• 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.

*** AMD "Zen" thread (inc AM4/APU discussion) ***

Please correct me if I can wrong. So we are looking at:

Ryzen 6C12tT at 65w TDP with 3.3Ghz base with 3.7Ghz boost scoring 1889/12544.

Vs

Intel 6850K 6C12T at 140w TDP with 3.6Ghz base with 3.8Ghz boost and 4Ghz single core boost scoring 1940/11170.

So the Ryzen CPU single core performance is 3% slower and multicore performance is 12% faster at a lower frequency, with 54% less power consumption and roughly 50% of the cost.

Is that not too good to be true?

Actually given what we know of the CPU design and the core configuration, I am not surprised that multithreading will be better with the AMD CPUs.
As for the price, given what know, then it looks like Intel was just overcharging us. And for that alone, we should boycott them from now on in principle.
 

Ty, took your numbers there and ran some calculations.
I see your CPU running at 4.6GHz, assuming that is all core turbo and CPU-Z scales linearly with clock speed (easy on the numbers)
Also assuming the Ryzen cpu does 3.7GHz all core.
Ryzen:
Single thread: 1888/3.7 ≈ 510
Multi-thread: 12544/3.7 ≈ 3390

6850k @ 4.6GHz:
Single thread: 2190/4.6 ≈ 476
Multi-thread: 14242/4.6 ≈ 3096

CPU-Z reference stock 6850k:
Single thread: 1940/4.0 = 485
Multi-thread: 11170/3.8 ≈ 2939

∴ Clock for clock the 6 core Ryzen unknown RAM speeds and timing is 5-7% faster than the 6850k on single threaded and 9.5-15% on multithreaded.
This means if the 6 core can get to 4.3 GHz it would have approximately equal performance in this test based on my assumptions and calculations. That sounds doable.
 
Last edited:
The CPC CPU had slightly buggy SMT and was running an AMD Fury X before the ReLive drivers,so was probably a worse case scenario,especially with some of the games they used too.

But even at Haswell level IPC for gaming,the Intel £100 to £250 range is going to look really wonky especially if there are 4C and 4C/8T Ryzen CPUs under £200.

Intel's pricing structure is about to be destroyed. They will have to do a complete rethink.
 
Ty, took your numbers there and ran some calculations.
I see your CPU running at 4.6GHz, assuming that is all core turbo and CPU-Z scales linearly with clock speed (easy on the numbers)
Also assuming the Ryzen cpu does 3.7GHz all core.
Ryzen:
Single thread: 1888/3.7 ≈ 510
Multi-thread: 12544/3.7 ≈ 3390

6850k @ 4.6GHz:
Single thread: 2190/4.6 ≈ 476
Multi-thread: 14242/4.6 ≈ 3096

∴ Clock for clock the 6 core Ryzen unknown RAM speeds and timing is 7% faster than the 6850k on single threaded and 9.5% on multithreaded.
This means if the 6 core can get to 4.3 GHz it would have approximately equal performance in this test based on my assumptions and calculations. That sounds doable.

You can check the 6850K reference speed in CPU-Z. lol
 
Ty, took your numbers there and ran some calculations.
I see your CPU running at 4.6GHz, assuming that is all core turbo and CPU-Z scales linearly with clock speed (easy on the numbers)
Also assuming the Ryzen cpu does 3.7GHz all core.
Ryzen:
Single thread: 1888/3.7 ≈ 510
Multi-thread: 12544/3.7 ≈ 3390

6850k @ 4.6GHz:
Single thread: 2190/4.6 ≈ 476
Multi-thread: 14242/4.6 ≈ 3096

∴ Clock for clock the 6 core Ryzen unknown RAM speeds and timing is 7% faster than the 6850k on single threaded and 9.5% on multithreaded.
This means if the 6 core can get to 4.3 GHz it would have approximately equal performance in this test based on my assumptions and calculations. That sounds doable.


In the image it shows 3.3Ghz as base clock(what task manager identifies it as) then the two other numbers show 3.45 and 3.56Ghz, so I'd bet that the turbo with all cores loaded was more like 3.5Ghz than 3.7Ghz, which would improve the Ryzen numbers a little. But even at the max realistic speed of 3.7Ghz, those numbers are great even before you factor in price, which makes them look epic.
 
You can check the 6850K reference speed in CPU-Z. lol


Well, I had an old version of CPU-Z and was unaware the 6850k was on the newer version and sorry Scramz I must have skimmed your post too quick... :(
Oh well, we've got a range now and 2 data points! :p

Edit: Updated my post to include the CPU-Z reference 6805k and the numbers got even better!
 
Well, I had an old version of CPU-Z and was unaware the 6850k was on the newer version and sorry Scramz I must have skimmed your post too quick... :(
Oh well, we've got a range now and 2 data points! :p

Edit: Updated my post to include the CPU-Z reference 6805k and the numbers got even better!
Either way, Ryzen is starting to look very impressive, let hope it is not all a big jokoe.
 
Please correct me if I can wrong. So we are looking at:

Ryzen 6C12tT at 65w TDP with 3.3Ghz base with 3.7Ghz boost scoring 1889/12544.

Vs

Intel 6850K 6C12T at 140w TDP with 3.6Ghz base with 3.8Ghz boost and 4Ghz single core boost scoring 1940/11170.

So the Ryzen CPU single core performance is 3% slower and multicore performance is 12% faster at a lower frequency, with 54% less power consumption and roughly 50% of the cost.

Is that not too good to be true?

Maybe too good to be true.

On the other hand, Intel have made pathetic progress over the last ~5 years. If we're looking at what we're used to historically. Somehow they've only gained some 20% IPC and 2.5x the core count since going from 32nm to 14nm.

But we'll find out soon enough.
 
I fixed that for you.....

MOAR_MHZ.jpg

Hah, great stuff.

Intel in full Netburst mode.
 
Just for everyone to reference a reasonably clocked i7-2700K (my own, frequency in sig):

cpu-z2700k_zpsr6nrfuix.png


Given those Ryzen results weren't even a high spec CPU, were 65W and only stock from Ryzen, a R7 1700X should be a decent improvement!

Sandy is finally about to become worth replacing :D
 
Last edited:
Back
Top Bottom