http://videocardz.com/61154/amd-radeon-rx-480-crossfire-3dmark-performance
>480 is at r9 390x, but tester was incompetent enough to include early single tests of r9 390x/gtx 980 of when they were recently released (stock clocks, early drivers)
http://wccftech.com/amd-rx-480-faster-than-nano-980/
>rx 480 performs the same as gtx 980, 480 is at same clock as videocardz test, 480 is about 500pts higher than vcz test
who do i trust /g/
WCCFtech is a POO-IN-LOO run AMD fanboy site that regurgitates rumors found on other more established and reliable websites
Never trust them.
>>55127295
reviewers in 12 days
>>55127295
Different clocks.
1266 Mhz and 1080 Mhz
Think of this as the "boost" speed. The card probably had a switch to change between the two like my current HD 7950 card does.
I'd assume the card can OC up to 1400-1500 Mhz probably, this is where the real performance of the card will shine.
>>55127366
videocardz:
>Clock speeds seems to be consistent for all CrossFire scores (1266 MHz). New benchmarks align with our previous single-GPU results. However the average score changed slightly due to lower performance in FireStrike Performance preset (run at 1230 MHz).
seems like they're the same in both tests
>>55127333
i can believe the amd fanboy part, but what does it have to do with >designated
>>55127434
http://wccftech.com/about/
founders are pooinloos and nearly all editors are pooinloos/muhammedans
>First benchmarks of new AMD GPU show up
>They promise amazing performance
>AMD and their shills promise the GPU will be an 'overclockers dream'
>Get hype
>Card gets released
>It's shit
Won't let you trick me this time, AMD!
>>55127551
>muslims
>>55127295
will amd have something to compete agains the 1080? a cf of 480 is a no go for me.
>>55127551
why dont you like indians
>>55127774
in 6 months i guess like with the 290/x, i wouldnt know
>>55127366
1288Mhz are the official stock clocks
They flashed different preview samples with different clocked bioses to identify the leakers.
>>55127910
why are they so retarded?
>>55127582
hurr durr asshole
why not wait til we see reviews? you might be right,but you might be shit wrong,
and so far benchmarks show it clocking to 1070 performance.
>>55127774
Vega 10 in late 2016 or very early 2017.
Basically Nvidia is not competing in the midrange and AMD is not competing in the high end for a few months. Classic duopoly.
>>55127974
>Basically Nvidia is not competing in the midrange
Heh, until they release the 1060, which won't be far away
>>55127966
>and so far benchmarks show it clocking to 1070 performance.
AHAHAHAH, no, the only remotely legit looking benchmarks of videocardz.com put it in between the 390x and the 980
>>55128005
it's going to be a while. 960 released 4 months after 970 and 980.
Not to mention Pascal sucks and 1060 is not going to be competitive with the 480. Maybe with the 470.
>>55127929
relax, theyre just anxious for amd's flagship
>Linking anything from WCCFtech
They literally copy/paste any news about AMD no matter how ridiculous it sounds. It's a website that posts Korean shoops where 480 RX is beating Titan X just so 2 days later they can show a benchmark where 480x loses to 980 gtx in a firestrike bench. Give me a break, wait for actual reviews, or even better - youtube videos made by owners.
>>55127295
Firestike with my r9 290 at 1ghz just returned 11696 basically identical to the vdz shill score. So I'm sure AMD somehow managed to make a new arch that's actually clock for clock slower than their old shit from three years ago.
290 xfire can put up 23000, 980 sli you would expect scores of 28k plus.
>>55127869
I'm not him but if you've been around the techweb scene at all, you'll see that pooinloo sites are trashy rumor mills: the tech equivalents of National Enquirer. If they don't have a press release to rehash, they just make shit up. They seem to mostly be beta Indian teenagers running these joints. Hell, even Anandtech started off like this, only he was legit for awhile before he found the Apple payola teat to suckle. Now he's drawing a salary and stock options for it--unless you really think he was an elite engineer that Apple just had to have...for their PR team.
It's not a bash on Indians.
>phonearena
>gsmarena
>wccftech
>>55127869
because 99.9% of all indians in tech shill for amd so you can't trust anything they have to say about gpu comparisons.
>>55127295
Just wait till it releases.
How is life with CF really? Does it still have microstutters?
>>55128596
>new arch that's actually clock for clock slower than their old shit from three years ago.
while I won't trust any Polaris benchmarks before release/NDA sunset, it's not impossible for it to be slower clock-for-clock in a few contrived scenarios due to the bandwidth cut, in cases where delta color compression doesn't pan out, etc.
>>55131854
Nah, the comment was retarded.
The 390 has far more cores than the 480, so the 480 needs faster clocks to be comparable.
>>55131599
Yeah it's horrible. Most people end up selling one of their cards. SLI on the other hand is working fine these days as long as your application supports it.
>>55129779
They sure help my portfolio though.
>>55128017
With no OC. Pascal cards are already clocked insanely high, but if the 480 isn't Fury tier with some OC it may reach the 1070 considering the small gap. Also, we have to see if it's something achievable or just for the lucky ones. Once again, it may be a beast, it may be just good. We'll see when it drops
>>55131868
4 cu isn't going to account for a 25% increase in base clock and dcc. A 40 cu Polaris would be slower at 1000 mhz than a 290 if that bench is correct.
I'm not going to believe something that not only defies logic but isn't even consistent with the benchmarks on the same site. Apparently Polaris ranges between an r9 290 and a Fury X.
>>55127333
>AMD fanboy
yeah they sure have me convinced with the 500 gtx 1080 hype articles they made
>>55132225
>SLI
>working
lol
I cant wait to fall for the RX480 meme, my 660ti is starting to give me the piss
https://www.youtube.com/watch?v=X6dCz4Wt1Ek
reminder that 480 could overclock very well based on PCB shots
>>55127295
Trust ur mum. Wait for the actual benchmarks/comparrisons fucking idiot.
