X is configured with iGPU disabled in BIOS):

12900KF oced @5.0 P-cores/4.0 E-cores

7950X @stock

So a stock 7950X is in fact faster (~14%) than an oced 12900KF

btw HEVC encoding with Premiere only uses 16 threads as much

The 12900kf can boost to 5.2Ghz so if this test uses anything lightly threaded you did UNDERclock it.
Also we are talking about how the 13900k compares so the 12900k isn't really all that relevant.
Also also if it only uses 16 threads then why did you OC the e-cores?
So you increased the 12900k from an normal allcore of 4.9 to 5Ghz just so you can say that you compared it OC to stock and it still lost...

 

[MarkPost]

The 12900kf can boost to 5.2Ghz so if this test uses anything lightly threaded you did UNDERclock it.

What are you talking about? 12900KF boost in MT tasks (actually 16 threads are MT you know) isn't 5.2 AT ALL (thats only and exclusively ST boost). MT freq is 4.9 ever. So, yes actually its OVERclock for MT tasks.

Also we are talking about how the 13900k compares so the 12900k isn't really all that relevant.

We are talking about that Premiere benchmark at Computerbase, and if its misleading or not. And the answer to the question is YES, it is, as I showed.

On the other hand, in this Premiere test, diff between 13900 and 12900 is only a freq one. It would be nice if some 13900 owner run an encoding with Premiere to see 13900 freq. Keeping in mind no IPC diffs and that 16 threads are used, I would say 13900 freq its not higher than ~5.2-5.3

Also also if it only uses 16 threads then why did you OC the e-cores?

Just because its my 24/12 config

So you increased the 12900k from an normal allcore of 4.9 to 5Ghz just so you can say that you compared it OC to stock and it still lost...

No as I said its my 24/12 config. Anyways running at 5.0 is closer to 13900 MT clocks than 4.9

 

[Henry swagger]

https://twitter.com/x/status/1591037459000721408

Alder lake and raptor lake are doing beautifully sales wise.. intel regaining market leadership

 

[Markfw]

https://twitter.com/x/status/1591037459000721408

Alder lake and raptor lake are doing beautifully sales wise.. intel regaining market leadership

The fact that ONE place thinks they are doing good in sales on two different markets does not mean Intel is regaining market leadership.

For example, in the most lucrative market, server, shat do they have ? NOTHING that can touch Genoa. You sound like like a marketing person who works for Intel.

 

[Hulk]

It's also strange that for the Surface Laptop 5 there are no AMD CPU options? It seems like Intel has some leverage and turned the screws on MS.

 

[A///]

The fact that ONE place thinks they are doing good in sales on two different markets does not mean Intel is regaining market leadership.

For example, in the most lucrative market, server, shat do they have ? NOTHING that can touch Genoa. You sound like like a marketing person who works for Intel.

Harry seems to flip flop between intel and amd when he's not missing his number 9 key on his keyboard. he reminds me of arne.

It's also strange that for the Surface Laptop 5 there are no AMD CPU options? It seems like Intel has some leverage and turned the screws on MS.

No? It's assumed AMD couldn't guarantee a steady supply of mobile chips to microsoft simply due to being constrained and having to produce nearly everything on a single node and these new surfaces come at a time when intel had new products but amd did not in mobile. im not entirely sure who is buying surface laptops because they're a terrible deal for what you get.

 

[naad]

https://twitter.com/x/status/1591037459000721408

Alder lake and raptor lake are doing beautifully sales wise.. intel regaining market leadership

Intel flooding the channel with cheap chips, while AMD is trying to keep margins, just look at the Q3 CCG margin drop YoY, from 37% to 20%, like datacenter they're giving away chips for peanuts.

I seriously doubt this is sustainable for Intel, but best of luck to them.

 

[A///]

Intel flooding the channel with cheap chips, while AMD is trying to keep margins, just look at the Q3 CCG margin drop YoY, from 37% to 20%, like datacenter they're giving away chips for peanuts.

I seriously doubt this is sustainable for Intel, but best of luck to them.

intel is letting these chips go for cheap maybe at a loss to say we're gaining market share. intel's the town bicycle, the trick on the corner waving her bum for people driving down the street to leer at her. The wash performance and high power and heat are to be ignored.

Five or six years later i find it perplexing how the tables have turned. intel is facing a black plague in server even with sapphire rapids launching in 1h23. it's way behind the times in every metric and at a higher cost.

 

[naad]

intel is letting these chips go for cheap maybe at a loss to say we're gaining market share. intel's the town bicycle, the trick on the corner waving her bum for people driving down the street to leer at her. The wash performance and high power and heat are to be ignored.

Five or six years later i find it perplexing how the tables have turned. intel is facing a black plague in server even with sapphire rapids launching in 1h23. it's way behind the times in every metric and at a higher cost.

Pat probably thought its better to make no profit and have fabs utilized than have even less profit and have fabs sitting idle, even if margins have to drop to single digits..
the old managment would never sacrifice margins

 

[A///]

Pat probably thought its better to make no profit and have fabs utilized than have even less profit and have fabs sitting idle, even if margins have to drop to single digits..
the old managment would never sacrifice margins

Pat? Pat Benetar?

 

[Markfw]

Pat probably thought its better to make no profit and have fabs utilized than have even less profit and have fabs sitting idle, even if margins have to drop to single digits..
the old managment would never sacrifice margins

Now THAT sounds reasonable. Give crap away, just to make it look like they are doing well.

 

[naad]

Pat? Pat Benetar?

Gelsinger, the CEO of Intel

 

[A///]

Now THAT sounds reasonable. Give crap away, just to make it look like they are doing well.

yes in market share, but come quarter reports the numbers do not lie. You can dress an ugly dog up with whatever cute outfit you want but it's still an ugly dog.

 

[A///]

Gelsinger, the CEO of Intel

Yeah that came to me way after I posted. i've called him gelsinger for so long i'd thought his name was pete or something equally un pat.