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

humbug · 29 Dec 2016 at 13:58

Silent_Scone said:
As above, but if you're having to ask that, you're missing the point.

Under 3 seconds difference in Blender with two cores less, in an ideal world the 6850K shouldn't be getting that close...Zen needs to overclock like a gem if it's got a hope in hell's chance of beating out BWE, if prices don't reflect.

It's the perfect example if you understand how heavily threaded Blender is...when most real world applications simply aren't.
Over 75% of CPUs tested by ASUS were capable of 4.3 or higher.

We already know the 8 core 16 Thread Intel 6900K is slightly slower than Zen Clock for clock in the same benchmark, so why try to confuse the whole thing by comparing apples to oranges? you have a 30% overclock on that Intel 6 core and its still slower.

What's 6 cores +30%?

~>Dg<~ said:
5820k ocd and a 6800K ocd can get the same results.or very similar.if its quicker in blender than the above two chips and slower at the rest its still a fail even if priced the same.if its slightly cheaper but same results well we had this power for a good while already.

so pricing is key.

Are you two completely loosing your marbles? Zen overclocks too, from what we know so far a lot higher than Sky Lake.

Thats desperate now... no one reading this thread is this stupid.

Silent_Scone · 29 Dec 2016 at 14:11

humbug said:
We already know the 8 core 16 Thread Intel 6900K is slightly slower than Zen Clock for clock in the same benchmark, so why try to confuse the whole thing by comparing apples to oranges? you have a 30% overclock on that Intel 6 core and its still slower.

What's 6 cores +30%?

Actually if you look closely, I ran the benchmark on the same build that AMD did, and it's slightly faster.

(6900K at 3.4 is 35.16 - 150 in Blender 2.77)

The point is, zen needs to have some range...like I keep saying, but the hardcore fanatics keep putting their walls up.

You put a demo up to awe people and then do it without boost showing the Intel counterpart running below it's stock boost clock, everyone should be asking for more than that.

humbug said:
Are you two completely loosing your marbles? Zen overclocks too, from what we know so far a lot higher than Sky Lake.

That's really not the case, sorry. I guess like with most of the hype residing within this forum, you'll have to wait and see for yourself.

humbug · 29 Dec 2016 at 14:20

Silent_Scone said:
Actually if you look closely, I ran the benchmark on the same build that AMD did, and it's slightly faster...

The point is, zen needs to have some range...like I keep saying, but the hardcore fanatics keep putting their walls up.

You put a demo up to awe people and then do it without boost showing the Intel counterpart running below it's stock boost clock, everyone should be asking for more than that.

Clock for clock core for core Zen is slightly faster than Broadwell, you cannot disprove that by using less cores very overclocked, core scaling, clock scaling it doesn't work like your trying to make out it does.
The only way you can get an accurate comparison is to use the same number of cores, thats what the original benchmark is and the result is Zen is slightly faster, if you want to disprove it go out and spend £1000 on a 6900K, others have and found the results to be correct.

Your apples to oranges comparison does not stack up to known apples for apples comparisons, of course it doesn't, nice try tho.

Silent_Scone · 29 Dec 2016 at 14:28

humbug said:
Clock for clock core for core Zen is slightly faster than Broadwell, you cannot disprove that by using less cores very overclocked, core scaling, clock scaling it doesn't work like your trying to make out it does.
The only way you can get an accurate comparison is to use the same number of cores, thats what the original benchmark is and the result is Zen is slightly faster, if you want to disprove it go out and spend £1000 on a 6900K, others have and found the results to be correct.

Your apples to oranges comparison does not stack up to known apples for apples comparisons, of course they don't, nice try tho.

Blender 2.77, same build AMD used during the demonstration at render sample 150. Faster.

Point of fact, we need to see Zen overclocking results.

humbug · 29 Dec 2016 at 14:32

Silent_Scone said:
Blender 2.77, same build AMD used during the demonstration at render sample 150. Faster.

Point of fact, we need to see Zen overclocking results.

Zen result was just under 36 Sec, thats just over 35 Sec, 1 to 2% margin of error.

Silent_Scone · 29 Dec 2016 at 14:36

humbug said:
Clock for clock core for core Zen is slightly faster than Broadwell, you cannot disprove that by using less cores very overclocked, core scaling, clock scaling it doesn't work like your trying to make out it does.
The only way you can get an accurate comparison is to use the same number of cores, thats what the original benchmark is and the result is Zen is slightly faster, if you want to disprove it go out and spend £1000 on a 6900K, others have and found the results to be correct.

Your apples to oranges comparison does not stack up to known apples for apples comparisons, of course it doesn't, nice try tho.

humbug said:
Zen result was just under 36 Sec, thats just over 35 Sec, 1 to 2% margin of error.

People have been running 2.78a which is consistently slower for me between 39 and 40 seconds. In cases like this, these details matter.

SiDeards73 · 29 Dec 2016 at 14:39

I see Scone managed to find more straws.... That result as Humbug says is within normal margins of error, nice try though.

Your really trying hard to downvote AMD aren't you? Wonder why?

humbug · 29 Dec 2016 at 14:42

Zen is already brining them out of hibernation ^^^^ 2017 is going to be a lot of fun.

Silent_Scone said:
People have been running 2.78a which is consistently slower for me between 39 and 40 seconds. In cases like this, these details matter.

Thats there problem and its got nothing to do with this, those who used the same version of Blander in the demo got the same results.

Silent_Scone · 29 Dec 2016 at 14:42

I ran it multiple times. You call them straws, I call them facts. Something people looking for a bargain tend to ignore

.

Like the entire point I'm making, we need Zen overclocking results. But you can't defend that, as you haven't seen anything...there's the issue.

humbug · 29 Dec 2016 at 14:44

Silent_Scone said:
I ran it multiple times. You call them straws, I call them facts. Something people looking for a bargain tend to ignore

You don't have a 6900K you ran it on a 30% overclocked 6800K and still lost.

Silent_Scone · 29 Dec 2016 at 14:47

humbug said:
You don't have a 6900K you ran it on a 30% overclocked 6800K and still lost.

Look at the screenshot in my previous post, 6900K @ 3.4.

6850K in the same version, 2.77

http://i.imgur.com/zLkAcQq.jpg

"Thanks Scone, I guess we just need to wait and see much leg room there is with Zen."

"Liar, margin of error. Who cares if Zen overclocks it's faster than BWE"

lol, no hope! Bye.

humbug · 29 Dec 2016 at 14:50

Silent_Scone said:
Look at the screenshot in my previous post, 6900K @ 3.4.

6850K in the same version, 2.77

http://i.imgur.com/zLkAcQq.jpg

Yeah, as i said thats a 30% overclocked 6800K, how you think you can prove anything with the wrong CPU is beyond me, its surreal.

Silent_Scone · 29 Dec 2016 at 14:57

humbug said:
Yeah, as i said thats a 30% overclocked 6800K, how you think you can prove anything with the wrong CPU is beyond me, its surreal.

2.77 is consistently quicker, that's all I'm saying. Digest it...

Good luck. Take your time

Rainmaker · 29 Dec 2016 at 15:00

Silent_Scone said:
People have been running 2.78a which is consistently slower for me between 39 and 40 seconds. In cases like this, these details matter.

I've noticed that too. Certainly PC World published a 'how to' guide for the benchmark, then (un)helpfully linked to 2.78a, so that's what a lot of people have done.

Even with v2.77 my FX8350 is loltastic slow. Ryzen will be a tasty upgrade provided the price is right, else I'll be grabbing an i7 or Xeon. I don't care what colour the badge is, I just want the right chip for the job at the price.

humbug · 29 Dec 2016 at 15:04

Silent_Scone said:
2.77 is consistently quicker, that's all I'm saying. Digest it...

Good luck. Take your time

RyZen vs 6900K was run on 2.77, your run there with the overclocked 6850K is also with 2.77, its all on 2.77 so the fact that its quicker than 2.78 is meaningless, its all done on 2.77.

Are YOU looking at your own picture? again... your run there is on 2.77.

whyscotty · 29 Dec 2016 at 15:07

2.78a is slower

humbug · 29 Dec 2016 at 15:10

humbug said:
RyZen vs 6900K was run on 2.77, your run there with the overclocked 6850K is also with 2.77, its all on 2.77 so the fact that its quicker than 2.78 is meaningless, its all done on 2.77.

Are YOU looking at your own picture? again... your run there is on 2.77.

here i'll paint it out for you.....

whyscotty · 29 Dec 2016 at 15:10

Silent_Scone · 29 Dec 2016 at 15:13

Thanks, Neil. Yep seems 2.77 is consistently quicker. Meaningless though apparently. 3.3 too.

Rainmaker · 29 Dec 2016 at 15:15

I'm honestly not sure if I'm missing something in this. What's the significance of the fact an i7 6900K can approximate the same time as Ryzen 8/16? All that tells me is that Ryzen appears to have Broadwell-E IPC (even without turbo, which I assume everyone here left enabled on their 6900s?). That being the case, if Ryzen is less than £1,000 then it's winner winner chicken dinner?

I may have taken the wrong end of the stick but some of the posts seem to be 'Ah but look, my 6900K gets the same(ish) time as Ryzen'. Well, yeah... but it's a £1,000 CPU, and shouldn't it rather be 'Holy crap AMD are finally matching Intel's top desktop chip, and at 90W TDP instead of 140W'? As I said, I may be missing the point.

Competitor rules

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

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