4x 980Ti vs 2x 1080Ti

[AntonChigurh]

Just now, pas008 said:

margin of error goes both ways

lol troll harder

Lol troll harder you didn't even look at the benchmarks did you?

[pas008]

Just now, N1ghtshade said:

Lol troll harder you didn't even look at the benchmarks did you?

lol 720p results more cpu bound and integrated gfx learn to read

and who the fuck plays at that

[othertomperson]

1 minute ago, pas008 said:

lol 720p results more cpu bound

and who the fuck plays at that

If that GPU is an R7 240 or Integrated Graphics as in the graph, abso-fucking-lutely. This is one of the most egregious misrepresentations of data I've seen this week, well done.

[AntonChigurh]

Just now, pas008 said:

lol 720p results more cpu bound

and who the fuck plays at that

Well it's on integrated graphics so they didn't have a choice what res they played. But true. If it's 2017 and you're still playing at 720p, you probably need to upgrade your prebuilt PC from 2004 just a tad

[AntonChigurh]

Just now, othertomperson said:

If that GPU is an R7 240 or Integrated Graphics as in the graph, abso-fucking-lutely. This is one of the most egregious misrepresentations of data I've seen this week, well done.

The entire point is the CPU

???

 

"egregious misrepresentation of data"

m8 U wot it's numbers on a chart. Numbers dont lie, and are pretty hard to misrepresent in a situation like this

[pas008]

1 minute ago, N1ghtshade said:

The entire point is the CPU

???

 

"egregious misrepresentation of data"

m8 U wot it's numbers on a chart. Numbers dont lie, and are pretty hard to misrepresent in a situation like this

troll harder

mr butt hurt

[DildorTheDecent]

2 > 1

 

If you're an enthusiast then jump on the multi card option. Maybe not four cards though. 

[AntonChigurh]

Just now, pas008 said:

troll harder

mr butt hurt

How is that even an argument? You've obviously run out of replies, so you've just resorted to calling people butt hurt. 11/10 ign, you win an internet

[othertomperson]

Just now, N1ghtshade said:

The entire point is the CPU

???

 

"egregious misrepresentation of data"

m8 U wot it's numbers on a chart. Numbers dont lie, and are pretty hard to misrepresent in a situation like this

The entire point is the CPU, which is why to cripple CPU performance by making the benchmark ridiculously GPU bound (by using an R7 240 or worse) is a misrepresentation of data.

 

http://www.gamersnexus.net/hwreviews/2822-amd-ryzen-r7-1800x-review-premiere-blender-fps-benchmarks/page-7

 

Here these are better. CPU bound, a series of benchmarks that compare just the CPU performance of a load of CPUs including the Haswell and Kaby Lake i7s.

[pas008]

Just now, N1ghtshade said:

How is that even an argument? You've obviously run out of replies, so you've just resorted to calling people butt hurt. 11/10 ign, you win an internet

lol you havent proved yourself at all

just rambling

[AntonChigurh]

3 minutes ago, othertomperson said:

The entire point is the CPU, which is why to cripple CPU performance by making the benchmark ridiculously GPU bound (by using an R7 240 or worse) is a misrepresentation of data.

They both used integrated graphics on all games except 2. And on the 2 they didn't they both used the same GPU. Clicc the link son, look at the rest of the benchmarks.

[AntonChigurh]

Just now, pas008 said:

lol you havent proved yourself at all

just rambling

Literally what I just called you out for doing. You're not so good at this whole argument thing, are you?

[Biggerboot]

Ignoring the benchmarks for a sec, I'll say my last experience with sli 5-ish years ago was unpleasant.  I experienced flickering and micro-stuttering.  AFR would rarely go smoothly for me.  That being said, it may have very well improved, but you really gotta look less at the framerates and more at other metric like compatibility and game specific issues.

 

SLIing 2 1080ti's is obviously something only an enthusiast should be doing, but it is the only way to get more performance than one.  So if you're in the market for the most performance possible and money isn't an issue, then that's you're only option.  Dual GPU at any other level is harder to recommend, but if you enjoy the tweaking process then the small issues may not be a factor to you.

 

3 or 4?  Just don't unless it's a mining or rendering rig.

[pas008]

Just now, N1ghtshade said:

Literally what I just called you out for doing. You're not so good at this whole argument thing, are you?

count the benchmarks in 1080p with higher settings or larger resolution they trade blows

[AntonChigurh]

Just now, pas008 said:

count the benchmarks in 1080p with higher settings or larger resolution they trade blows

Yes, and in anything lower, the 7700k wins. So if its a tie in 1080p, and the 7700k wins at lower res, tell me, whats the better cpu?

 

/thread

[othertomperson]

4 minutes ago, N1ghtshade said:

They both used integrated graphics on all games except 2. And on the 2 they didn't they both used the same GPU. Clicc the link son, looks at the rest of the benchmarks.

I did click the link. This link seems to be aggregating data from a massive range of different tests, including synthetic benchmarks, single threaded synthetic benchmarks, good gaming benchmarks, very very bad gaming benchmarks, and sticking them all together and coming to the conclusion that the two are the same? It's absurd. This is exactly why sites like CPU boss and Game Debate don't give good data, because you can't just aggregate such disparate data and produce anything reliable.

 

There is a huge difference between the 7700K and 4790K in any competent test most of the time.

 

2 minutes ago, Biggerboot said:

Ignoring the benchmarks for a sec, I'll say my last experience with sli 5-ish years ago was unpleasant.  I experienced flickering and micro-stuttering.  AFR would rarely go smoothly for me.  That being said, it may have very well improved, but you really gotta look less at the framerates and more at other metric like compatibility and game specific issues.

 

SLIing 2 1080ti's is obviously something only an enthusiast should be doing, but it is the only way to get more performance than one.  So if you're in the market for the most performance possible and money isn't an issue, then that's you're only option.  Dual GPU at any other level is harder to recommend, but if you enjoy the tweaking process then the small issues may not be a factor to you.

 

3 or 4?  Just don't unless it's a mining or rendering rig.

Just as well PC hardware in no way progresses in half a decade then, or your anecdote might be out of date!

[pas008]

Just now, N1ghtshade said:

Yes, and in anything lower, the 7700k wins. So if its a tie in 1080p, and the 7700k wins at lower res, tell me, whats the better cpu?

 

/thread

do you even remember what you are arguing about?>

[Biggerboot]

1 minute ago, othertomperson said:

Just as well PC hardware in no way progresses in half a decade then, or your anecdote might be out of date!

It very well could be.  I'm still not a heavy enough AAA gamer to really justify SLI for myself.  If your SLI rig is serving you well then more power to you.  I wouldn't recommend it to a first time builder though (and my first complete build did feature 2x250s).

[AntonChigurh]

Just now, othertomperson said:

I did click the link. This link seems to be aggregating data from a massive range of different tests, including synthetic benchmarks, single threaded synthetic benchmarks, good gaming benchmarks, very very bad gaming benchmarks, and sticking them all together and coming to the conclusion that the two are the same? It's absurd. This is exactly why sites like CPU boss and Game Debate don't give good data, because you can't just aggregate such disparate data and produce anything reliable.

 

There is a huge difference between the 7700K and 4790K in any competent test most of the time.

I beg to differ. You need a large range of tests to show the uses for each cpu. For instance, if someone benchmarked a 7700k and a Xeon in GTA V, people would think "oh look this xeon's shit, why would you buy it?". Now if you benchmarked them both in something like multi-core synthetic benchmarks, the xeon would whoop ass. It's all about having both so you can show which is best for each situation

[AntonChigurh]

3 minutes ago, pas008 said:

do you even remember what you are arguing about?>

Yes. 7700k/Kaby lake is better than Haswell for gaming

 

Do you even remember why you tried arguing? Cause you were dead wrong from the start

[pas008]

Just now, N1ghtshade said:

Yes. 7700k/Kaby lake is better than Haswell for gaming

 

Do you even remember why you tried arguing? Cause you were dead wrong from the start

lol as you can see at higher resolutions they are trading blows

 

19 hours ago, N1ghtshade said:

Why in god's name do you have a Haswell in a gaming pc?

hahaha

19 hours ago, N1ghtshade said:

You have a Haswell in your rig. Think that might be the problem buddy

here

19 hours ago, N1ghtshade said:

That's the point. You'd probably get a lot better performance out of something like a 7700k. Gaming does better with fewer, faster cores.

yeah alot better huh

[othertomperson]

4 minutes ago, Biggerboot said:

It very well could be.  I'm still not a heavy enough AAA gamer to really justify SLI for myself.  If your SLI rig is serving you well then more power to you.  I wouldn't recommend it to a first time builder though (and my first complete build did feature 2x250s).

Well I wouldn't anymore, but that's because the prices of single GPU nearly always lines up with the dual-GPU equivalent. If we were back in the stage where two 270Xs were still a couple of hundred less than a 290X it would make sense again financially.

 

I also don't think any new system builder really wants to be jumping straight into multiple 1080 Tis. At the same time 5 years ago the market wasn't as striated by resolution as much as now. A 1080 Ti (and especially two) was geared at 4K and won't give any 1080p benefit at all. Anything over a 560, however, was always going to be used for 1080p or 1440p. The only things two 580s got you were framerate and time before upgrading.

 

3 minutes ago, N1ghtshade said:

I beg to differ. You need a large range of tests to show the uses for each cpu. For instance, if someone benchmarked a 7700k and a Xeon in GTA V, people would think "oh look this xeon's shit, why would you buy it?". Now if you benchmarked them both in something like multi-core synthetic benchmarks, the xeon would whoop ass. It's all about having both so you can show which is best for each situation

So if I took that Xeon system and benchmarked it in Crysis 3 at 4K on a GTX 950, and ran Cinebench, then added the two scores together and divided by two, you think that would give me a number that was in any way indicative of its performance? Aggregating numbers across different hardware, different benchmarks, different testing methodologies will give you utterly meaningless numbers, which is exactly what this has done. This site gives the false impression that these two CPUs are identical because it's averaged across a lot of disparate results when they are not directly comparable or even necessarily valid.

[AntonChigurh]

Just now, pas008 said:

lol as you can see at higher resolutions they are trading blows

 

hahaha

here

yeah alot better huh

You told me this before. And I said that, yes, they are at higher res. But look at the lower resolutions. Not gonna repeat myself again. If they're trading blows at 1080p, and the 7700k wins at 720p, which is better for gaming?

 

And yes, it is. I just proved it to you

[pas008]

1 minute ago, N1ghtshade said:

You told me this before. And I said that, yes, they are at higher res. But look at the lower resolutions. Not gonna repeat myself again. If they're trading blows at 1080p, and the 7700k wins at 720p, which is better for gaming?

 

And yes, it is. I just proved it to you

lol wow you are lost arent you

should I quote your shit quotes again?

[AnonymousGuy]

On 4/5/2017 at 5:54 PM, paps511 said:

What do you think would be better, 4x 980Ti or 2x 1080Ti?

 

Hypothetically of course.........

4 GPUs is more about waving around a money dick than getting real performance gains.  Unless you're doing rendering or whatever the fuck uses multi-gpu without SLI.  

 

We're not debating SLI 1070 vs 1080Ti so it's really a wasted effort arguing whether or not you should SLI 1080Ti's.  If you want to get the max performance, you don't really have a choice.