That's why I wat to see benchmarks soon. I'm debating whether the 1070 or 480 is better for me. I'm leaning more toward the 1070 seeing how long I've kept my current card.
That's why I wat to see benchmarks soon. I'm debating whether the 1070 or 480 is better for me. I'm leaning more toward the 1070 seeing how long I've kept my current card.I'll take fury performance with ultra low power consumption for 200 ish dollars. Actually that is perfect for me. I have a 2600k at 4ghz and would prefer a cool running card as I live in Texas and it's summer time. Also only shelling out $200 is nice too. I'm willing to wait it out. For a little while I considered a 1070 but if I'm honest with myself a 480 Is plenty of gpu for me. I don't even know if 4ghz 2600k would be enough for a 1070 anyways.
That's why I wat to see benchmarks soon. I'm debating whether the 1070 or 480 is better for me. I'm leaning more toward the 1070 seeing how long I've kept my current card.
Because the 480 I don't think would be strong enough as I'm planning to get a 1440p monitor later. I also enjoy max settings most of the time. As of now I don't know if I would be able to do that with a single 480 and I don't ever plan on running two cards. Obviously Vega would be great but I don't want to wait that long.Then why are you looking at a 1070 when AMD has had better drivers over the life of their last gen cards. If you had gotten a 780 then you would have had less long term. AMD recently has had the better support.
IMO, if the 48 CU rumor does turn out to be true, then Polaris 10 is not a 232mm^2 chip, but considerably bigger.
Why ? CU's for AMD on average take about 33% of a die, CU + MTU around 43%.
You would be better off buying an Rx480 now, mining with it and then waiting for benchmarks of the actual flagships from nvidia and amd. Save the $150 bucks as a down payment on Vega or big Pascal.That's why I wat to see benchmarks soon. I'm debating whether the 1070 or 480 is better for me. I'm leaning more toward the 1070 seeing how long I've kept my current card.
Yea I've never done mining but I'm curious to see how well the 480 will really perform because it sure is affordable.You would be better off buying an Rx480 now, mining with it and then waiting for benchmarks of the actual flagships from nvidia and amd. Save the $150 bucks as a down payment on Vega or big Pascal.
You know these cards are coming, just like we know the 1080 and 1070 are going to fall off a cliff when big Pascal drops.
Sent from my XT1575 using Tapatalk
Can you provide some background reading on this?
only pictures, and its only approximates:
tahiti , remember its both the MTU and CU in the green
http://www.3dcenter.org/dateien/abbildungen/AMD-Tahiti-Die-Shot-markiert.kleine vorschau2.jpg
FIJI is more shader heavy then normal so has higher % CU ( normally 44-48 shaders a rop, 64 for FIJI)
http://www.3dcenter.org/dateien/abbildungen/AMD-Fiji-Die-Shot.jpg
Tonga is a bit of an odd ball, like Fiji higher number of CU a ROP but CU to die size is quite a bit worse then Fiji:
http://www.pcgameshardware.de/screenshots/759x427/2015/06/Tonga_Overlay_20150603102014-pcgh.jpg
http://cdn.wccftech.com/wp-content/uploads/2013/12/AMD-Hawaii-GPU.jpg
Thank you. Also, if you don't mind commenting, in your opinion how likely is it 232mm2 chip could have 48 CU's?
I see only 6 pin connector on Sapphire GPU.
There are, but they may be more geared towards compute (or GP100 cores really are quite different to GP104).
https://devblogs.nvidia.com/parallelforall/inside-pascal/
TLDR is NV claiming improved scheduling and some new instructions. It's not just a straight die shrink (again unless GP104 is quite a different architecture, not just a different layout), but it doesn't appear to matter very much for gaming so far.
Where does it say 1.08GHz ?So with this leak today with the RX480 (presuming it's accurate) running at 1.08ghz are we looking at fury performance if the clocks are set at 1.266ghz?
They say this 200 card puts VR into the mainstream.
But I thought the barrier to entry was the high price for the headsets. How does the 480 change the game so to speak with VR?
If the 3dmark leaks hold true and the 36CU (2304sp) RX480 is at 980/Fury level, 44CU (2816sp) is 22% extra hardware, 48CU (3072sp) is 33% extra. Nevermind all the extra hardware that's also enabled, AMD also disables TMUs and other units when cutting CUs.
At the same 1266MHz clocks in the leak, and assuming somewhat linear scaling, 44CU/+22% puts P10 at 1070/TX tier, 48CU/+33% with a 100MHz bump is probably at 1080 territory.
If so, this also explains why Fiji was so lackluster and an afterthought of a GPU just to test HBM and have something to match stock 980Ti, all the crucial R&D went to Polaris (and Vega) and it could very well pay off.
Should P10 overclock well and the full variants speculated here fill the $300 bracket, we're in for a treat and a reset of the GPU market pricing that's long overdue. Hell, the $200-230 480 already does that in some ways, although not at the full effect this 3072sp part at the stated max $300 price tag could.
As you say, this also explains the 1070 and 1080 coming so early out of the gate, with a $100 early adopter tax, and with so little supply that you see it out of stock for weeks and it sells out within minutes.
Maybe Raja saying AMD was ahead of the transition to 14nm in the mainstream segment back then in January wasn't smoke and mirrors after all.
RV770 did have 800sp in it when the reviews came out. Who knows, maybe this ends up being true.
Where does it say 1.08GHz ?
So with this leak today with the RX480 (presuming it's accurate) running at 1.08ghz are we looking at fury performance if the clocks are set at 1.266ghz?
I think the long stagnation on 28nm may have biased some expectations about what kind of average yields should be obtained. Nvidia doesn't seem to be doing particularly stellar on yields either, with even the ultra-specialized P100 being a cut-down chip. And the GP104 in the GTX 1070 is cut down by a full 25% of shaders, more than expected. While that latter decision could be the result of market segmentation, I think it has to do with the rush to production and the fact that TSMC is still a relative newcomer to 16FF+.
We'll probably be seeing mediocre yields for 6 months or so. Maybe up to a year. For reference, it was in October 2013 that AMD released the first wave of R9 200 series rebrands, which (temporarily) did away with cut chips for Pitcairn and Tahiti. That was about 22 months after the first 28nm GPU (7970) hit the review sites.
Samsung's FinFET process is known to be denser than TSMC's. The Apple A9 smartphone chip was dual-sourced on TSMC and Samsung, and the Samsung version was about 9% smaller. And that was with the Samsung 14LPE process. 14LPP (the version GloFo adapted) is supposed to provide even smaller size, as well as better performance. Also, don't forget that even on the same process node, there is a trade-off between density and speed. Nvidia could have crammed more shaders in GP104 in the same die size if they wanted, but then they wouldn't be able to hit the 2000+ MHz clock speeds that has everyone so impressed. Since at least Kepler, Nvidia has preferred to go for less dense and higher clocking designs, while AMD's GCN products are generally oriented around denser designs with lower clock rate caps.
In addition, we've all been assuming that Polaris 10 is a 232mm^2 chip based on some information written in an old LinkedIn profile. This belief was bolstered by AMD's positioning of P10 as a "mainstream" part, the modest TDPs, and the low pricing announced at Computex. But it's at least possible that we could be wrong, that the 232mm^2 GPU was something other than Polaris 10 (a future product, an engineering sample that never made it into production, whatever else) and that Polaris 10 is actually closer to 300mm^2. Only time will tell. Fortunately, we don't have all that much longer to wait.
A few posts ago, that leak comes from a reviewer who posted a youtube video. Claims bios and spec different to other samples (AMD sending different "RX 480" to different people, his buddy's sample runs at 1227mhz, TPU's sample runs at 1266mhz). His one runs at 1080mhz. A play on the GTX 1080.
"This is an overclockers dream!"Unfortunately everything nVidia says has to be verified by outside sources.
Hopefully that's the case then. A slightly bigger, faster clocked chip at $300 might be the sweet spot for a lot of people.
So are you saying (speculating, of course) nVidia knows AMD had a Polaris GPU up their sleeve that will ~match the 1080 and are going to sell it for $300. So, their response is to early release theirs at 2x the price, +$100 F you over even harder tax? Surely even nVidia isn't that bold?
A few posts ago, that leak comes from a reviewer who posted a youtube video. Claims bios and spec different to other samples (AMD sending different "RX 480" to different people, his buddy's sample runs at 1227mhz, TPU's sample runs at 1266mhz). His one runs at 1080mhz. A play on the GTX 1080.