-
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.
***The DigitalFoundry Face Off Thread***
- Thread starter shankly1985
- Start date
- Joined
- 25 Nov 2011
- Posts
- 20,679
- Location
- The KOP
Compared to others they do a reasonably good job of side by side comparisons and a bit more useful detail rather than dwelling heavily on just average fps - however as per the other thread there is the odd issue where they could have widened their testing methods a bit and got a better picture.
As an example they hamstrung the 970 in the first part of the test - the limits they place on both cards to get a "level" playing field actually has a bigger negative impact on the 970 than the 390 - as can be seen in the last bit where the 970 isn't hamstrung and somehow manages twice the percentage performance increase (that is comparing the framerate in the same scene and not at the same timeline where the scenes are different) ratio compared to the core MHz increase once it was working as normal + the overclock. Would have been interesting to see the 390 OC results against it.
As an example they hamstrung the 970 in the first part of the test - the limits they place on both cards to get a "level" playing field actually has a bigger negative impact on the 970 than the 390 - as can be seen in the last bit where the 970 isn't hamstrung and somehow manages twice the percentage performance increase (that is comparing the framerate in the same scene and not at the same timeline where the scenes are different) ratio compared to the core MHz increase once it was working as normal + the overclock. Would have been interesting to see the 390 OC results against it.
Last edited:
Not very close at all.
Permabanned
- Joined
- 12 Sep 2013
- Posts
- 9,221
- Location
- Knowhere
Claiming supposed facts and claiming bias because you perceive that to be the case after looking at the forum posters names and seeing those you believe are red orientated has little value and the funniest thing is you're wrong. The 980ti sli does better more consistently and after watching the videos and there views the message I got was if you go dual card go Green. So sorry Choppers but you're very wrong
Last edited:
- Joined
- 25 Nov 2011
- Posts
- 20,679
- Location
- The KOP
Lol just goes to show doesn't it. Well done LambChop
To be fair to LambChop, when I saw the thread title and then who posted it, AMD wining was my first thought as well, but after watching the first video linked, I was a little surprised that it wasn't a AMD whitewash as I initially expected. So well done Shankly for finding and posting up a useful site with some good comparisons videos.
I do agree with fs123, though, I'm not entirely sure of the knowledgeability of the presenters. going from a 4790k @4.6GHz to a 5960X @4.4GHz is definitely not almost a doubling of CPU horsepower, especially as far as games are concerned. But I would have to watch more to see if it is just silly comments like that or a something else.
I do agree with fs123, though, I'm not entirely sure of the knowledgeability of the presenters. going from a 4790k @4.6GHz to a 5960X @4.4GHz is definitely not almost a doubling of CPU horsepower, especially as far as games are concerned. But I would have to watch more to see if it is just silly comments like that or a something else.
Ok, I've just watched a video, he says a fury is beating the 980ti in single card performance and has a clear advantage in shadow of mordor and better dual card scaling. There's a fraction of a second where the FuryX creeps over the 980Ti in single card perf.
Nothing like not saying what you see is there, these guys would have been hopeless on Catchphrase.
Edit - Added more, the said they also are not doing a like for like comparison on the division benchmark. Yeah these guys are great.
Nothing like not saying what you see is there, these guys would have been hopeless on Catchphrase.
Edit - Added more, the said they also are not doing a like for like comparison on the division benchmark. Yeah these guys are great.
- Joined
- 25 Nov 2011
- Posts
- 20,679
- Location
- The KOP
It's not just about getting a tiny bit of fps over the competition...
These guys so much more then anyone else that I know off..
Take note of the more important frame latency it's this what makes a game playable and smooth..
So far they have been very non bias imo apart from I would have liked to see the 390 overclocked also in the 970 vs 390 I have nothing bad to say about them.
These guys so much more then anyone else that I know off..
Take note of the more important frame latency it's this what makes a game playable and smooth..
So far they have been very non bias imo apart from I would have liked to see the 390 overclocked also in the 970 vs 390 I have nothing bad to say about them.
LambChop, you actually argue that AMD has better multigpu scaling? Ever since they moved to XDMA crossfire, they have had better scaling (when it actually is working). Nvidia is still stuck at using SLI bridges. Interesting to see how NVlink improves their scaling.
There is a reason why CF generally has a bit better raw framerate scaling :S and that isn't necessarily a positive thing. (Though you really need to do some tweaking of both to get a fair comparison).
Look at the frame variance with crossfire.
There is no issue with SLI bridges, that isn't the problem in the slightest.
Soldato
- Joined
- 1 May 2013
- Posts
- 10,011
- Location
- M28
what hardware are these guys using, what brand and models of cards, clocks etc?
SLI bridge reduces transfer over PCI-E, but comes with a latency penalty. So it's a tradeoff really. It works smoothly, but you lose a bit of scaling.
AMD has other issues that scaling, they have stuttering occasionally. And I don't think they are all GPU memory problems (would be nice to see 8GB card in crossfire benched to see if it helps on stuttering).
Permabanned
- Joined
- 12 Sep 2013
- Posts
- 9,221
- Location
- Knowhere
Definitely, You hear plenty of questionable info on DF video's but the graphs are handy to get a look at performance and consistency.
So they tell us how the Ti monsters the Fury in Bf4 and although closer in dual card it's still an SLI win, then they tell us how SOM does well in crossfire providing stutter free performance unlike the rest, that's then followed by how disappointing crossfire is in GTAV compared to SLI, Which in turn is followed by pointing out how terribly crossfire spikes in Unity and then how bad the gameplay stutter is in Far cry 4 with crossfire.
It sounded too me like Crossfire got torn a new one overall yet all you heard was AMD did better in SOM, Wow! The force is strong in you young Choppers.
Last edited:
Didnt watch the videos or read any of the comments.....You know what they say.....Look before you leap Mr Chop.
The way DF do these video's is good. Having also done many side by side comparison video's, they are quite tricky to get the same walk through that is repeatable and then you get people who always insist that such and such isn't fair etc etc.
I am with you on that one. Being on the red team I think these videos are excellent and right now all I am concerned about is getting to the bottom of any issues for MY card on newly released games. There have been a fair number of problems on ROTTR for both cards and it really is good to see these videos getting to the nitty gritty and comparing this stuff.
IMHO (I have been in IT Support for over 20 years, but have no real knowledge of GFX engineering or monitor Technology) it seems to come down to a few issues with the FIJI cards....from what I can gather from these videos
No or little optimization for HBM, OR that the 4GB limit is actually an issue for extremely high texture packs (I put my settings down to High instead of Very High and it sorted 99% of my stuttering issues).
CPU bottle necking (I have i5 3570K and should be upgrading but waiting for Zen) which could be causing some of the frame spikes on FIJI cards. I hope that DX12 can sort out this issue as I believe it sorts out the CPU stuff or whatever.
Well in my mind these videos have given me what I think is the best insight into these issues and make logical sense to me. I just hope AMD can sort the issues with driver updates or give all of us FIJI owners a free 8GB HBM2 Polaris upgrade in 4 months time.
Last edited: