AMD Ryzen 9 7950X Allegedly 40% Faster than 5950X in CPU-Z Bench Multi-Threaded

[Ellertis]

But more expensive. The 13700k is very likely to be in the $400 range, the 5950x was once $800. A part that costs half as much (at starting msrp) will be faster. That’s competition at play. Finally, The 7950x is going to be at least $700 probably more. But it will be fast for sure gotta give amd kudos for that.

Wouldn`t it be reasonable to expect the 7900x to be faster than the 13700k and a bit more expensive too

 

[erek]

All of a sudden I have the urge to sell my 5800X...

what about your 5800X3D?

 

[Tropick]

God I love CPU wars. I still remember the late 90s.

You're not kidding, It feels great, been waiting for this moment since FX. Finally the two titans are in a proper arms race again, AMD pushing the hi-cache chiplet route and Intel hitting back with their split P+E architecture. Exciting times.

People need to take a step back from their brand loyalty for a second and just appreciate the pace of innovation we've been seeing here lately. So many are quick to forget (or even think fondly of) the stagnation we saw in the early to mid 2010s. AMD's weak heavy machinery architectures paving the way for Intel to begin resting on their laurels put the industry on its ass for way too long. Feels so good to see the two firing on all cylinders in the CPU space again. Here's to hoping we see a proper K6 vs. Pentium era once again!

 

[Makaveli]

They're showing the 7950x with a ~40% uplift over zen3 5950x, but the 7700x is only doing ~20% over the 5800x, everything still to be seen basically

I think the reason for this is the 5950x was really limited in what it can do at 142w PPT. The 7950x has a much larger power budget to work with besides all the arch changes.

 

[maxfly]

Good times... to be a leaker, hahaha! If true it's friggin great. I just hope the rest of the stack kicks ass at both multi and single. Well I'll be honest. I'm only interested in single.
Competition is guud.

 

[ymdhis]

The 13700k is looking powerful. Faster than the mighty 5950x.

It's so incredible isn't it, that after something like seven years of Intel CPUs only bumping the clockspeed by 200MHz each time, we are finally seeing processors which actually improve on previous generations.

 

[dj-electric]

People who call E-cores "toy cores" "nerfed cores" and so are quite in denial of how capable they are, even when the results of using them for general purpose multithread computing is staring them in their face.

 

[BlaezaLite]

How about "Helper Cores"? But as long as there is competition, there will always be uplift, so good job Intel! Good job AMD! What's going to happen if one just decimates the other? Complacency again?

 

[ThrashZone]

People who call E-cores "toy cores" "nerfed cores" and so are quite in denial of how capable they are, even when the results of using them for general purpose multithread computing is staring them in their face.

Hi,
Think the real term is thermal defective cores
If e cores can't do the same speed as p cores at the same temperature then they are defects which at one time were binned out.

 

[dj-electric]

If e cores can't do the same speed as p cores at the same temperature then they are defects which at one time were binned out.

That's not how silicon works. Different architecture works in its own frequencies and voltage to frequency curve. Nothing here is "defective".
Silicon isn't run on X frequency universally. E cores and P cores are fabricated entirely differently. They are not the same.

In case you never bothered knowing how the CPU cores look in the SOC before calling them defective, this is ADL-S, with the 8 P cores and 8 E-cores to the right.

These "toy cores" and "defective cores" are built to be very silicon space efficient, and carry an IPC similar to what people met on Skylake \ Zen2

 

[Kelutrel]

That's not how silicon works. Different architecture works in its own frequencies and voltage to frequency curve. Nothing here is "defective".
Silicon isn't run on X frequency universally. E cores and P cores are fabricated entirely differently. They are not the same.

In case you never bothered knowing how the CPU cores look in the SOC before calling them defective, this is ADL-S, with the 8 P cores and 8 E-cores to the right.

These "toy cores" and "defective cores" are built to be very silicon space efficient, and carry an IPC similar to what people met on Skylake \ Zen2

Wow! This is an awesome picture! Amazing!

 

[Nordic]

Simple, the 5950x was gimped hard by its power limit at stock while the 5800x wasn't.

On my own 5950x and by what I have seen with others, the stock power settings are really close to the sweet spot. I am able to increase performance by increasing the power limit at the cost of ever increasing power used. By setting the power limit to the max the motherboard settings my 5950x uses twice as much power.

 

[Denver]

The crucial point is that the i9 gets that score with aggressive OC. "In CPU-Z the CPU reaches as high as 6.18 GHz with all eight Performance cores and 4.69 GHz with all sixteen Efficient cores."

Intel Core i9-13900KF CPU reaches 6.18 GHz with liquid chiller attached in new overclocking test - VideoCardz.com

Intel Core i9-13900KF overclocked to 6.18 GHz thanks to exotic cooling Extreme Player is back with another test of the flagship Intel Raptor Lake-S CPUs. This is not the first test of unreleased hardware from Extreme Player, we have already seen his coverage of Core i9-13900K as well as Core...

videocardz.com

 

[hs4]

Hi,
Think the real term is thermal defective cores
If e cores can't do the same speed as p cores at the same temperature then they are defects which at one time were binned out.

Here is official TSMC FINFLEX™ diagram, the way to produce "thermal defective cores" in N3E node. Designing a core with 3-2 Fin transistors will increase the clock speed, while using 2-1 Fin transistors will result in a "thermal defective" core with low clock speed and high efficiency.

 

[tracker1]

The crucial point is that the i9 gets that score with aggressive OC. "In CPU-Z the CPU reaches as high as 6.18 GHz with all eight Performance cores and 4.69 GHz with all sixteen Efficient cores."

Intel Core i9-13900KF CPU reaches 6.18 GHz with liquid chiller attached in new overclocking test - VideoCardz.com

Intel Core i9-13900KF overclocked to 6.18 GHz thanks to exotic cooling Extreme Player is back with another test of the flagship Intel Raptor Lake-S CPUs. This is not the first test of unreleased hardware from Extreme Player, we have already seen his coverage of Core i9-13900K as well as Core...

videocardz.com

Shocked... Okay not really... It's pretty much assumed that AMD has a slightly better IPC I'm the coming generation and Intel slightly more tolerance for power use and that they'd be very close and trading blows.

I'm willing to give the edge to AMD for lower power use and a better platform longevity. Upgraded a 2600 to a 5700 last month. Have never been able to jump that much with Intel. Been running a 5950X since shortly after launch and very happy overall. Looking forward to gen5 nvme and ddr5@6000.

Also despite e chores being good enough in a lot of tasks,. I'm running Linux and don't expect the scheduler updates to really solidify for a couple more years. Intel put a lot of effort with Microsoft for this model. And even then had to disable avx512… and now physically cutting it off. They broke standard capability reporting and it will take time. When AMF introduces their own little cores it'll probably be more solid in Linux.

I am really happy to see the dramatically improved performance and competition. The lower power and mobile space has gotten very interesting. I'm not using a 5900hx based mini PC as my home server... Lower power, more performance, yes please. And replacing my router with an Intel N6005 box

It's pretty great all around.

 

[Garrus]

That's not how silicon works. Different architecture works in its own frequencies and voltage to frequency curve. Nothing here is "defective".
Silicon isn't run on X frequency universally. E cores and P cores are fabricated entirely differently. They are not the same.

These "toy cores" and "defective cores" are built to be very silicon space efficient, and carry an IPC similar to what people met on Skylake \ Zen2

Zen 2 or Skylake, total nonsense. You bought the marketing. Anandtech tested E core performance exhaustively. The 12900K has 8 E cores and it often loses to 4 Skylake cores. So half the speed of Skylake or Zen 2 is more like it.

"Having a full eight E-cores compared to Skylake's 4C/8T arrangement helps in a lot of scenarios that are compute limited. When we move to more memory limited environments, or with cross-talk, then the E-cores are a bit more limited due to the cache structure and the long core-to-core latencies."

The issue is one single E core can appear to equal a Skylake core, all by itself. But when you run 8 of them together, the core performance doesn't scale right because of their limited design. Look at the MT tests and you'll get almost half the performance as what you expected.

E cores are small for a reason. Lacking cache, high latency, lacking hyper threading, don't expect much from 16 of them. CPUs are more than the raw compute core. I don't want any E cores at all, thank you.

CPU Benchmark Performance: E-Core - The Intel 12th Gen Core i9-12900K Review: Hybrid Performance Brings Hybrid Complexity (anandtech.com)

 

[Minus Infinity]

The 13700k is looking powerful. Faster than the mighty 5950x.

Well it is a 12900K with higher clocks, more cache and architectural tweaks.

AMD is going to suffer on marketing, becuase they are releasing the same processors just in Zen 4 form, but Intel has increased core counts effectively pushing everything up a tier. 7600X will be expected to compete against 13600K but now is a lower end cpu same with 7700X vs 13700K. And Intel is going to keep increasing e-core counts with Arrow Lake hitting 40 IIRC. AMD does not appear to have an answer to this unless they also go hybrid in Zen 5 and use Bergamo cores as e-cores. AMD luckily does have v-cache coming though.

While I have no real interest in Raptor Lake, and will repalce my old Zen 1700X system with a 7900X (hopefully with v-cache) come Arrow Lake I might be a lot more inclined to go that way if AMD keeps to current core counts. No way a regular 16 core 8950X would compete with a 32 core (8P + 24 E) 14900K say. There are rumours AMD is going hybrid with Zen 5 and IMO they have no other option. I don't having 32 full cores is wise in the desktop market especially from power use terms.

 

[A Computer Guy]

Well AMD also will use big.little in zen5 so that`s all will be an interesting confrontation

I think they should do big, medium, little.
4 big superfast cores, 4 medium duty all round cores, 8 little slow cores.
Then have a robust ability to pin software threads to particular core types.

 

[hs4]

Zen 2 or Skylake, total nonsense. You bought the marketing. Anandtech tested E core performance exhaustively. The 12900K has 8 E cores and it often loses to 4 Skylake cores. So half the speed of Skylake or Zen 2 is more like it.

"Having a full eight E-cores compared to Skylake's 4C/8T arrangement helps in a lot of scenarios that are compute limited. When we move to more memory limited environments, or with cross-talk, then the E-cores are a bit more limited due to the cache structure and the long core-to-core latencies."

The issue is one single E core can appear to equal a Skylake core, all by itself. But when you run 8 of them together, the core performance doesn't scale right because of their limited design. Look at the MT tests and you'll get almost half the performance as what you expected.

E cores are small for a reason. Lacking cache, high latency, lacking hyper threading, don't expect much from 16 of them. CPUs are more than the raw compute core. I don't want any E cores at all, thank you.

CPU Benchmark Performance: E-Core - The Intel 12th Gen Core i9-12900K Review: Hybrid Performance Brings Hybrid Complexity (anandtech.com)

I read the AnandTech article you indicated, among 20 MT tests
8xE core win: 12 tests
i7-6700K win: 3 tests
tie (difference within 5%): 5 tests
Contrary to your description, the E-core cluster appears to have better MT characteristics than Skylake 4C/8T.

I have done some tests myself comparing E and P cores, and it appears that the E core provides 1.5 times the power efficiency when both P and E cores are run at maximum clock, and the same MT performance in 2/3 the area when they are run at the same clock. These comparison are still true for the runtime benchmarks in productive applications.

By the way, @OneRaichu already showed difference in core-to-core latency matrix between ADL and RPL. The matrix indicates that the cache and latency problems seem to have eased considerably.

 

[A Computer Guy]

I read the AnandTech article you indicated, among 20 MT tests
8xE core win: 12 tests
i7-6700K win: 3 tests
tie (difference within 5%): 5 tests
Contrary to your description, the E-core cluster appears to have better MT characteristics than Skylake 4C/8T.

I have done some tests myself comparing E and P cores, and it appears that the E core provides 1.5 times the power efficiency when both P and E cores are run at maximum clock, and the same MT performance in 2/3 the area when they are run at the same clock. These comparison are still true for the runtime benchmarks in productive applications.

By the way, @OneRaichu already showed difference in core-to-core latency matrix between ADL and RPL. The matrix indicates that the cache and latency problems seem to have eased considerably.

I wonder if the OS can optimize better with the P/E cores since the are so different. For example putting all the low priority threads on the e-cores automatically so the p-cores are not bogged down by a bunch of low performance operations.

 

[Deleted member 185088]

Sub high end would be 7900x in my mind


Yeah well, you might dislike them but if you buy a high core count cpu, its probably for MT performance. And smaller cores WILL give you more MT performance while being more power efficient AND area efficient

Are they more efficient than AMD's cores? Not sure about that, on laptops AMD trashes Intel in battery life and performance (when on battery), probably those e-cores just like p-cores are pushed beyond their sweet spot.

 

[ncrs]

I wonder if the OS can optimize better with the P/E cores since the are so different. For example putting all the low priority threads on the e-cores automatically so the p-cores are not bogged down by a bunch of low performance operations.

That's what Intel Thread Director for Windows 11 does. Unfortunately there is no support for Windows 10 or earlier, and the support for Linux is still lacking.

 

[Garrus]

I read the AnandTech article you indicated, among 20 MT tests
8xE core win: 12 tests
i7-6700K win: 3 tests
tie (difference within 5%): 5 tests
Contrary to your description, the E-core cluster appears to have better MT characteristics than Skylake 4C/8T.

I have done some tests myself comparing E and P cores, and it appears that the E core provides 1.5 times the power efficiency when both P and E cores are run at maximum clock, and the same MT performance in 2/3 the area when they are run at the same clock. These comparison are still true for the runtime benchmarks in productive applications.

By the way, @OneRaichu already showed difference in core-to-core latency matrix between ADL and RPL. The matrix indicates that the cache and latency problems seem to have eased considerably.

Holy Moly. 8 E cores are about the same as 4 Skylake Cores as you admit. Before you were suggesting it was 4 = 4. What a moving of goalposts. And btw listing which one wins, when you win by a few percentage points is intentionally misleading. The fact that it loses in ANY tests, when it is 8 versus 4 cores, is the point. Half the speed.

Anyways I tried to help. Bizarre. If it was truly as fast as you say, the 12900k would have been much faster than it is, and you know it. I'm done. Don't expect, and don't lie to people, 16 E cores in Raptor lake will not be like 16 Skylake Cores. This is why there is no point in making comments, nobody actually cares what the truth is.

Try it: "I admit, it is about half the speed as I thought it was." "It takes 8 E cores, not 4, to consistently equal or beat 4 Skylake Cores". You're welcome.

 

[GreiverBlade]

Well it is a 12900K with higher clocks, more cache and architectural tweaks.

AMD is going to suffer on marketing, becuase they are releasing the same processors just in Zen 4 form, but Intel has increased core counts effectively pushing everything up a tier. 7600X will be expected to compete against 13600K but now is a lower end cpu same with 7700X vs 13700K. And Intel is going to keep increasing e-core counts with Arrow Lake hitting 40 IIRC. AMD does not appear to have an answer to this unless they also go hybrid in Zen 5 and use Bergamo cores as e-cores. AMD luckily does have v-cache coming though.

While I have no real interest in Raptor Lake, and will repalce my old Zen 1700X system with a 7900X (hopefully with v-cache) come Arrow Lake I might be a lot more inclined to go that way if AMD keeps to current core counts. No way a regular 16 core 8950X would compete with a 32 core (8P + 24 E) 14900K say. There are rumours AMD is going hybrid with Zen 5 and IMO they have no other option. I don't having 32 full cores is wise in the desktop market especially from power use terms.

Actually in my view, Intel needed to go hybrid to keep up with AMD (consumption, multi threaded), thus their "more core" is not automatically better, plus AMD only needed to add 3DV cache to the 5800X to compete with ADL previously

Reviews will always tell and for now, hybrid is not looking too bright for me

It would be hilarious if a CPU with so much efficient core count still ended up drawing more power than one with double "P" core count
(I.e.: 8P/24E Vs 16 )

With such drastic move from Intel, I expected way more, I guess I was a bit too enthusiastic, although the disappointment is not on the level of their ARC GPUs

 

[hs4]

Holy Moly. 8 E cores are about the same as 4 Skylake Cores as you admit. Before you were suggesting it was 4 = 4. What a moving of goalposts. And btw listing which one wins, when you win by a few percentage points is intentionally misleading. The fact that it loses in ANY tests, when it is 8 versus 4 cores, is the point. Half the speed.

Anyways I tried to help. Bizarre. If it was truly as fast as you say, the 12900k would have been much faster than it is, and you know it. I'm done. Don't expect, and don't lie to people, 16 E cores in Raptor lake will not be like 16 Skylake Cores. This is why there is no point in making comments, nobody actually cares what the truth is.

Try it: "I admit, it is about half the speed as I thought it was." "It takes 8 E cores, not 4, to consistently equal or beat 4 Skylake Cores". You're welcome.

Did you really read the article? In the article you showed, "12900K E-core", i.e. 8xE-core, were compared.

And have you never seen the Alder lake die photos, Intel would have distributed a lot of them in PR. Including cache and corresponding ring bus area, each P-core is about 10.7 mm² and each E-core is about 13.4 mm² in a cluster of four. 24 E-cores are placed in the same area of 8 P-cores. Intel itself treats a cluster of four E-cores as if they were a single P-core. Have you ever seen an illustration like this? (I have directly taken these from Intel Newsroom Press Kit: 12th Gen Intel Core)

In the 12900K, comparison between 4xE cores and 1xP core become:
- 25% more area
- Equal power consumption
- 50% more MT performance
The relationship is 1.2x area efficiency and 1.5x power efficiency.