• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Mhz for Mhz Comparison... 2013 tech v's Modern tech... who can join me?

pugheaven

pugheaven

pugheaven

Soldato
Joined
22 Apr 2008
Posts
3,914
Location
Bryn Celyn Wales
Hi Guys,

EDITED: I originally asked for 4.6ghz and multi cores, however, that makes no sense, but that would mean a lot of people won't be able to post so,

I've amended this to 4.0Ghz which allows the Ryzen crew in as well, so can any posts now be for 4.0Ghz CPU's please. Thanks guys, would really be good to see a mhz per mhz comparison on all the different chips variations... IT DOESN'T MATTER HOW MANY CORES NOW, SIMPLY AS WE NEED TO CONCENTRATE ON THE SINGLE CORE SCORE ANYWAY!

Using CPU-Z just as a benchmark... This is everyone who's contributed as of yet:

CPUZ SCORES @4Ghz (give or take error on bus)
E5 Xeon 1860 V2: Single Core = 411.8 (Pugheaven)
Ryzen7 3700X: Single Core = 481.7 (EdButler)
Ryzen5 2400G: Single Core = 455.8 (sandys)
Ryzen Threadripper 2970x: Single Core = 464.7 (sandys)
Ryzen5 3600: Single Core = 476.6 (Humbug)

4ghz_1.JPG


4ghz_2.JPG



The reason I'm interested in this now is the below video I watched which to me was a really interesting article, so rather than play games, would prefer to look at raw power and the easiest way I suppose is a quick benchmark that takes single and multi into consideration and everyone usually has the app... just be interesting to see if we've come anywhere in the last 7-8 years... I believe AMD have but I'm not convinced with Intel...

 
Last edited:
AMD is definitely faster except in the most Intel favourable code.
Zen2's IPC is better than in Intel's (still) "latest" Sky Lake and that's aging Ivy Bridge architecture CPU, whose performance has been nibbled also by countless vulnerability patches.

Of course you would now get crazy amount of cores at its price.
Here are 12 core 3900X results at stock speed.
On no doubt pretty crappy VRM board, which might limit clocks little and with single channel utter **** slow memory to seriously cripple Ryzen.
 
I would need to setup my chiller to get 4.6Ghz out of my 1800X lol. You'd probably need to run dual channel memory for a direct comparison.
 
Robert896r1 said:
@pugheaven here you go: https://valid.x86.fr/58tlfx

As another comparison point, here is my daily. Lots of oc headroom in the newer platform: https://valid.x86.fr/58tlfx
Click to expand...
nice speeds... I agree this is my problem in a way though in the fact that INtel seems to throw mhz at things now rather than genuinly increase the IPC and be innovative hence probably why AMD caught them up in a way clock for clock, hence why I wanted to see in that one for example is there any chance you could just do one run at 4600mhz for me mate on all cores just to see what yours comes out with? We know that thwe new platform can do 5.2ghz, but what I wanna see here is a like for like in mhz to just see what improvement of which there will be have been made in the last few years as I'm not convinced other than increasing mhz for Intel... seems like AMD is the only one that's actaully made what I'd call progress... as a lot of i7's ok they were 4 core but could hit 5ghz 8 years ago...

Would really be appreciated if you could as I'd love to see it mate. Cheers P.S Just for interest what's that run temp wise maxed out sday with OCCT or AIDA64 with FPU testing? I believe they run hot by all accounts?
 
pugheaven said:
I believe they run hot by all accounts?
Click to expand...
Intel's TDPs are done with "artistic freedom".
If cooling can cope with the heat, under full AVX load 8 core 9900K chasing advertised clocks is like 200W TDP CPU.
10 core 10900K? Yep, toward 250W.

For comparison AMD's 16 core is ~140W max, unless you change BIOS settings from stock.
 
EsaT said:
Intel's TDPs are done with "artistic freedom".
If cooling can cope with the heat, under full AVX load 8 core 9900K chasing advertised clocks is like 200W TDP CPU.
10 core 10900K? Yep, toward 250W.

For comparison AMD's 16 core is ~140W max, unless you change BIOS settings from stock.
Click to expand...
Yeah it's a lot, this old 22nm chip I have here is taking 215W on full load when doing FPU calculations... that's a lot of heat generated that...
 
kitfit1 said:
Not 4.6Ghz...........................at the moment. But will be when i put my SS on it at the weekend.

YSARFKR.jpg
Click to expand...

Thats' as near as damn it though mate, so that's a good one to compare to so,if we just took that figure for now is looking like the AMD over an old Xeon is around 17% faster give or take clock for clock for 1 CPU... just based on CPUZ which obviously isn't the be all and end all, but at least it's consistant in it's findings... I wonder if you could do one at as close to 4Ghz as possible mate just for me, as I've amended this thread to a 4Ghz to get more different CPU's in here to just get a ROH.

CPUZ SCORES @4Ghz
E5 Xeon 1860V2: Single Core = 411.8, Multi Thread = 4062.1
 
kitfit1 said:
It's a fair assumption you read little to nothing of Pugs OP ? He's not looking for you or anyone to post the biggest clock they can get, what he is asking is that we can all post 4.6Ghz clocks using CPU-Z benchmark........................................maybe read the post again :rolleyes:
Click to expand...

Actually, there's a continued problem with CPU Z that the same machine running CPU - Z validating multiple times in short order will replace the URL with the same exactly URL instead of creating a unique one. I did 4.6 before and then 5.2 so the 4.6 got replaced with 5.2

@pugheaven see above why both URL's ended up the same. Here's a re-run of 4.6 https://valid.x86.fr/8t6qgz
 
pugheaven said:
Thats' as near as damn it though mate, so that's a good one to compare to so,if we just took that figure for now is looking like the AMD over an old Xeon is around 17% faster give or take clock for clock for 1 CPU... just based on CPUZ which obviously isn't the be all and end all, but at least it's consistant in it's findings... I wonder if you could do one at as close to 4Ghz as possible mate just for me, as I've amended this thread to a 4Ghz to get more different CPU's in here to just get a ROH.

CPUZ SCORES @4Ghz
E5 Xeon 1860V2: Single Core = 411.8, Multi Thread = 4062.1
Click to expand...

Actually Pug, although CPU-Z said 4.565Ghz it was actually running at 4.08Ghz all core PBO enabled. Which as @Robert896r1 said, effectively gives a completely wrong impression of the real speed.
My view is that so everyone is singing from the same hymn sheet would be just to run CPU-Z bench but show HWiNFO64 in the same screenshot. Starting HWiNFO64 before CPU-Z, that way everyone can see the max clock and there would be no doubt at all.
 
OK my Ryzens at 4ghz both with similar memory timings though one a Ryzen (zen) and the other a TR (zen+) hard to directly compare as one is a pared down cache wise Zen and the other not, but looks very unimpressive for the refresh TR vs OG Ryzen (despite 2xxx code 2400G uses 14nm Zen)

EDIT - Ah just noted that you wanted 8c/16t, I don't have one of those, so will hide images in spoiler tags in case some one is interested anyway

Zen

zen4Ghz.png


Zen Plus

zenplus4-Ghz.png
 
Thanks for everyone so far we're now getting a feeling for single core performance now at 4Ghz for varying chips and it's quite interesting actually...

CPU-Z SCORES @4Ghz (give or take error on bus)
E5 Xeon 1860 V2: Single Core = 411.8 (Pugheaven)
Ryzen7 3700X: Single Core = 481.7 (EdButler)
Ryzen5 2400G: Single Core = 455.8 (sandys)
Ryzen Threadripper 2970x: Single Core = 464.7 (sandys)
Ryzen5 3600: Single Core = 476.6 (Humbug)

Would now love to see some Intel's as well and some more different Ryzens if possible... be nice to get everyone a single core score at 4ghz... however upto now... from fastest to slowest on a single core there is a maximum of what, circa 14% difference with mine being the slowest expectidly.
 
You must log in or register to reply here.
Back
Top Bottom