Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [VC]GTX 1060 specifications leaked - faster than RX 480
New Posts  All Forums:Forum Nav:

[VC]GTX 1060 specifications leaked - faster than RX 480 - Page 49  

post #481 of 735
Is it confirmed anywhere that the 1060 is going to be better than 970 in terms of performance, and if so, how much do you think will be? I'm in need of a new graphics card and 970 fell in price, but I'm waiting to see what 1060 offers.

Also 1070 prices are RIDICULOUS, how long does it take to stabilize? I'm not paying €500 for that thing.
post #482 of 735
Quote:
Originally Posted by Aymanb View Post

Is it confirmed anywhere that the 1060 is going to be better than 970 in terms of performance, and if so, how much do you think will be? I'm in need of a new graphics card and 970 fell in price, but I'm waiting to see what 1060 offers.

Also 1070 prices are RIDICULOUS, how long does it take to stabilize? I'm not paying €500 for that thing.

Im with you on the prices. Especially for the components and PCB's. When they called the 1080 the "Founders Edition" they mean that the people buying them are creating the financial foundation for the next run. Im waiting. And Im waiting a long time before I move on anything
post #483 of 735
Quote:
Originally Posted by DMatthewStewart View Post

Hold on a minute. That is entirely wrong. We are comparing the 120w usage to a device that uses 150w. That means that, since we are trying to equate the performance, our max amount of power used is 150w. 1060 uses 120w. That means its 20% more power effecient because its only using 80% of the total wattage calculated, or total wattage pool. 120w is 80% of 150w. So youre calculating it incorrectly just as they have done. Thats strictly full TDP. Before we even touch on performance. So 150w is 100% of the pool of power that we are rating. The 1060 uses 120w. Its 20% more power effecient

And where does 1.15 come from?

Dividing 150/120 gives us no number related to power efficiency. The original graph is incorrect because it was calculated incorrectly. Working it backwards and then typing it out seems right at first. Its not. Unless Im missing something here. Im open minded to being totally wrong but this seems way too easily identifiable as incorrect.

If the cards had the same performance and same electricity consumption. Efficiency is 1.

E.g 120 watts divided by 120 watts divided by the equal performance would act as a modifier of 1.

However if one card uses 30 less wants. That means one card is 25 percent more efficient.

Even. 150 divided by 120.

1.25 I.e 25 percent more efficient. However of this card also performs 15 percent better while consuming 25 percent better it acts as a modifier.

I.e 1.15. IF IT performed the same this modifier would be 1 but since it's performs 15 percent better, this becomes 1.15.

There for to get the efficiency you have to multiply both numbers.

1.25 * 1.15 = efficiency.
post #484 of 735
Quote:
Originally Posted by variant View Post

That's impossible to predict. Polaris and Vega don't share the same architecture. Polaris is Graphics IPv8 similar to Tonga/Fiji, while it seems Vega is Graphics IPv9 which was developed for 14nm.
Quote:
Originally Posted by EightDee8D View Post

Better hope it's same ipv8 cuz we know what huge expectations and hype does. tongue.gif

^^^FINALLY somebody who gets it.

If you keep your expectations low at worst you won't be disappointed, and at best you'll be very pleasantly surprised.

If you hype your expectations through the roof then the best you could hope for is "it did what it was supposed to do".
post #485 of 735
Quote:
Originally Posted by magnek View Post


^^^FINALLY somebody who gets it.

If you keep your expectations low at worst you won't be disappointed, and at best you'll be very pleasantly surprised.

If you hype your expectations through the roof then the best you could hope for is "it did what it was supposed to do".

Well, earlier, when I also said that it's IP9, I wasn't implying that it will be better. I was implying that it will be different.
Just to show another interpretation of the IP9 thingy... biggrin.gif Of course it will be a disappointment - and occasionally burning houses - as any other AMD card has done...
post #486 of 735
No I'm saying staying conservative, not generating false hope or hype (or worse, actively generating false hype for nefarious purposes) is the key to long term happiness. biggrin.gif
post #487 of 735
Quote:
Originally Posted by magnek View Post

No I'm saying staying conservative, not generating false hope or hype (or worse, actively generating false hype for nefarious purposes) is the key to long term happiness. biggrin.gif

I think the big thing that's going to happen with Vega is they are going to bring back double precision and make it a compute architecture again. 4gb of memory and the professional market don't mix. 16gb does and both tahiti and particularly hawaii were solid professional cards. Most of the improvements and changes will be with this I believe.

Considering the 4096 shader leak, I was wondering why there wasn't an increase in shaders vs fiji and I think it's because the double precision had to be taken out due to the effect on power and die area with Fiji. With 14nm finfet they can do this again while having something with reasonable power consumption and performance. What this means is more of the improvements will be for professional tasks and not gaming. So performance I could see being around fiji + 20-25% mostly from clocks. These changes or rarely improve gaming performance and typically lower frequency.

One notable exception would be if the broke down the CU's into smaller clusters.

If they reduce the number of cores in a CU from 64 to 32, it could up IPC up 10-15%. But other than that Polaris + 65% or Fur + 20-25% are pretty safe bets if the card has 4096 shaders because there wasn't an noticeable IPC increase from Polaris to Hawaii.
post #488 of 735
Quote:
Originally Posted by tajoh111 View Post

I think the big thing that's going to happen with Vega is they are going to bring back double precision and make it a compute architecture again. 4gb of memory and the professional market don't mix. 16gb does and both tahiti and particularly hawaii were solid professional cards. Most of the improvements and changes will be with this I believe.

Considering the 4096 shader leak, I was wondering why there wasn't an increase in shaders vs fiji and I think it's because the double precision had to be taken out due to the effect on power and die area with Fiji. With 14nm finfet they can do this again while having something with reasonable power consumption and performance. What this means is more of the improvements will be for professional tasks and not gaming. So performance I could see being around fiji + 20-25% mostly from clocks. These changes or rarely improve gaming performance and typically lower frequency.

One notable exception would be if the broke down the CU's into smaller clusters.

If they reduce the number of cores in a CU from 64 to 32, it could up IPC up 10-15%. But other than that Polaris + 65% or Fur + 20-25% are pretty safe bets if the card has 4096 shaders because there wasn't an noticeable IPC increase from Polaris to Hawaii.
They said 15% ipc increase from 290 to rx480, but actually it regressed. maybe rop/rbe bottlenecks ? but if you compare fiji vs pol10's perf/tflops there's like 10% increase.

i think 4k shader vega gpu will be -/+ 5% of 1080 stock. considering they fix all the bottlenecks.
post #489 of 735
Quote:
Originally Posted by EightDee8D View Post

They said 15% ipc increase from 290 to rx480, but actually it regressed. maybe rop/rbe bottlenecks ? but if you compare fiji vs pol10's perf/tflops there's like 10% increase.

i think 4k shader vega gpu will be -/+ 5% of 1080 stock. considering they fix all the bottlenecks.

bandwidth bottleneck, there had been indications that VRAM OC increases perf, this indicates that bandwidth is insufficient.

if vega will use HBM, then yes they'll get more out of the chip, supposedly it'll be sufficient enough to satiate vega's compute units.
post #490 of 735
Soooo.... where are the reviews? its 7.7.2016 smile.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Rumors and Unconfirmed Articles
This thread is locked  
Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [VC]GTX 1060 specifications leaked - faster than RX 480