Intel 12th-Gen Core Alder Lake Architectural Benchmark

[Steve]

As a follow up to our feature on how CPU cores and cache impact gaming performance, we're testing different Intel 12th-gen Core Alder Lake CPUs configurations to see how fast those P-Cores and E-Cores really are.

Read the full article here.

https://www.techspot.com/review/2374-intel-alder-lake-architecture/

 

[deathtrap]

Thank you for that great piece of article.
I agree to everything you wrote 100%.
I find it sad, that there won´t be a pure 8 P-Core version, because now I need to gamble how long a 12600 non-K will be usable in modern games, or if I wait till Zen4 or Zen3D; if that stacked product ever happens below $300.

 

[Geralt]

So, the hybrid architecture sucks for gamers. Thank God AMD is still not implementing that architecture. I prefer full P cores always. But Intel had to invent this hybrid crap bringing problems with DRM sometimes. It's a company with a bad soul.

 

[Nintenboy01]

Wonder if Intel will rethink this approach in the future or just limit it to laptops in future architectures. But 8 cores/16 threads will probably be sufficient until the next console gen except for bad PC ports.

 

[elementalSG]

Interesting article with mixing up the P and E cores and benchmarking their performances. Will there also be some investigation into Power Consumption of P vs E cores?

 

[Michael7]

Very good read. So it looks it's worth to wait for the non K series with no E cores. It's a shame they don't make Alder Lake with just 8 P cores and while the hybrid architecture might less appealing to a PC use case, gamers especially, it will no doubt make a big difference in mobile/laptop segment but for that we have to wait a few more months to see the tests.

 

[Puiu]

So, the hybrid architecture sucks for gamers. Thank God AMD is still not implementing that architecture. I prefer full P cores always. But Intel had to invent this hybrid crap bringing problems with DRM sometimes. It's a company with a bad soul.

AMD will do something similar with Zen 5 + Zen4D in 2023 (2nd AM5 CPU generation). But the Zen4D cores are expected to be much more powerful than the E cores Intel is using. I hope we'll see 16 Zen5 + 16-32 Zen4D cores.

 

[Geralt]

AMD will do something similar with Zen 5 + Zen4D in 2023 (2nd AM5 CPU generation). But the Zen4D cores are expected to be much more powerful than the E cores Intel is using. I hope we'll see 16 Zen5 + 16-32 Zen4D cores.

OMG, the crap is contagious! OK, if those E cores will be more powerful, I have hopes anyway.

 

[Vulcanproject]

Ahhh I remember when AdoredTV essentially said that the FX series would age better and the gap would close against old i5 and i7 2nd gen parts, here the 2600k. The lolz.

The incredible thing is the 2600k is obviously very slow but still (kind of) viable at 1080p nearly eleven years later. Especially seeing as though you can dump another 20 percent clock speed on it quite easily.

 

[maxxcool7421]

I just want to see how a 5900x with 3d stacks of huge as& v-cache compares .. I do a ton of Vm work AND gaming on the same rig (4ghz 1950x) and am deeply interested in seeing 12 cores @ 4.8ghz all cores with a shovel full of on die cache.

 

[Ludak021]

If we look at first Ryzen here (1800X, year 2017) at 4c it's a lot worse than Skylake at 4c (year 2015).

 

[theruck]

Any tests for the iGPU planned?

 

[HardReset]

If we look at first Ryzen here (1800X, year 2017) at 4c it's a lot worse than Skylake at 4c (year 2015).

It might be 2+2 or 4+0, and that could affect quite a lot. Also it was not designed to be 4 core part, like 7700K was.

Article is pretty good but again, FX-8350 BE has 8 cores, not 4. So clear fact there's nothing to discuss. Otherwise we could also say 12900K is only 8 core because we don't count those 8 efficiency cores for some reason

 

[Shadowboxer]

So E cores clearly aren’t for gamers. But these Alder lake CPUs are clearly the best choice for gamers right now. We won’t see games using more than 8 P cores at full load for a long long time. BF2042 uses 45% of my 5800X in 128 player mode, so it’s got ample headroom and that’s by far the most demanding gaming load I’ve put on it.

 

[Shadowboxer]

If we look at first Ryzen here (1800X, year 2017) at 4c it's a lot worse than Skylake at 4c (year 2015).

It’s also worse than 2013s Haswell, Ryzen 1xxx was a very poor choice for gamers. But in all fairness Ryzen has savagely improved since then

 

[maroon1]

I wish if you did 8 E core vs 8 P core (not 4 cores vs 4 cores)

E cores does not support HT and we know a lot modern games don't work well with 4 threads only

If you look at 9th gen i3 and 10th gen i3 gaming benchmarks you will see massive difference in gaming because 9th gen core i3 did not support HT........... 4 cores/4threads is not enough for gaming anymore

It’s also worse than 2013s Haswell, Ryzen 1xxx was a very poor choice for gamers. But in all fairness Ryzen has savagely improved since then

Keep in mind that all the CPU that review tested have SMT (4 cores/8threads), except for FX8350 and E cores

He should either used i5 7400 in the test or turned off HT...... or he could compared 8 cores vs 8 cores ..... We all know that (4 cores/4threads) is not good for gaming anymore.....

In battlefield 5, Core i3 10100 has 50% higher 1% lows than core i3 9100F

https://static.techspot.com/articles-info/2033/bench/BFV_1080p-p.webp

And also 40% higher lows in Tomb raider

https://static.techspot.com/articles-info/2033/bench/SotTR_1440p-p.webp

I'm 100% sure that very poor 1% lows on E cores is because it is only 4 cores/4threads....

 

[maxxcool7421]

It might be 2+2 or 4+0, and that could affect quite a lot. Also it was not designed to be 4 core part, like 7700K was.

Article is pretty good but again, FX-8350 BE has 8 cores, not 4. So clear fact there's nothing to discuss. Otherwise we could also say 12900K is only 8 core because we don't count those 8 efficiency cores for some reason

FX was an ''enhanced 4 core'' cpu with a FPU that had markedly lower output. Even When Excavator shipped you could not run all the shared rescources at 8x8 execution levels.. which is why in most use cases outside Server-integer tasks a simple 4 core intel chip could still win out .. and in some cases I3 cpus would win.

anandtech.com/show/8316/amds-5-ghz-turbo-cpu-in-retail-the-fx9590-and-asrock-990fx-extreme9-review/8

anandtech.com/show/14804/amd-settlement

I love AMD .. I have AMD in every rig (4x rigs) but my media box .. but Excavator was never a true 8 core cpu.

 

[HardReset]

FX was an ''enhanced 4 core'' cpu with a FPU that had markedly lower output. Even When Excavator shipped you could not run all the shared rescources at 8x8 execution levels.. which is why in most use cases outside Server-integer tasks a simple 4 core intel chip could still win out .. and in some cases I3 cpus would win.

FX was 8 core CPU. You can run 8 tasks all assigned to different cores. Something you cannot do with 4 core CPU. Not even with HT. That's definite proof that FX-8350 is 8 core CPU.

On low loads yes but tbh quad cores are quite useless for even low level multitasking.

anandtech.com/show/8316/amds-5-ghz-turbo-cpu-in-retail-the-fx9590-and-asrock-990fx-extreme9-review/8

anandtech.com/show/14804/amd-settlement

I love AMD .. I have AMD in every rig (4x rigs) but my media box .. but Excavator was never a true 8 core cpu.

What about that settlement? AMD just decided to pay small amount of money to avoid more expensive court expenses. It doesn't prove anything.

Of course, there were no 8-core Excavator models ever released.

 

[Ludak021]

It might be 2+2 or 4+0, and that could affect quite a lot. Also it was not designed to be 4 core part, like 7700K was.

Article is pretty good but again, FX-8350 BE has 8 cores, not 4. So clear fact there's nothing to discuss. Otherwise we could also say 12900K is only 8 core because we don't count those 8 efficiency cores for some reason

Well, I was talking about the comparison in the article itself, 6700 skylake 4c and 1800x (set to 4c). In this configuration, as artcle itself presents, 1800x, set to use 4c, is comparable to haswell.
I said nothing more and nothing less. Just a comment made after looking at the article. It's an observation.

 

[hahahanoobs]

Good thing E-cores were never meant for gaming on their own! There is a specific reason why it's called a hybrid design. You don't disable the combustion engine in a Prius to run pure electric, right? You run both to get the most.

This is a curiosity test. Nothing more.

 

[dad0ts]

How you switched all p-cores off for e-core only testing? This is not possible to do in bios setup. Also are e-cores always stable when oc to 4.2 ghz? Their stock is only 3.9 ghz max. Also e cores have no HT, so 1.5x-2x perf. drop comes only from this, even i7 2600k has HT, as you not switched smt off. So this is comparable only to amd FX with 4 threads(which is usually not enough for modern games), all other cpus had 8 threads. So core-to-core latency is not the main factor affecting poor "e-core only" performance here.

 

[dad0ts]

I just want to see how a 5900x with 3d stacks of huge as& v-cache compares ..

5900x/5950x has extra 32 mb L3 cache on a separate die, so it is hit by core-to-core latency when used from another die cores. It has not 64 mb L3, but 32+32 mb L3. So the best amd zen3 for games is 5800x: all 8 cores and 32 MB L3 cache are on the same die within the same CCX, latencies are consistent and minimal.

 

[maxxcool7421]

FX was 8 core CPU. You can run 8 tasks all assigned to different cores. Something you cannot do with 4 core CPU. Not even with HT. That's definite proof that FX-8350 is 8 core CPU.

On low loads yes but tbh quad cores are quite useless for even low level multitasking.

What about that settlement? AMD just decided to pay small amount of money to avoid more expensive court expenses. It doesn't prove anything.

Of course, there were no 8-core Excavator models ever released.

https://regmedia.co.uk/2019/08/27/amd-eight-core-settlement.pdf

//shrug// that looks like fault admittance to me .. "here is money to go shut up because you are right in that the shared decoders + shared scheduler and only four classic FPUS prevent sustained true 8x8 execution except for specific circumstances."

 

[AKBrian]

It's the right conclusion, but for different reasons. Core latency isn't the only killer. L3 cache access latency is, as well. The E- cores have no dedicated L3; they share the P- cores' L3 cache, so it's an effective "off core" traversal each time.

 

[Strawman]

So, the hybrid architecture sucks for gamers. Thank God AMD is still not implementing that architecture. I prefer full P cores always. But Intel had to invent this hybrid crap bringing problems with DRM sometimes. It's a company with a bad soul.

You are absolutely right, this hybrid architecture sucks so much in gaming it actually tops the charts in most games. Winning by a huge margin in some of them as well. You hit the nail on the goddamn head my man