Geforce GTX 1060 Thread: faster than RX 480, 120W, $249

[trane]

That's familiar, reminds me of a certain claim of Radeon R9 390X performance under 100W during gaming for Radeon RX 480 'as if it were fact' before launch.

So true. All we have seen leaked marketing slide from Nvidia without any substantiation at all.

 

[Sweepr]

You still need to admit few if any of your threads concerning upcoming products carry a [Rumor] tag even though some clearly should.

You think it's a rumor that NVIDIA claims the card is faster than Radeon RX 480? Did you expect anything different from them?

Ps: Can't add more than two words to the title.

 

[SolMiester]

My take on this is the 480 reference will match or bet the 1060, however the 1060 will OC better than ref 480 and therefore beat it. 3rd party 480 with decent cooler system (again), will change the positions..

NV doesnt need to show false benchmarks or claim dreams of OC greatness, they never have...

 

[Silverforce11]

NV doesnt need to show false benchmarks or claim dreams of OC greatness, they never have...

Really?

"Pascal, 10x faster than Maxwell!"

"1080, 2x faster than Titan X.. "

"65C, cool as a cucumber!" & "Premium design, craftsmanship" when describing their throttling reference cooler.

Surely though, these companies have a product to sell and their marketing needs to generate hype. Understandable. But for you to say they don't falsely hype their products, is very strange denial of very recent official statements.

 

[wilds]

I remember "crazy overclockability" from JHH regarding the GTX 1080. In terms of absolute MHz, it was. But the fact that he was advertising it by up selling the 1080 FE reference cooler ruined it. The 2.1 GHz 1080 they had only could maintain those clocks in a 60fps capped demo at 100% fan speed.

I think that is much worse than 1 AMD employee calling Fury X an "Overclocker's Dream".

 

[Thala]

That's familiar, reminds me of a certain claim of Radeon R9 390X performance under 100W during gaming for Radeon RX 480 'as if it were fact' before launch.

You still don't get my point. None of this 100W speculation made it to the threads headline as if it were fact. Apparently for good reason.
What you are doing here is the exact opposite, pulling a speculation out of context and putting it into the threads headline as if it were fact.

I am not saying that it is utterly impossible, that GTX 1060 beats RX480. I even consider it likely for some games and very unlikely for others.

 

[krumme]

Issue of course is, that RX480 is close to 980 at DX11 but its quite a bit ahead in DX12.

And its not like we are 2 years away from dx12 now. All the comming performance AAA games next 12 month is dx 12. It seems crazy to buy for anything but dx12 at this moment imo. Not since 3d introduction in the nienties have we seen such a change. DX9 was big but nothing like this because it also implies better cpu perf.

 

[railven]

And its not like we are 2 years away from dx12 now. All the comming performance AAA games next 12 month is dx 12. It seems crazy to buy for anything but dx12 at this moment imo. Not since 3d introduction in the nienties have we seen such a change. DX9 was big but nothing like this because it also implies better cpu perf.

I personally feel at least for PC gaming, DX12 isn't going to matter for the mainstream until all those popular MMOs/F2P update to it.

 

[Mopetar]

I personally feel at least for PC gaming, DX12 isn't going to matter for the mainstream until all those popular MMOs/F2P update to it.

I don't know how much those really matter since it's not as though many require terribly powerful GPUs in order to run the game reasonably well. Existing games don't have a lot of incentive to overhaul their engines for what would be marginal performance gains, at least not anytime soon, especially if most of the consumer-base won't benefit for another 2-3 years before most are running hardware that can take advantage of the DX12 features.

 

[jj109]

While the hardware nerds are all looking at the new GPUs, looks like a ton of people already bought their GPU for the summer.

June Steam Hardware Survey

DIRECTX 12 GPUS FEB MAR APR MAY JUN
NVIDIA GeForce GTX 970 6.38% 6.79% 6.97% 6.93% 7.35% +0.42%
NVIDIA GeForce GTX 960 4.09% 4.31% 4.45% 4.60% 4.98% +0.38%

Both AMD and Nvidia are late to the mid-range party.

 

[desura]

Looks like a solid advance. It should be about the same performance as the 970?

Looking at how long they were stuck at 28nm, I bet 14nm will be here for twice as long. Looks like a good buy come December when winter hits.

 

[desura]

My take on this is the 480 reference will match or bet the 1060, however the 1060 will OC better than ref 480 and therefore beat it. 3rd party 480 with decent cooler system (again), will change the positions..

NV doesnt need to show false benchmarks or claim dreams of OC greatness, they never have...

Last time I built my PC, I figured that Intel was a better value than AMD for power draw reasons.

 

[Flapdrol1337]

You think it's a rumor that NVIDIA claims the card is faster than Radeon RX 480? Did you expect anything different from them?

Ps: Can't add more than two words to the title.

Yeah I expect something different.

I expect nvidia to pretend the rx480 doesn't exist and compare to the 960, look at that performance gain!

 

[bill3]

My take on this is the 480 reference will match or bet the 1060, however the 1060 will OC better than ref 480 and therefore beat it. 3rd party 480 with decent cooler system (again), will change the positions..

NV doesnt need to show false benchmarks or claim dreams of OC greatness, they never have...

It's interesting it's quite a lot lower specced than 480. 192 bit bus, only 4.4 teraflops (5.8 for 480), less TMU's, 3/6GB ram vs 4/8, etc.


Given that, all the talk of how "inefficient" the 480 is is kinda funny. NV is using 80% of the TDP for a card with significantly less brute force. Even if for safety's sake you designated the 480 as a 170 watt card, Nvidia has little to no real efficiency advantage. When I have read approximately 1.4 trillion comments the last few days how AMD crapped the bed on so called "efficiency" (meaningless term anyway since a GPU's job is to play games not conserve watts).

Given Nvidia is claiming 15% more performance in 480, first of all that must be at 1080P, at higher resolution 480 will pull ahead due to 256 bit bus. Second, it's likely a best case scenario, using TWMTBP Nvidia favored games. I suspect in reality they more like trade blows even at 1080P.

All in all, pretty happy with my 480 purchase then.

Also pricing, if the rumors of 249 for 3gb 1060 and 299 for 6g are true...YIKES, Nvidia done goofed. I cant imagine that's the true pricing though. I expect maybe 249 MSRP for the 6GB card.

Should be fun, I love new card launches and benchmarks!

 

[SolMiester]

Really?

"Pascal, 10x faster than Maxwell!"

"1080, 2x faster than Titan X.. "

"65C, cool as a cucumber!" & "Premium design, craftsmanship" when describing their throttling reference cooler.

Surely though, these companies have a product to sell and their marketing needs to generate hype. Understandable. But for you to say they don't falsely hype their products, is very strange denial of very recent official statements.

None of the above are false graphical benchmarks or false claims on OC greatness TBF.
I havent heard 2 x faster than titan anywhere!

 

[Silverforce11]

None of the above are false graphical benchmarks or false claims on OC greatness TBF.
I havent heard 2 x faster than titan anywhere!

Then you didn't watch JHH on stage for the launch.

He also said this of the 1080 reference: "Crazy good overclocking!", when all the review sites show 5-12% performance gains from OC, which throttled back after 20 minutes. -_- Certainly nowhere as good as Maxwell OC.

 

[SolMiester]

Then you didn't watch JHH on stage for the launch.

He also said this of the 1080 reference: "Crazy good overclocking!", when all the review sites show 5-12% performance gains from OC, which throttled back after 20 minutes. -_- Certainly nowhere as good as Maxwell OC.

No, i havent watched the launch..and AFAIK, 2100mhz is crazy good OC if you use partner cards...
Fury X didnt OC period...

 

[Silverforce11]

No, i havent watched the launch..and AFAIK, 2100mhz is crazy good OC if you use partner cards...
Fury X didnt OC period...

If you didn't watch it, that's why you missed out on all the marketing over-hype statements.

Actual in-game boost clocks @ stock 1860mhz, OC to -> 2100mhz 1860 is about 13%. That's why reviews see around 10% performance gain from OC. It's not OC from 1600 or 1700mhz.

It's better than Fury X stock volt OC, but not by much. Far from the 20-25% performance gain for Maxwell OC.

 

[SolMiester]

Isnt 1860mhz the boost speed? Did you ever think that perhaps NV class boost and OC over standard speed an OC?

 

[Azix]

how can it be faster than a 980? 980 is 4.6 tflops, this 1060 at 1700 Mhz is 4.352 teraflops. Cutting it very close. How will they beat a 980 with that, much less overclocked 980s? eg. for 980ti vs 1080 the 980ti is 6.5tflops, 1080 is 8.62 tflops. The difference is over 32% but is the 1080 that much faster than a stock 980ti? Not to mention a situation where the 1060 actually has a lower tflop rating than a stock 980. if it wins you can bet its some driver shenanigans

 

[iiiankiii]

Isnt 1860mhz the boost speed? Did you ever think that perhaps NV class boost and OC over standard speed an OC?

Bro, don't be dense. When the card boosts to 1860mhz, that's the performance that is recorded on the benchmark. That's the number shown in the graph. Therefore, when you overclock to say 2050mhz, you don't magically get gain 443mhz worth of extra performance from the base clock of 1607mhz, you gain 190mhz on top of the boost clock. That's how it works. It's called math and logic.

 

[SolMiester]

Bro, don't be dense. When the card boosts to 1860mhz, that's the performance that is recorded on the benchmark. That's the number shown in the graph. Therefore, when you overclock to say 2050mhz, you don't magically get gain 443mhz worth of extra performance from the base clock of 1607mhz, you gain 190mhz on top of the boost clock. That's how it works. It's called math and logic.

Dont be insulting...bro!..My understanding is no boost is guaranteed either...
I asked a question that perhaps thats how NV see the boost OC, you dont need to insult me ..

 

[Silverforce11]

Isnt 1860mhz the boost speed? Did you ever think that perhaps NV class boost and OC over standard speed an OC?

See they classing it like that is already false marketing hype because it's fake.

Go check out OC reviews for the 1080, they don't get 25% extra performance at 2.1ghz. Because that's not how OC works for NV GPUs. You don't look at base clocks and compare that to your final OC clocks.

As @iiiankiii said, the performance at stock, is from the higher than paper spec boost clocks.

980 and 980Ti reference boost to 1.2 to 1.25ghz, well above paper spec. When they OC to 1.5ghz, they gain around 20-25% more performance. That's good OC.

The worse thing is the performance in benchmarks aren't even reality because gamers have their GPU inside a case, not an open bench optimal condition where boost clocks are higher. After 10 minutes gaming inside a case, as you may know by now, the reference 1080 throttles hard. Some reviewers find it running at base clocks. That's a loss of over 250mhz of performance compared to the first few minutes of it boosting at 1.89ghz.

A gamer sees benchmarks, woah, nice numbers, comes home, games in his case and now it's under performing by 10-15% unless he cranks up the fan so now it's noisy.

So please, next time don't say stuff like NV doesn't falsely hype it's products, it happens to every single reference card out there.

 

[SolMiester]

I disagree....the hype machine is AMD, the PR machine is NV

 

[tviceman]

Then you didn't watch JHH on stage for the launch.

He also said this of the 1080 reference: "Crazy good overclocking!", when all the review sites show 5-12% performance gains from OC, which throttled back after 20 minutes. -_- Certainly nowhere as good as Maxwell OC.

If you didn't watch it, that's why you missed out on all the marketing over-hype statements.

Its as if Nvidia and AMD are both feeding off each other's stupidity to appease dumb, rabid fans.

....Hey! They made ridiculously bold claims! Lets make bolder ones! .....Hey! Their claims aren't even close to reality! Lets try and top it!