Cinebench R15 Extreme Edition: Unofficial mod

humbug · 12 Feb 2019 at 08:45

kitfit1 said:
With the score between runs being that close Danny, a margin of error has to be taken into account. The thing that has struck me though is moorhen2's multi score of 599. Considering the cpu had to be run at 5.2Ghz to get that score, the score itself is very unimpressive. I'm 100% shure that if my 2700x could run the same bench at 5.2Ghz it would leave the 9900K in the dust.

Yeah, first gen Ryzen 6 core, £160 running at 3.85Ghz = 335 CB vs Pastymunchers 5Ghz 9600K scoring 325 CB.

I can't match his single core but that's mainly from lack of clock speed. still tho its impressive to see a 2 year old AMD lower end CPU beat Intel's latest £270 i5 in MT.

Can't wait to see what Ryzen 3600 can do.

pastymuncher · 12 Feb 2019 at 10:20

Like I have said many times before, multicore performance isn't very important to me because the games I play tend to be single threaded so pure grunt is more important to me. That is why I stuck with Intel and didn't switch to AMD. Not that it was much of a upgrade from my 4790k though. Anyway seeing is you have a extra 6 threads 10 points more than mine isn't very impressive. Disable your HT and post that result up. I will agree that for the price the performance and double the threads of mine is good.

Tee Hee Johnson · 12 Feb 2019 at 11:43

Danny75 said:
What about the single core run? Would be nice if everyone could show their single core result along with the multicore so we can see another example of how core strength compares between CPUs now. But be warned it can be like watching paint dry. I went to make a cup of tea and the 4770K was still at it when I came back.

I'll do a proper run tonight with screengrab of single and multi core.

humbug · 12 Feb 2019 at 15:32

pastymuncher said:
Like I have said many times before, multicore performance isn't very important to me because the games I play tend to be single threaded so pure grunt is more important to me. That is why I stuck with Intel and didn't switch to AMD. Not that it was much of a upgrade from my 4790k though. Anyway seeing is you have a extra 6 threads 10 points more than mine isn't very impressive. Disable your HT and post that result up. I will agree that for the price the performance and double the threads of mine is good.

Disable HT? that's like me telling you to run your CPU at the same clock speed as mine, mine would humiliate yours. of course the point is its an equally ridiculous argument and i acknowledged your single threaded performance over mine.

It is what it is and yes when its an older generation CPU costing at the time over £100 less than your newer CPU costs now.... yet still a more powerful CPU, it is damned impressive.

Gaming, i find its a pretty solid match for the 1070, i'm not going to see higher performance with a 9600K, even with an RTX 2080 there is only a <10% difference between the 9600K and a 2600X, tho the 2600X is faster than my 1600 i'm not going to see any difference given the RTX 2080 is twice as fast an my 1070.

You also have a 1070.

Its all relative

kitfit1 · 12 Feb 2019 at 17:47

pastymuncher said:
Disable your HT and post that result up. I will agree that for the price the performance and double the threads of mine is good.

Got home from work and read your post pasty, my heating has been on all day so PBO only booted to 4.23Ghz instead of 4.25Ghz as it normally does. Turned SMT off, as you can see.

moorhen2 · 12 Feb 2019 at 17:47

kitfit1 said:
With the score between runs being that close Danny, a margin of error has to be taken into account. The thing that has struck me though is moorhen2's multi score of 599. Considering the cpu had to be run at 5.2Ghz to get that score, the score itself is very unimpressive. I'm 100% shure that if my 2700x could run the same bench at 5.2Ghz it would leave the 9900K in the dust.

That is a very bold statement to make. Like for like there is only 21 points difference in multi core when I run my 9900k at the same frequency as your 2700x, and a higher single core score. So saying your cpu at 5.2ghz would leave a 9900k in the dust is a bit strong to be honest.

I don't think the OP wanted this to be a contest, just wanted comparisons I think.

humbug · 12 Feb 2019 at 17:55

moorhen2 said:
That is a very bold statement to make. Like for like there is only 20 points difference in multi core when I run my 9900k at the same frequency as your 2700x, and a higher single core score. So saying your cpu at 5.2ghz would leave a 9900k in the dust is a bit strong to be honest.

I don't think the OP wanted this to be a contest, just wanted comparisons I think.

9900K at 4.7Ghz, what is thought to be the new 3600X running at an unknown not final clock frequency. but thought to be around 4.3Ghz to 4.6Ghz all core boost.

3600X engendering sample 2057 Cinebench points
9900K 4.7Ghz all core boost 2040 Cinebench points

https://youtu.be/Jc7lqFaTvPo?t=186

Rumored Roadmap.

kitfit1 · 12 Feb 2019 at 18:05

@humbug, there really is no need to make this another AMDvIntel thread. As moorhen2 has said it's not a contest.

The reason i said what i said moorhen2 is that i have run my 2700X under both phase change and LN2 in the not too distant past, i know for a fact that when it gets over 4.8Ghz it does indeed leave a 9900K in the dust. But off course it isn't under phase or LN2 at the moment.

humbug · 12 Feb 2019 at 18:21

kitfit1 said:
@humbug, there really is no need to make this another AMDvIntel thread. As moorhen2 has said it's not a contest.

The reason i said what i said moorhen2 is that i have run my 2700X under both phase change and LN2 in the not too distant past, i know for a fact that when it gets over 4.8Ghz it does indeed leave a 9900K in the dust. But off course it isn't under phase or LN2 at the moment.

Its a Cinebench thread, and like it or not the CPU market is a duopoly and if this thread was put here just to say "hay look at this piece of software" and nothing more then its in the wrong room, it should be in here. otherwise given that its in the CPU room its about CPU's, if people can't handle debates about that Duopoly cropping up in it then perhaps those people would do well not reading so as not to be offended by it.

kitfit1 · 12 Feb 2019 at 18:27

humbug said:
Its a Cinebench thread, and like it or not the CPU market is a duopoly and if this thread was put here just to say "hay look at this piece of software" and nothing more then its in the wrong room, it should be in here. otherwise given that its in the CPU room its about CPU's, if people can't handle debates about that Duopoly cropping up in it then perhaps those people would do well not reading so as not to be offended by it.

I don't dissagree with your sentiment. But why post stuff about CPU's that are not only not confirmed yet, but arn't even released yet ? It's pointless because NONE of us on here could post a benchmark of any sort from any CPU in that list.

humbug · 12 Feb 2019 at 18:32

kitfit1 said:
I don't dissagree with your sentiment. But why post stuff about CPU's that are not only not confirmed yet, but arn't even released yet ? It's pointless because NONE of us on here could post a benchmark of any sort from any CPU in that list.

Because in a debate about performance i think its relevant, its whats coming up and soon, nothing wrong with with looking forward.

Pooh · 12 Feb 2019 at 20:43

kitfit1 · 12 Feb 2019 at 21:02

Pooh said:

WOW

That's just 6 more than my 2700X single core and i was at 4.35Ghz.........................................that is well within any margin of error. Peeps really should keep posting single core results as well, thanks Pooh.

Kei · 12 Feb 2019 at 22:36

46 single core score Threadripper 1920x @ 4.15GHz

iakhtar · 12 Feb 2019 at 23:07

8086k @5.2
Ram 4000mhz

Tee Hee Johnson · 12 Feb 2019 at 23:26

My 8700K at 5.1Ghz, 3200Mhz RAM.

Single core took a while lol

Apone70 · 13 Feb 2019 at 12:36

418 down from 1644 with my 8700k at 5ghz

Apone70 · 13 Feb 2019 at 12:52

single core 57 @5ghz

Phil2008 · 13 Feb 2019 at 23:18

Side by side results with a "4gb R9 290 gpu" and a "[email protected] cpu"

Distracted · 14 Feb 2019 at 07:51

The test is better in terms of the time it takes stressing the CPU to get a more accurate reading of its potential. But the results are too close together to draw meaningful comparisons imho.

Right now 1 point represents far too much change and the score could be increased by an order of magnitude to deliver something of value in terms of results.
Saying this is obviously easier than doing it, but this is just my opinion.