Core i9-10990XE 22-core Processor Last Gasp of the X299 Platform?

[techguymaxc]

At all core 5.0 ghz, this thing is going to draw around 600w. Other reviews show the 18 core drawing in excess of 500w when running 4.8 ghz.

Even with a 360mm custom loop, the 4.8 ghz 10980xe was running at 100*C at Hardware Unboxed.

Here is total system power consumption of various overclocked Cacade Lake parts.

The 14000 CB20 score was done at 5.0 ghz.

Skylake has about a 125 point multiplier in CB20 which means (# cores w/ht) x (frequency) x 125 = CB20 score.

The point is, the 3960x will use about half the power to get a similiar MT performance.

Not quite.

This won't be 5GHz all-core, 4GHz base clock is indicated in the CPU-Z screenshot and was mentioned in the article.

5GHz is a single-core Turbo clock. This is in line with other recent Intel HEDT SKUs.

BOOORING.

Intel is releasing press announcements "hey look we managed to beat a couple more cores out of the Skylake/14nm+++++++++++ dead horse", meanwhile AMD is all "YOU get more cores! And YOU get more cores! EVEREYBODY GETS MOAR COREZ!!!!!111"

Seriously Intel, you gotta learn when to stop. You already look ridiculous, every extra pitiful increment of your current dead uArch just makes you look more and more desperate. And desperation isn't a good look for a company your size. Just eat that humble pie, admit that you can't compete with AMD in its current form, stop wasting resources on poor dead Skylake, and focus on getting desktop 10nm parts to consumers.

Such a simple-minded view.

Not every workload scales with core count, let alone to the degree necessary to invalidate the existence of this (still) hypothetical CPU.

AMD is not your favorite sports team. They're not going to reward you with free computer parts for cheering them on. If you like their products, they expect you to buy them. It is a mutually-beneficial exchange.

 

[Vya Domus]

Not every workload scales with core count, let alone to the degree necessary to invalidate the existence of this (still) hypothetical CPU.

So let's see, if something doesn't scale well with core count then this CPU wont be any better than the rest of Intel's line up, correct ? If it does scale well, AMD has CPUs out there which will have up to 200% more cores. Meanwhile this has 4 more cores and a higher base clock, not exactly ground breaking.

So, I ask you, how would you validate the existence of this (still) hypothetical CPU ? What exactly does this product bring to the table that's new and exciting ?

Let's face it, we can beat around the bush all day, it wont change the fact that Intel simply got obliterated this round in HEDT. Usually I try to avoid expressions like those but there's really no other way to put it, it's like 6950X vs FX-9590 back a few years ago, except the roles are reversed and the gap is maybe even worse. It's not about your favorite team, it's an objective observation.

 

[techguymaxc]

So let's see, if something doesn't scale well with core count then this CPU wont be any better than the rest of Intel's line up, correct ? If it does scale well, AMD has CPUs out there which will have up to 200% more cores. Meanwhile this has 4 more cores and a higher base clock, not exactly ground breaking.

So, I ask you, how would you validate the existence of this (still) hypothetical CPU ? What exactly does this product bring to the table that's new and exciting ?

Let's face it, we can beat around the bush all day, it wont change the fact that Intel simply got obliterated this round in HEDT. Usually I try to avoid expressions like those but there's really no other way to put it, it's like 6950X vs FX-9590 back a few years ago, except the roles are reversed and the gap is maybe even worse. It's not about your favorite team, it's an objective observation.

I had a feeling this argument would rear it's head.

So it's not that workloads don't scale, it's that scaling in most workloads only goes so far. At that point, performance is derived from the combination of IPC and clock speed. It's the balanced breakfast of CPU performance scaling.

 

[Vya Domus]

I had a feeling this argument would rear it's head.

So it's not that workloads don't scale, it's that scaling in most workloads only goes so far. At that point, performance is derived from the combination of IPC and clock speed. It's the balanced breakfast of CPU performance scaling.

Cascade Lake-X and Zen 2 are a very close match both in IPC and clock speed, they're both just as balanced, the only real difference is that one of them comes in higher core counts, none of the balance gets lost along the way. In Intel's case we are talking about a CPU with 4 more cores and a little bit higher clocks, let's face it that wont change much, it's kind of ridiculous to believe that it will be more "balanced" in any real way.

 

[Makaveli]

22 cores OMG... Wait, amd has 64 cores to the consumer marked. Go home Intel, you are drunk and still produced on 14 NM combined with 22 cores regardless of fake or real, we all know what that means for power consumption

And still doesn't Change the fact that x299 is a living dinosaur now with only pcie gen 3 and so on. A living fossil from the past.

Frankly now I think of it, 380 watt tdp, 4 ghz baseclock and 5 ghz boost single, dual or maybe all core boost? Who knows. My point is that this smell of I9 9990XE all over again just now with 22 cores and just as high power consumption and heat output

For the Prosumer market.

No consumer is spending 4k on a 64 core threadripper.

Just wanted to add that correction.

 

[Nihilus]

Not quite.

This won't be 5GHz all-core, 4GHz base clock is indicated in the CPU-Z screenshot and was mentioned in the article.

5GHz is a single-core Turbo clock. This is in line with other recent Intel HEDT SKUs.

So if 4 ghz is base all core boost should be around 4.5 ghz? The CB20 run was done at 5.0 ghz overclock in any case.

The 3960X is using 30% less power (going from the graph).

22 cores at 5.0 ghz should be closer to 700w total power consumption. So around 75% more power. The 10990xe at 5.0 ghz should be neck and neck with the 3960x stock on blender.

 

[T1beriu]

So much fake.

 

[thesmokingman]

For the Prosumer market.

No consumer is spending 4k on a 64 core threadripper.

Just wanted to add that correction.

Then they're more likely to buy a middling 22core cpu that has nearly 400w tdp?

 

[GorbazTheDragon]

Let's not forget that the mesh parts tend to underperform relative to the ring parts due to the higher latency to IMC... I really wonder how slapping an extra row of core tiles will affect the performance of the mesh... There's a reason the XCC parts are 6 instead of 4 wide.

As far as per core throughput goes, zen 2 most definitely matches these mesh intel parts, the ring parts still have better memory latency so have a tendency to pull ahead in some stuff, and they can be clocked faster... That said, the diminishing returns you get as far as power consumption goes makes the argument for pushing those clocks for "workloads that don't scale well past 10 cores" really marginal...

The only real advantage 2066 has at this point is the wide PCIe interface, 20x4.0 of AM4 is pretty close to the overall bandwidth but 4.0 devices are still too sparse on the market to make use of that... Intel needs to push down the price of this platform more for it to be attractive, I'm talking of order 2/3rd the prices for the motherboards and a prospective 22 core be well under $900... Like 1151, it's a really dead end platform and will be well outclassed by it's AMD counterparts in 2-3 years time.

 

[techguymaxc]

Cascade Lake-X and Zen 2 are a very close match both in IPC and clock speed, they're both just as balanced, the only real difference is that one of them comes in higher core counts, none of the balance gets lost along the way. In Intel's case we are talking about a CPU with 4 more cores and a little bit higher clocks, let's face it that wont change much, it's kind of ridiculous to believe that it will be more "balanced" in any real way.

Let's compare with 10980xe, the current flagship part for LGA2066.
18 cores
3GHz base clock
4.6GHz Boost
4.8GHz Turbo Boost Max 3.0

Compare with a hypothetical 10990xe:
22 cores
4GHz base clock
4.xGHz Boost
5.0GHz Turbo Boost Max 3.0

Which of these chips is faster?

As for the comparisons with Zen 2 uarch, some workloads still run better on Intel. Adobe products are a good example of this.

For me, I find Intel chips to be faster for my video editing workflow, thanks to higher clock speeds and AVX 512. Personally, I am holding out for Zen 3 before I make a final decision on my next workstation chip. Rumor has it AMD will finally introduce AVX 512 support with Zen 3.

 

[Vya Domus]

For me, I find Intel chips to be faster for my video editing workflow, thanks to higher clock speeds and AVX 512. Personally, I am holding out for Zen 3 before I make a final decision on my next workstation chip. Rumor has it AMD will finally introduce AVX 512 support with Zen 3.

The throughput of a 10990XE under AVX 512 will be lower than what a 3990X can do with just AVX2 if we compare flagship with flagship, it will also be more likely to get that throughput since more things are written for AVX2 and whatever was written with AVX512 probably has a AVX2 fallback as well. And if we talk about a potential 48 core part, that will be about the same while still having an advantage under more circumstance.

Intel no longer has an advantage with AVX 512, not until they get more cores. I'm also yet to find out how of the much commercial software out there uses that, it's not very practical to implement and comes with some other issues.

Adobe products are a good example of this.

Are they really ?

I'm being kind of pesky, I know, but that's the reality. No matter how you spin it Intel lost pretty much every big advantage they had.

Rumor has it AMD will finally introduce AVX 512 support with Zen 3.

I really hope they don't. That brings more problems than improvements. They've explicitly stated their dislike for larger SIMD and for good I reason, they don't have a place in a world where GPUs can do the same things orders of magnitude faster and easier. If they will support it I hope they do it by fusing 2x 256 bit instructions, it's just not worth ruining their power envelope with the amount of cores that they have.

 

[champsilva]

The throughput of a 10990XE under AVX 512 will be lower than what a 3990X can do with just AVX2 if we compare flagship with flagship, it will also be more likely to get that throughput since more things are written for AVX2 and whatever was written with AVX512 probably has a AVX2 fallback as well. And if we talk about a potential 48 core part, that will be about the same while still having an advantage under more circumstance.

Intel no longer has an advantage with AVX 512, not until they get more cores. I'm also yet to find out how of the much commercial software out there uses that, it's not very practical to implement and comes with some other issues.



Are they really ?

View attachment 142228
View attachment 142229

I'm being kind of pesky, I know, but that's the reality. No matter how you spin it Intel lost pretty much every big advantage they had.



I really hope they don't. That brings more problems than improvements. They've explicitly stated their dislike for larger SIMD and for good I reason, they don't have a place in a world where GPUs can do the same things orders of magnitude faster and easier.

3990X = $4000, 10980XE = $978

Let's say this 10990XE = $1300, im putting more than $300 for 4 core CPU.

And he didn't said his workflow. And not everything is about rendering.

• For Adobe Premiere Pro and After Effects, the following CPUs are our recommendations depending on your budget:
  
  • AMD Ryzen 7 3800X ($399)
  • Intel Core i9 9900K ($499)
  • Intel Core i9 10920X ($689)
  • AMD Ryzen 9 3950X ($749)
  • Intel Core i9 10940X ($784)
  • Intel Core i9 10980XE ($979)
  • AMD Threadripper 3960X ($1,399)

Source: https://www.pugetsystems.com/labs/a...-Editing-2019-1633/#PremiereProCPUPerformance

The threadripper performs good with Premiere and After, but the price difference isn't worth. Its like 5% performance for $400 difference.

 

[Vya Domus]

Its like 5% performance for $400 difference.

Uhm, sometimes "it's like" 35% better unless I'm reading the wrong scores in the source you provided. That's significant and can totally be worth it for 400$.

And not everything is about rendering.

Indeed but often AMD's offerings beat Intel's in performance even outside of rendering.

3990X = $4000, 10980XE = $978

And ? The 3990X will demolish it, of course it's a lot more expensive. Those are the flagships, it's kind of ridiculous but that's the reality. How else would you put it ?

 

[techguymaxc]

The throughput of a 10990XE under AVX 512 will be lower than what a 3990X can do with just AVX2 if we compare flagship with flagship, it will also be more likely to get that throughput since more things are written for AVX2 and whatever was written with AVX512 probably has a AVX2 fallback as well. And if we talk about a potential 48 core part, that will be about the same while still having an advantage under more circumstance.

Intel no longer has an advantage with AVX 512, not until they get more cores. I'm also yet to find out how of the much commercial software out there uses that, it's not very practical to implement and comes with some other issues.

The scenario you describe here assumes sustained peak throughput across all cores.

Are more cores better? Sure, if you can take advantage of them. Most workloads simply don't.

Are they really ?

View attachment 142228
View attachment 142229

I'm being kind of pesky, I know, but that's the reality. No matter how you spin it Intel lost pretty much every big advantage they had.

Intel still leads in workloads that only scale to mild-moderate thread counts, again, thanks to IPC and clock speed.

I'm not saying this hypothetical CPU is the perfect CPU for every workload, but it would bring more performance to me.

I really hope they don't. That brings more problems than improvements. They've explicitly stated their dislike for larger SIMD and for good I reason, they don't have a place in a world where GPUs can do the same things orders of magnitude faster and easier. If they will support it I hope they do it by fusing 2x 256 bit instructions, it's just not worth ruining their power envelope with the amount of cores that they have.

I hope they introduce AVX512 support, it's pretty much the only thing keeping me from switching at this point. They likely don't need to increase AVX unit width to 512-bits natively, I would be fine with a fused approach.

 

[Vya Domus]

Intel still leads in workloads that only scale to mild-moderate thread counts

What's a "mild-moderate" thread count ?

We're scraping the bottom of the barrel here.

The scenario you describe here assumes sustained peak throughput across all cores.

It doesn't assume anything about that, both Zen 2 and Cascade-Lake drop in clock speed immediately as soon as there is a vector load. Zen 2 much less so, taking that into account a 3990X would probably end up being a fair bit quicker than a 10990XE under vector loads.

Listen, you don't have to believe me that AVX512 isn't as amazing as you think it is:

That's pretty much the most ideal case for AVX 512 and two 7742 with just 256 bit AVX are just as fast as two 8280s.

Those are server CPUs but there is no reason to believe it's any different on desktops. And keep in mind those are 28 core Xeons, not 22 where it would have been faster. It's not about sustained peak, it's just how the math adds up.

Most workloads simply don't.

You wouldn't touch any of these CPUs to run workloads that don't. Come one, this argument is ... uninspired, to put it in a more elegant way. No, whoever buys these has a clear purpose for them in mind where they actually make sense, because they're made exactly for when the workloads scale well.

I've always found this argument to be exceedingly bizarre, clearly neither AMD or Intel actually believes in what you're saying, otherwise they wouldn't put dozens of cores in their CPUs and try and sell them for a premium. It's painfully obvious that core counts are the priority for both.

 

[GorbazTheDragon]

Intel still leads in workloads that only scale to mild-moderate thread counts, again, thanks to IPC and clock speed.

Stop. The raw FP throughput for a ~4.1GHz zen2 core is the same as a ~5GHz xLake core. It ends there.

The advantage xLake has is that its cores have better throughput in certain programs which weren't yet optimised for zen 2 cores, my understanding is that in some cases updates to the software have changed that.

The real advantage that xLake really lies in the lower memory latency, so anything with unpredictable memory accesses will be less gated due to the longer wait times on memory accesses. This can also interact back with the optimisation and how the programs play with the OoO schedulers in the different architectures, because the penalty for bad branch predicts becomes worse for distant memory accesses.

 

[Vayra86]

At all core 5.0 ghz, this thing is going to draw around 600w. Other reviews show the 18 core drawing in excess of 500w when running 4.8 ghz.

Even with a 360mm custom loop, the 4.8 ghz 10980xe was running at 100*C at Hardware Unboxed.

Here is total system power consumption of various overclocked Cacade Lake parts.
View attachment 142120

Redirect Notice

The 14000 CB20 score was done at 5.0 ghz.

Skylake has about a 125 point multiplier in CB20 which means (# cores w/ht) x (frequency) x 125 = CB20 score.

The point is, the 3960x will use about half the power to get a similiar MT performance.

Soon we'll discover it was Intel who was mainly responsible for climate change all along Six. Hundred. Watts?!

 

[Makaveli]

Then they're more likely to buy a middling 22core cpu that has nearly 400w tdp?

Nope neither of these cpu's are for normal consumers, they are for Prosumers.

Anyone building with these should generally be earning income with these machines not playing games.

Let's compare with 10980xe, the current flagship part for LGA2066.
18 cores
3GHz base clock
4.6GHz Boost
4.8GHz Turbo Boost Max 3.0

Compare with a hypothetical 10990xe:
22 cores
4GHz base clock
4.xGHz Boost
5.0GHz Turbo Boost Max 3.0

Which of these chips is faster?

As for the comparisons with Zen 2 uarch, some workloads still run better on Intel. Adobe products are a good example of this.

For me, I find Intel chips to be faster for my video editing workflow, thanks to higher clock speeds and AVX 512. Personally, I am holding out for Zen 3 before I make a final decision on my next workstation chip. Rumor has it AMD will finally introduce AVX 512 support with Zen 3.

My gut says it will be implemented the same way AVX2 was in Ryzen one.

So Zen3 will process AVX512 in 256x2.

And they will wait for 5nm to do AVX512 in a single clock cycle.