AMD Presents New Horizon

Ieldra said:
Well I just ran it with samples at 100 and I got 26s with six cores @ 4350MHz which seems more less right to me. I'm clocked 26% higher and I have 25% fewer cores.

Edit:

If Zen at 3,4 is 36s then (26% clock increase = ~20% reduction in execution time ) so we're at ~ 28.8s assuming Zen at 4350mhz. Odd. I get less than that, and that's not even accounting for having 2 less cores yet
Click to expand...

Nope.

Run Broadwell-E at 8C/16T at 3.2GHz, and scored 33% faster at altered 100 sample rate....this still does not come close to what was shown by AMD today.
 
100 samples isn't enough to explain it either. Unless the 6900K under performed massively.
 
Shintai said:
100 samples isn't enough to explain it either. Unless the 6900K under performed massively.
Click to expand...

No, the problem now is the 6900k's performance (based on AMD"s numbers) is totally anomalous based on all the other tests lol.
 
Ieldra said:
No, the problem now is the 6900k's performance (based on AMD"s numbers) is totally anomalous based on all the other tests lol.
Click to expand...

AMD better have a very good explanation. Or they simply did like they did the last 10 years and straight out lied. The BF1 (GPU limited) and Dota 2 (Rigged to hell) gaming session wasn't exactly confidence inspiring.
 
jeremyshaw said:
Yeah, unfortunately with AMD's uploaded file and renderer not being the same as their website, it is nearly impossible to really get to the bottom of this.

I want to see if changing the render tiles to the default 64x64 will make the performance what you expect it to be. CPUs generally run slower with a larger tile size (under the Performance tab). Then again, the default sample size is 128 (as of 2.78a), not 100... Light bounces is set to 8 (min/max) in the file, but that is to be expected, since the default in 2.77 is 8. It is 12max 3min in 2.78a (again, as a default, the file changes this setting).

Either way, no matter how I analyze it, AMD's uploaded scene and instructions are clearly different from what they were running in the TDP demo.
Click to expand...
Yeah, it only took us about 12 hours to get to the bottom of that. Hopefully we will hear from AMD tomorrow on the specifics. But I have to say.....WTF!?

Does AMD know how many people WANT to be on its side?

I get it.

I am a tech "journalist" so I tell the facts and then give you subjective and objective analysis. But I want AMD to win. But this bullshit...is bullshit.

<redacted>This reeks of Roy Taylor.</redacted>
 
I smell an editorial brewing. Damn shame that AMD has banned me from all events, I could have asked them on-site. Let's give AMD a chance to answer via email however. Maybe it is just a simple thing.
 


This one just raises further questions. Different times than from the live demo. And why is the 95W Zen system drawing so much power vs the 140W 6900K system. Both uses around 180-190W.

~34 seconds from the Live demo is now ~25 seconds.
 
stashix said:
May I ask how? I didn't watch the stream, it started kinda late CET.
Click to expand...

Stuttered as hell on a 7700K because they made sure the load from 3rd party application (streaming) was locked to cores with high enough priority. Dota 2 can run with 60 FPS on an old Core 2.

One CPU got 8cores/16 threads, the other 4/8. Now tell 3rd party app to use 7-8 threads and run at a higher priority than the game.
 
My i7-3770k oc'd with small bus speed over clock. Win7 x64 cpu parking disabled.
GonFdQt.png
 
Halving the time of the Blender tests because they used 100 cycles instead of 200, seems to be a simple thing. I imagine that whoever set up the demo picked a number. In a large organization, I highly doubt that the same person who checked the "100" box for the demo was also involved in linking the Blender download to the community. The default 200 selection versus using 100 in a demo is not some sort of conspiracy. After all, BOTH cpus used the same setting. This is just a friggin' minor communications issue. One guy didn't know that the other guy changed a default.

Select 100 runs, try your rig, then compare to the demo. (Or, just cut all the published times using default values in half.) C'mon. This is taking a molehill and bulldozing it into something it isn't....
 
Shintai said:
Stuttered as hell on a 7700K because they made sure the load from 3rd party application (streaming) was locked to cores with high enough priority. Dota 2 can run with 60 FPS on an old Core 2.

One CPU got 8cores/16 threads, the other 4/8. Now tell 3rd party app to use 7-8 threads and run at a higher priority than the game.
Click to expand...
And here I was thinking something like this on the first page

Sigh
AMD, really?
 
c3k said:
Yeah, I did.

DId you even read the posts preceding mine? The ones implying that AMD did this on purpose? That there was a secret AMD conspiracy which has been exposed?

This 100 vs. 200 is minor. Get over it.
Click to expand...

The 100 vs 200 doesn't even come close to explaining the discrepancies. Read the thread.
 
Hitti2 said:
My i7-3770k oc'd with small bus speed over clock. Win7 x64 cpu parking disabled.View attachment 12586
Click to expand...
Ok, been reading up on another overclock site, I guess amd used 100 samples while above quote is 200 samples...

New results of 100 samples. Lol... Man I would'nt pay 1000+usd for a cpu that is 15 seconds faster in rendering etc... My i7-3770k is just fine...
blender 100samples.jpg
 
With 100 samples I get 46,69 seconds, yeah sample size doesn't explain it. 3770K @ 4,4 GHz, RAM @ 1600 MHz 9-9-9-24 1T.
ryzen.jpg
 
This is the kind of under-handed gimmickry bs that I absolutely hate.

It's too soon to call AMD out but where there is smoke there is fire.

Maybe I am jumping the gun here?

The 6700K dota2 stuttering seemed like absolute BS to me. Who is buying this crap? And why would AMD do this? Why not just run some benchmarks and if it's even or under then be honest about it. Compete on price if you can't compete on performance.

Has anyone here been to Microcenter? The AMD motherboard section is ... TINY .... the locked case with the AMD cpu's .... fully stocked, all the time ... for years.

At this point I hope Intel just comes out and beats the living shit out of them on price and shuts them down. I could care less about competition if it means one side is always bs'ing around.

For the time being, I'm going to just hold my tongue and see what the reviews says. In my heart of hearts I wish AMD would just release something, anything that beats AMD or Nvidia.
 
1:26 for an E3-1245 running at 3.6 Ghz. About what I expected. Not the best but better than I will need for a long time.
rizen.PNG
 
Blender 64 bit results on my sig rig...without the OC (everything dialed back to stock clocks).

1:51.08 normal process priority.
1:43.66 real-time process priority.


 
Last edited:
SixFootDuo said:
This is the kind of under-handed gimmickry bs that I absolutely hate.

It's too soon to call AMD out but where there is smoke there is fire.

Maybe I am jumping the gun here?

The 6700K dota2 stuttering seemed like absolute BS to me. Who is buying this crap? And why would AMD do this? Why not just run some benchmarks and if it's even or under then be honest about it. Compete on price if you can't compete on performance.

Has anyone here been to Microcenter? The AMD motherboard section is ... TINY .... the locked case with the AMD cpu's .... fully stocked, all the time ... for years.

At this point I hope Intel just comes out and beats the living shit out of them on price and shuts them down. I could care less about competition if it means one side is always bs'ing around.

For the time being, I'm going to just hold my tongue and see what the reviews says. In my heart of hearts I wish AMD would just release something, anything that beats AMD or Nvidia.
Click to expand...
Mmm
There are 2 things for me here

For normal people, eg customers, it would be enough that Ryzen would come close enough in a comparable workload
No one was actually expecting something that would flat out kill Intel

We were all just hoping for a close enough
So BS'ing isn't very helpful, nor does it them any good word of mouth wise

But I can see why they would do it
AMD is being traded on the stock market
Making things look incredible might bump the value of AMD stocks, even if it's just ever so shortly lived



Waiting for credible 3rd party benchmarks
And the prize point for Ryzen
 
Dont know if I fucked up something on my end and didn't do anything special. I just downloaded blender, opened the file, and clicked render and this is what I got.

5mrwZ9Z.png
 
SixFootDuo said:
Compete on price if you can't compete on performance.
Click to expand...

AMD can't afford to compete on price; investors are expecting Zen to return AMD to profitability. If it fails to do that, AMDs stock tanks again, and its highly likely they either go through restructuring (Chapter 11) or start selling off entire divisions (The GPU division comes to mind).
 
Assuming that they were running the latest version of Dota 2, this idea that someone running a core 2 duo could get 60+ fps isn't accurate, as 7.0 which was just released a few days ago introduced all sorts of cpu performance issues even for those running high end hardware. Not that I necessarily buy into any of these live demos for getting an idea on performance anyway. So much debate and so much trolling around basically a "sneak peek". Wait for the reviews once everyone has the hardware next year :)
 
If the 6-core Cannon Lake or Coffee Lake or whatever fuck the next mainstream desktop CPU is called is not compatible with my existing 1151 Xeon board, I will DEFINITELY be going back to AMD.
 
JosiahBradley said:
Sigh, you guys make me want to get out of bed and start benchmarking. :brb: Fine you win.

Also did anyone compile blender from source not just grab a binary? Lots of prepackaged apps on Windows that come from Linux projects are compiled at sort of lowest common denominator settings for compatibility. Compare that to say Gentoo where blender is going to fly being custom compiled with every x64 extension known man enabled. Don't have a Windows compiler handy but it might get you much faster results.
Click to expand...

Do we know if the windows version of blender is compiled against openblas or intel mkl? I'm guessing not unless you compile yourself.
 
rgMekanic said:
I used to livestream all the time, it's hard as hell on the machine depending on the codec and resolution you're streaming at, especially in demanding games (not that DOTA is demanding). For most BIG streamers, they do use 2 machines, 1 running the game, and running into a capture card of a 2nd thats doing the encoding and streaming.

EDIT: just as an edit, my W3690 (essentially a 990X) can only reliably stream at 720, and thats heavily game dependent.
Click to expand...
Bull, I stream to...from the same PC I play on...shadowplay only gives 1-3% drop in FPS...stop the FUD!
 
Brent_Justice said:
Blender Render: 1:14:51 so 1 minute and 14 seconds, almost 1 minute and 15 seconds time to complete on my i7 4770K @ 4.6GHz / DDR3 1600
Click to expand...

New Result with Render 100 Samples - 37.85 (37 seconds, that's a huge difference for me, don't understand what that means or why mine went down so much)

I ran it twice, second run at 100 Samples was 37.48

So I put it back to 200 just to double check, and it's 1:13.98 at 200 Samples

For whatever reason, that samples is making a whopping difference on my system.

So yeah, I am consistently getting 37 seconds at Render Samples 100
 
[MADHACKER]Weapon said:
prgHyoz.jpg


1 second slower then AMD's chart, 8 seconds faster then the live test. 6900K system in my sig.
Click to expand...

Interesting. So the graphic provided by AMD looks like it might be accurate, then. Now I want to know wtf they were doing that caused a 10 second difference in their live demo.

FWIW, I got ~58.5s last night with my 5820k @4.4. That was with WoW running in the background along with several other programs, so it's in line with others' scores at 200 samples.
 
Last edited:
thefordmccord said:
So, what about dual channel memory vs quad channel? Do you think that AMD was using dual channel memory in their Intel system?
Click to expand...

I found this after some searching:

http://www.forbes.com/sites/patrick...onstrations-and-product-details/#6cb657c51411

"These systems were otherwise identical except that one system was running on AMD’s new 8 core Ryzen processor while the other ran on Intel’s i7-6900K processor. Both systems ran their processors on stock air coolers, meaning that there isn’t any unfair cooling advantage for one chip or the other. Both processors ran DDR4 2400 memory, however the Intel system ran in quad-channel while the AMD system only ran in dual-channel since the Intel system is capable of quad-channel, and that is theoretically the fastest configuration of the RAM."



edit:

Now Reddit has a slide with system specs, and it seems the Intel had dual channel, not quad:


https://i.redd.it/r39v4xwzre3y.jpg

Looks like Forbes screwed up...
 
bpizzle1 said:
Interesting. So the graphic provided by AMD looks like it might be accurate, then. Now I want to know wtf they were doing that caused a 10 second difference in their live demo.

FWIW, I got ~58.5s last night with my 5820k @4.4. That was with WoW running in the background along with several other programs, so it's in line with others' scores at 200 samples.
Click to expand...

Upstream it was commented upon that successive runs needed a restart to flush the cache, otherwise subsequent runs had faster times than the first run. Brent or Dan did that (restarts) and had consistent times within .1 second. Without the restart, the times varied by 10s of seconds.

I suspect that may be cause of the live demo difference to which you're referring. (Unless I've misunderstood your post.)
 
I hope someone sent them a email, maybe they can answer with a response once they get into the office.
 
What are you guys talking about both systems in the demo were about a min. She even says that these tests take about a min to complete before we start. Every test done so far by us falls directly in line with what they have shown. Are we watching the same thing. Or are you just on something...lol
 
Imhotep said:
What are you guys talking about both systems in the demo were about a min. She even says that these tests take about a min to complete before we start. Every test done so far by us falls directly in line with what they have shown. Are we watching the same thing. Or are you just on something...lol
Click to expand...

The "minute" remark was in relation to the Handbrake test.
 
At AT forums the version is that default Blender setting is 128. So, will anyone give it a shot? Since 100 samples clearly get close enough to demoed system, i believe that one should get it close to livestream version.
 
You must log in or register to reply here.
Back
Top