My new Intel 11900K @5500Mhz vs. old Intel 8700K @5300Mhz CPU performance test

[humbug]

Yes, definetely AMD did fantastic job at ZEN3 arch comparing to ZEN2. However, they are still behind Intel in this perspective. Below 5950x with 3800Mhz Ram 16-18-18-38 "CR1". My score is even at CR2.

The 11900K is around 55ns in this,

In Sandra Coffeelake is around 85ns Zen 3 around 80ns, the reason its different is because AIDA64 is measuring from the L3 to the IMC where as Sandra is measuing from the L3 to the IMC and back again, there is a small difference between Intel and Zen 3 in L3 to IMC Transfer Latency, its slightly faster on Intel one way but slightly faster on Zen 3 full round, it isn't enough to make any real difference.

Again this also has no effect on games. That is (Again) Intercore Latency, this is the time it takes for one CPU core to communicate with another, Intel use a Ring Bus for that, Zen 3 use something else, the time it takes for one core to communicate with another in Zen 2 is around 53ns, for Coffee Lake its around 52ns, slightly lower, in games Coffee Lake and Zen 2 are about = or Coffee Lake is a little faster, the clock speed difference makes Coffee Lake that much faster in reality.

For Zen 3 the time it takes for one core to comunicate with another is around 30ns, roughly 40% faster than Coffee Lake and Zen 2.

So, when you look at low resolution gaming slides, and i have lost of them, i'll give you one example below.... where the high frame rates depend on the very fast intercore comunication Zen 3 is streaks ahead of Zen 2 and to a lesser extent tho still significant also Coffee Lake.

 

[Toasty]

Running radiators in a bucket of ice or outside in winter temps makes sense for trying to hit records in overclocking and benching not so much for gaming where the % difference will be imperceptible to human eyes compared to 5800x with regular AIO.

 

[humbug]

Running radiators in a bucket of ice or outside in winter temps makes sense for trying to hit records in overclocking and benching not so much for gaming where the % difference will be imperceptible to human eyes compared to 5800x with regular AIO.

I was just trying to be funny

 

[Toasty]

I know but he said main reason he got was for gaming when essentially it’s been shown there’s nothing in it in actual gaming with Zen 3 only in synthetics and you don’t need your radiators outside with Zen 3.

 

[torwak]

I believe you are mixing the terms like optimized architecture for software and CPU computing power. RAM speed has nothing to do with L1-L2-L3 cache. Memory is the external fast storage to store data for CPU. Our subject is also has nothing to do with inter-core communication. I even didn't talk about mesh memory or ring bus arch, but somehow you got there. I am totaly lost in the translation.

If you like AMD stay with it and use it, but please don't give a speech about things you really don't know. Check the below, but don't get a heart attack ok?

 

[humbug]

I know but he said main reason he got was for gaming when essentially it’s been shown there’s nothing in it in actual gaming with Zen 3 only in synthetics and you don’t need your radiators outside with Zen 3.

Yea, i don't care what CPU people decide to get, and its not as if the 11900K is an objectivity bad CPU, it isn't, not really, but Its also not good and it shouldn't exist and it has a silly price, so if he wanted Intel IMO he would have been better off with a 10900K but heyho someone needs to buy them i guess.
I think the 10900K, or rather the 10850K is a rather good CPU.

What i don't like seeing is people reasoning their purchase by putting down the competition, especially when they get it so completely wrong.

I would do that only if I take my radiators to balcony at winter night. Otherwise any torture test will may cause degradation at the silicon quality becasue of the inefficient cooling.

I have bought this CPU for gaming especially for the low memory latency that AMD can't achieve. Not for rendering farm.

I have never referred to the intercore latency. It's a complete disaster for AMD either.

If you wanted the outright best gaming CPU, the 5800X is it.

 

[Destination]

Give it a cinebench r23 single and multi just the standard run rather than the 10 minutes thing.
Oddly, do you really feel that cooling will be insufficient to the point of degrading your silicon in just 10 minutes?

 

[Destination]

I believe you are mixing the terms like optimized architecture for software and CPU computing power. RAM speed has nothing to do with L1-L2-L3 cache. Memory is the external fast storage to store data for CPU. Our subject is also has nothing to do with inter-core communication. I even didn't talk about mesh memory or ring bus arch, but somehow you got there. I am totaly lost in the translation.

If you like AMD stay with it and use it, but please don't give a speech about things you really don't know. Check the below, but don't get a heart attack ok?

This is quite remarkable, sucha low score, it is actually the first time I've seen such low numbers generated on this series, this was said to be one of the weak points in the lower chips in the 11 series, as they couldn't match the 10 series in latency, but you've done well.

Out of interest, does this config show any actual improvement over say a gear 1 3600 speed when it comes to gaming, I always recall Steve Burke saying you need to test it in an application or game to see that it is having real world effect rather thna just seeing better figures reported which perf might actually become retrograde due to other factors?
I'd love to see a comparative run with some game demos.

 

[Grim5]

How much power does it pull at all core 5.5?

 

[McGraw]

Not at all. I have custom loop. The temperatures were around 80-85C at CB20. The power draw was a lot. I hit almost 290W.

How did you get such strong chips?

 

[humbug]

This is quite remarkable, sucha low score, it is actually the first time I've seen such low numbers generated on this series, this was said to be one of the weak points in the lower chips in the 11 series, as they couldn't match the 10 series in latency, but you've done well.

Out of interest, does this config show any actual improvement over say a gear 1 3600 speed when it comes to gaming, I always recall Steve Burke saying you need to test it in an application or game to see that it is having real world effect rather thna just seeing better figures reported which perf might actually become retrograde due to other factors?
I'd love to see a comparative run with some game demos.

Yea reviews have it at about 50 to 60ns, they, must all be wrong.

 

[Shac]

Hmmm they have the 5900X at 65.8 ns

However my 5950x will do 57.9 ns and probably tighter if i spent time optimising settings.

 

[Journey]

I have bought this CPU for gaming especially for the low memory latency that AMD can't achieve. Not for rendering farm.

Any chance you could show us some of the improvements in games from the 8700k, not just benchmarks, since you bought it for games not as a benchmarking system, right? Found any where you get more than a 10% improvement across the 1% low, and the average rates?

 

[Grim5]

This is quite remarkable, sucha low score, it is actually the first time I've seen such low numbers generated on this series, this was said to be one of the weak points in the lower chips in the 11 series, as they couldn't match the 10 series in latency, but you've done well.

Out of interest, does this config show any actual improvement over say a gear 1 3600 speed when it comes to gaming, I always recall Steve Burke saying you need to test it in an application or game to see that it is having real world effect rather thna just seeing better figures reported which perf might actually become retrograde due to other factors?
I'd love to see a comparative run with some game demos.

If the numbers are real and not photoshopped then his gaming performance should be 10-15% higher than any 11900k has reviewed

 

[Destination]

I just assumed it was a function of artificial figures due to gear 2. I believe that the numbers might work,I just want to see if they stack into actual benefit.
Again I reiterate that Steve from GN runs a test when he OCs to check that perf be it processor or memory, that it is actually increasing performance.

Thus my previous 2 posts asking for specific information.

 

[Deleted User 456458]

I just assumed it was a function of artificial figures due to gear 2. I believe that the numbers might work,I just want to see if they stack into actual benefit.
Again I reiterate that Steve from GN runs a test when he OCs to check that perf be it processor or memory, that it is actually increasing performance.

Thus my previous 2 posts asking for specific information.

You can find gaming oc’d results here: https://kingfaris.co.uk/blog/10900k-vs-11900k

All the settings and timings are also presented.

 

[Joxeon]

If the numbers are real and not photoshopped then his gaming performance should be 10-15% higher than any 11900k has reviewed

Yeah seems a lot higher than the samples the reviewers received or anyone else here has reported, maybe he's benching on LN2 or maybe he just got extremely lucky.

 

[Destination]

You can find gaming oc’d results here: https://kingfaris.co.uk/blog/10900k-vs-11900k

All the settings and timings are also presented.

Yes,but that is a very specific and indeed very odd set of testing criteria.
no one buys a chip, disables hyperthread and then plays games.
Also the results showed the new processor sucking dick.
I would prefer our poster shows what he has managed, as it seems spectacular.

 

[torwak]

This is quite remarkable, sucha low score, it is actually the first time I've seen such low numbers generated on this series, this was said to be one of the weak points in the lower chips in the 11 series, as they couldn't match the 10 series in latency, but you've done well.

Out of interest, does this config show any actual improvement over say a gear 1 3600 speed when it comes to gaming, I always recall Steve Burke saying you need to test it in an application or game to see that it is having real world effect rather thna just seeing better figures reported which perf might actually become retrograde due to other factors?
I'd love to see a comparative run with some game demos.

41.0ns latency is not from my system. CAPFRAMEX achieved it using PATRIOT VIPER GAMING DDR4-4800 C20 unlinked (1:2) at 11900K with APEX XIII. My own system score (43.1ns - Gear1 - Linked 1:1 for gaming) is lower below becasue I have budget GSKILL 3866 18-18-18-36 memory kit using it with 11900K and APEX XIII. I don't want to invest 500USD to DDR4 peformance kits to have additional 4-5FPS at games. Also Alder Lake / DDR5 will be available soon.

https://twitter.com/CapFrameX/status/1383784384403107849?s=20
https://twitter.com/CapFrameX/status/1383506937182969862?s=20

 

[torwak]

How did you get such strong chips?

I did nothing special. I have ordered it from the competitor the day it released.

Asus APEX XIII BIOS shows me "83" for the silicon prediction (SP). 83 is somewhere golden sample for 11900K.