these run quite nicely

great to hear, where can I subscribe to your blog?

cool fucking thread dude why dont you get kidnapped by candleja

>>56706910
KE

File: k.jpg (110KB, 584x584px) Image search: [Google]

110KB, 584x584px

>>56706866

>>56706910
Who posted th

>>56706866
You bought.....Unironically.......two 1070's........for 900$ instead of a single 1080 or pascal titan?

wait are you the faggot who made a thread bragging about his dual 970's a few years back?

>>56708076
you do realize a single pascal titan or 1080 isn't that much better than a single 1070, but dual SLI offers about 80% scaling?

>>56708113
And you know SLI is pretty much junk in real-world performance.

>>56708125
80% is real world

>>56708132
>80% scaling
>real world

oh you poor stupid fuckstick......

>>56706866
Nobody's dumb enough to buy 2x1070's right?
Is this a joke?

>>56708153

I think its the same faggot autismo who bought 970's for SLI

and he bought them AFTER the 3.5gb scandal broke.

>>56708146
>the jelly is real
You'd know SLI just werks these days in almost every game you'll play right?

File: nvidia-geforce-gtx-1080-3840x2160-2-way-sli-scaling.png (75KB, 649x861px) Image search: [Google]

75KB, 649x861px

>>56708146
>>56708153

you're right. it's closer to 100% a lot of the time.

File: perfrel_1920_1080.png (193KB, 999x730px) Image search: [Google]

193KB, 999x730px

>>56708173
>>56708186
I wish I lived in the same fantasy world as you guys.

>average of 18% scaling at 1080p
>average of 33% scaling at 1440p
>even discounting the games where SLI is broken the average doesn't hit 50%

Only the biggest retards around using multi-GPU setups. The problems and shitty scaling aren't worth it. You buy the most expensive single card you can afford.

Enjoy your ebin new setup, OP. Be sure to let us know how just mad you are six months from now. :^)

>>56708275
most AAA games will have about 90% scaling though. if you play indie games and that sort of thing then yeah stay away from SLI.

>>56707589
>>56708061
Mfw you cunts can't even do it right.

Worse than OP.

>>56708275
dumbass. the reason it doesn't show good scaling at low res is because the games become CPU as opposed to GPU bottlenecked.

>>56709325
Not the guy you're quoting, but you're retarded. These benchmarks are done with literally top tier performance. There is not bottleneck on that chart. The only bottleneck is your fucking brain.

>>56709365
holy shit you're dumb.

>>56708301
In your dreams, cuck. There were no indie games in TPU's benchmark suite.

>>56709325
>1440p
>low res

Godfuckingdamn Nvidia users are retarded.

>>56709510
are you actually this stupid? it becomes less bottlenecked by the CPU and moreso by the GPU at 1440p. take it up to 4k and that's when you'll start seeing 90% scaling.

>>56706866
How no one adressed the bait that is the crossfire bridge in the middle?

>>56708113
>isnt that much better
>30% difference

Ok. Keep trying to convince yourself

>>56709655
i mean 30% is nothing compared to the 90% improvement that SLI'ing two will give you

>>56709714
90% from sli. Fucking what planet are you on. You are talking aboslute bollocks.

>>56709725
https://i.4cdn.org/g/1474441820356.png

>>56709733
So you think spamming that one single results chart makes it true?

see: >>56708275
that for a start

>>56709777
dude the last 10 posts have been discussing how that chart isn't worthwhile

>>56709792
I can see 2 and they were probably both from you.
You seem a bit trigger happy with numbers, no wonder you think its 90% if 2 posts equal 10

>>56709829
okay, let's be conservative and say 70%. that's still 40% better than 30%.

>>56708153
shouldn't it be called a 2140?

>>56709840
I always wonder what retards were like, now I know

>>56710050
no you don't. because I'm not one.

>>56706910
DELETE THI

>>56708132
>>56708146
>>56708186
>MUH SCALING


the microstutter still is a fucking pain in the ass.
Last time I had SLI was with two GTX770's.

nevar again.

>>56710231
Also are you OP? and is that picture bait?. It seems to be as theres a crossfire att. in the photo

>>56708315
Do it right, then

>>56706866
T. Manchild

I like how all of you talk about scaling.
I like how none of you asked what CPU the OP has.
Seeing how he is about bragging on a budget, I doubt he has a X CPU, meaning the max amount of PCIE lanes he has available is 20.
I like how his cards will be running either 16x4 or 8x8 tops, meaning they both get scaled down to fuck and back.
I like how it's 2016 and none of you realize that unless you run a high end CPU (500 bucks +) that SLI will NOT scale well.

I have a 6700k with a single 1080. Dual 1070's, would not perform any better than the single card on this system, just like 90% of systems out there.

Why do you still do this dumb shit?

>>56710363
my CPU is a haslel overclocked to 4.5Ghz and I have a high bandwidth SLI bridge

>>56710410
Which model?
Overclocking has nothing to do with anything

>people are still falling for the SLI meme
loving
every
laugh

>>56710363
GPUs can't saturate an x8 lane

>>56710458
only reason its still around is for nvidia to milk 2-300% more money out of the stupid people that buy their already way overpriced jewware

>>56710410
Which CPU?

>>56710363
more PCIe lanes on the CPU makes about 1fps difference

https://youtu.be/rctaLgK5stA?t=309

Well, since this thread at least is kinda about the 1070, which brand should I buy if I'm getting one?

>>56710499
4670k but before you say anything look at this
>>56710522

>>56710535
So a 16 lane CPU.
Meaning you're either running both cards at 50%, or one at 75 and one at 25.
For 300 less you could have bought a 1080 and gotten better performance out of your machine than you have right now.

As for >>56710522
You want me to take a youtube video as a liable source of this when on paper and in god knows how many online articles you will find that what I am saying is true?

Not trying to make fun of you, we all make uneducated purchases, it happens. But if you can, return them and opt in for a single 1080 and see for yourself that you will see a performance increase. Not only that, but you no longer have to worry about SLI support in games and how it scales it down.

>>56710626
>You want me to take a youtube video as a liable source of this when on paper and in god knows how many online articles you will find that what I am saying is true?

nope. I'd ask for literally only 1 that contradicts linus' findings. bear in mind that said article needs to be benching pcie 3.0 since the bandwidth has changed drastically with each revision.

>>56710655
Google it man, you will see tons of articles of how SLI fails unless you have a high end CPU.

It's not rocket science.

Also, you're taking linus as a viable source.... aka the guy who has run 4x sli titans knowing and admitting that 2 of them aren't doing shit.

>>56710711
>4
there's your problem. he mentioned in his video that a 4th card literally won't do a single thing unless you have the additional lanes from your cpu.