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

[raghu78]

And absolutely no new DX12 titles will be TWIMTBP? What's good for one is good for the other.

And why only link CoDIII? That's the only one that shows 480 ahead of 1060 and that's the one you linked.. See this is what I'm talking about on these forums. Nobody can keep things real.

I pointed out COD because its a DX11 game where Rx 480 wins big and also its much more popular than a Anno 2205 where GTX 1060 wins even bigger or an AC Syndicate.

https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/6.html
https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/7.html

The fact is AMD has brought GPUOpen to combat Gameworks so that AMD can counter the huge losses in Gameworks titles like AC Syndicate. The difference is GPUOpen is open source and does not harm Nvida GPU perf. It allows developers to bring console optimizations onto PC and shader extensions which map directly to GCN opcodes. I think AMD has learnt from Nvidia who is the market leader and this trend will hurt the PC gamer as he will have to choose GPU based on games he plays as we will see more games where it plays vastly better on one GPU vendor.

 

[Keysplayr]

New comparison from Hardware.fr:

Gainward GTX 1060 vs RX 480 Sapphire Nitro+: €280 match

- Power: Gainward GTX 1060 draws 83W less under load (TW3 Hairworks Off)
- Noise: 10.7 dBa less under load (!), 6.8 dBa less than Nitro+ at Silence Mode
- Overclock:
208 MHz GPU (+ 11%) and 2300 MHz memory (+ 15%) vs +1380 MHz (+ 3%) and 2250 MHz memory (+ 12.5%)
- Actual Gain from Overclocking (Overall): 12.3% vs 6.1%



www.hardware.fr/focus/118/rx-480-sapphire-nitro-vs-gtx-1060-gainward-match-280.html

Well, that's gotta be annoying.

Nice find!

 

[Keysplayr]

I pointed out COD because its a DX11 game where Rx 480 wins big and also its much more popular than a Anno 2205 where GTX 1060 wins even bigger or an AC Syndicate.

https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/6.html
https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/7.html

The fact is AMD has brought GPUOpen to combat Gameworks so that AMD can counter the huge losses in Gameworks titles like AC Syndicate. The difference is GPUOpen is open source and does not harm Nvida GPU perf. It allows developers to bring console optimizations onto PC and shader extensions which map directly to GCN opcodes. I think AMD has learnt from Nvidia who is the market leader and this trend will hurt the PC gamer as he will have to choose GPU based on games he plays as we will see more games where it plays vastly better on one GPU vendor.

Problem is, we can go back and forth throwing out single benchmarks all day. We need to look at the overall. And in this case, the 1060 is the clear better choice. For MANY reasons. Not just benchmarks.

 

[Leadbox]

Problem is, we can go back and forth throwing out single benchmarks all day. We need to look at the overall. And in this case, the 1060 is the clear better choice. For MANY reasons. Not just benchmarks.

How many points do you get for that post?



This is the reason I put in that mod warning. A worthless post only meant to incite.





esquared
Anandtech Forum Director

 

[Keysplayr]

How many points do you get for that post?

How many are you willing to give me?





And you need to stop. You are not helping matters.

I will lock the thread if it continues.


esquared
Anandtech Forum Director

 

[Cookie Monster]

How many points do you get for that post?

This is exactly the type of posts we dont need.. zero relevance to the thread. Just adding more mud to the mud-slinging.

 

[raghu78]

Problem is, we can go back and forth throwing out single benchmarks all day. We need to look at the overall. And in this case, the 1060 is the clear better choice. For MANY reasons. Not just benchmarks.

I do not agree. Whether the GTX 1060 or the Rx 480 is the better choice comes down to the performance in games you are interested and how well the GPUs have aged from both vendors in the past. Kepler is murdered by GCN GPUs which were supposed to be its competitors. Maxwell gets the same treatment in DX12.

http://www.hardwarecanucks.com/foru...iews/73040-nvidia-gtx-1060-6gb-review-21.html

I bet you were one of the GTX 970 supporters. Just look at the GTX 970 getting murdered by R9 390 (and R9 290x which is its equal) in DX12 and losing badly even in DX11 at 1440p when they were supposed to be equals. Now pushing the same argument in 2016 that GTX 1060 is better performing based on 2013-2015 games when there is clear evidence that Rx 480 is faster on avg in DX12 games is shortsighted. I can agree that power efficiency is Nvidia's major advantage and thats why in mobile they will be dominant. But for the desktop gamer interested in perf and perf/$ the Rx 480 is the better card for any person who keeps the GPU for 2+ years.

 

[Keysplayr]

This is exactly the type of posts we dont need.. zero relevance to the thread. Just adding more mud to the mud-slinging.

I agree and I'm trying my hardest not to retaliate in kind.

 

[Leadbox]

This is exactly the type of posts we dont need.. zero relevance to the thread. Just adding more mud to the mud-slinging.

What he is and what he posts are inseparable to me, but I take your point with :thumbsup:

 

[Gikaseixas]

I do not agree. Whether the GTX 1060 or the Rx 480 is the better choice comes down to the performance in games you are interested and how well the GPUs have aged from both vendors in the past. Kepler is murdered by GCN GPUs which were supposed to be its competitors. Maxwell gets the same treatment in DX12.

http://www.hardwarecanucks.com/foru...iews/73040-nvidia-gtx-1060-6gb-review-21.html

I bet you were one of the GTX 970 supporters. Just look at the GTX 970 getting murdered by R9 390 (and R9 290x which is its equal) in DX12 and losing badly even in DX11 at 1440p when they were supposed to be equals. Now pushing the same argument in 2016 that GTX 1060 is better performing based on 2013-2015 games when there is clear evidence that Rx 480 is faster on avg in DX12 games is shortsighted. I can agree that power efficiency is Nvidia's major advantage and thats why in mobile they will be dominant. But for the desktop gamer interested in perf and perf/$ the Rx 480 is the better card for any person who keeps the GPU for 2+ years.

Couldn't agree more. At this point one should pick according to his/her needs and features.
AMD and NV both have high and lows. I'm actually having a hard time recommending one over the other to a friend of mine.

Sent from my HUAWEI MT7-L09 using Tapatalk

 

[raghu78]

Couldn't agree more. At this point one should pick according to his/her needs and features. AMD and NV both have high and lows. I'm actually having a hard time recommending one over the other to a friend of mine.

Sent from my HUAWEI MT7-L09 using Tapatalk

If your friend will keep the GPU for a long time (2+ years) I think the Rx 480 in time will be the better card across the board. I think we are seeing a pattern with Kepler and Maxwell. Look at tpu's charts at the time of 780 ti, 970 launch and today. btw tpu has only one DX12 game which is RoTR so their chart is quite disadvantageous to AMD.

https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/23.html
https://www.techpowerup.com/reviews/ASUS/GTX_970_STRIX_OC/25.html
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_780_Ti/27.html

Look at R9 290X vs 780 Ti at 780 Ti launch and R9 290X vs GTX 970 vs 780 Ti at 970 launch and perf today. R9 290x went from 7% slower vs 780 Ti at 1440p on launch to 6% faster today. If you take DX12 / Vulkan games it gets ugly for both Kepler and Mawell.

 

[Arachnotronic]

If your friend will keep the GPU for a long time (2+ years) I think the Rx 480 in time will be the better card across the board. I think we are seeing a pattern with Kepler and Maxwell. Look at tpu's charts at the time of 780 ti, 970 launch and today. btw tpu has only one DX12 game which is RoTR so their chart is quite disadvantageous to AMD.

https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/23.html
https://www.techpowerup.com/reviews/ASUS/GTX_970_STRIX_OC/25.html
https://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_780_Ti/27.html

Look at R9 290X vs 780 Ti at 780 Ti launch and R9 290X vs GTX 970 vs 780 Ti at 970 launch and perf today. R9 290x went from 7% slower vs 780 Ti at 1440p on launch to 6% faster today. If you take DX12 / Vulkan games it gets ugly for both Kepler and Mawell.

780 Ti is based on an architecture that first debuted in early 2012. 290X is a late 2013 architecture.

Pascal is a thoroughly modern architecture. Not at all the same thing IMO.


Sent from my iPhone using Tapatalk

 

[JustMe21]

I was doing a build and video card replacement, so I got the opportunity to play with the MSI Radeon RX480 and EVGA Geforce GTX 1060 SC Gaming cards. I also compared it to my Gigabyte GTX 960 Windforce 2X OC Edition card. The only issue I saw with the 1060 was on the Graphics Test 2 on the TimeSpy benchmark. The textures would shimmer and the light rays were more opaque than on my GTX 960 or the Radeon RX 480. I did a clean install of the 368.81 drivers for both the 1060 and 960 and even did the 960 tests before and after the 1060 card. The highest wattage usage I saw was around 200 W for the 960 and 1060, but on average, the 1060 was usually using less power. On the RX 480, the highest wattage usage was around 225 W.

 

[raghu78]

780 Ti is based on an architecture that first debuted in early 2012. 290X is a late 2013 architecture.

Pascal is a thoroughly modern architecture. Not at all the same thing IMO.

Sent from my iPhone using Tapatalk

GCN architecture debuted in early 2012. Anyway Hawaii brought more ACEs compared to Tahiti which was the major architectural improvement. GTX 970 meanwhile launched in late 2014, roughly a year after R9 290X. It should be more modern than Kepler and do well against R9 290X right. Why does it lose to R9 290X today even in DX11 while getting hammered in DX12. Sorry but the past 2 generations have seen AMD GPUs age much better. Anybody denying it is just in plain denial. oh btw Maxwell and Kepler were also thoroughly modern architectures when they launched. We know now how that turned out.

 

[Arachnotronic]

GCN architecture debuted in early 2012. Anyway Hawaii brought more ACEs compared to Tahiti which was the major architectural improvement. GTX 970 meanwhile launched in late 2014, roughly a year after R9 290X. It should be more modern than Kepler and do well against R9 290X right. Why does it lose to R9 290X today even in DX11 while getting hammered in DX12. Sorry but the past 2 generations have seen AMD GPUs age much better. Anybody denying it is just in plain denial.

Hawaii brought several improvements beyond the additional ACEs, including improved geometry processing, more L2$ and greater L2 bandwidth, more ROPs, etc.

780 Ti was literally a DP focused GPU that had to fill in as a premium gaming GPU because a true successor to GK104 wasn't ready. Ever wonder why we never got a GM104?


Sent from my iPhone using Tapatalk

 

[bystander36]

GCN architecture debuted in early 2012. Anyway Hawaii brought more ACEs compared to Tahiti which was the major architectural improvement. GTX 970 meanwhile launched in late 2014, roughly a year after R9 290X. It should be more modern than Kepler and do well against R9 290X right. Why does it lose to R9 290X today even in DX11 while getting hammered in DX12. Sorry but the past 2 generations have seen AMD GPUs age much better. Anybody denying it is just in plain denial.

What makes the most sense, at least to me, is simply the fact that AMD predicted/engineered the current trends in dev development. The console wins likely are the biggest reason. As such, Nvidia has been playing catch up on these trends. They had no idea what was going to happen, while AMD did.

Now that many of these trends are clear as day, Nvidia has been redesigning their newer GPU's to over come these new changes. Each new generation of GPU will likely age better than the previous ones for Nvidia, as they were no where near on track with their designs as their current ones are much closer. That isn't to say they don't have more work to do.

 

[raghu78]

Hawaii brought several improvements beyond the additional ACEs, including improved geometry processing, more L2$ and greater L2 bandwidth, more ROPs, etc.

780 Ti was literally a DP focused GPU that had to fill in as a premium gaming GPU because a true successor to GK104 wasn't ready. Ever wonder why we never got a GM104?

Sent from my iPhone using Tapatalk

Sorry but the argument of extra L2 cache, more ROPs holds true for GK110 also vs GK104 as it does for Hawaii vs Tahiti. btw Hawaii's geometry processing was improved wrt Tahiti. But we all know that geometry processing was Kepler's greatest strength against older GCN products. Hawaii reduced the gap but Kepler was still better and Maxwell improved further on it. So that is not the reason for GCN GPUs to do better against Kepler or Maxwell.

http://www.anandtech.com/show/10325/the-nvidia-geforce-gtx-1080-and-1070-founders-edition-review/29

 

[zinfamous]

What makes the most sense, at least to me, is simply the fact that AMD predicted/engineered the current trends in dev development. The console wins likely are the biggest reason. As such, Nvidia has been playing catch up on these trends. They had no idea what was going to happen, while AMD did.

Now that many of these trends are clear as day, Nvidia has been redesigning their newer GPU's to over come these new changes. Each new generation of GPU will likely age better than the previous ones for Nvidia, as they were no where near on track with their designs as their current ones are much closer. That isn't to say they don't have more work to do.

I actually think nVidia knew what was going to happen, simply bet against it and/or didn't think it was going to be a threat to them.

In the end, it may not be. AMD still has a massive hill to climb and has to grab back a lot of market share this generation or they are sunk. I think they do have the better strategy, but even the better strategy can be defeated by better marketing.

nVidia still wins in raw performance in the minds of the many, regardless if AMD is set to win in design (we can pretty much claim, looking back, that AMD has been winning in design for the last 2 generations because it has been something of a long game with them--but the reality is that they weren't winning performance at that time when nVidia simply had the better hardware for yesterday's API through those generations)

 

[bystander36]

I actually think nVidia knew what was going to happen, simply bet against it and/or didn't think it was going to be a threat to them.

In the end, it may not be. AMD still has a massive hill to climb and has to grab back a lot of market share this generation or they are sunk. I think they do have the better strategy, but even the better strategy can be defeated by better marketing.

nVidia still wins in raw performance in the minds of the many, regardless if AMD is set to win in design (we can pretty much claim, looking back, that AMD has been winning in design for the last 2 generations because it has been something of a long game with them--but the reality is that they weren't winning performance at that time when nVidia simply had the better hardware for yesterday's API through those generations)

I'm not referring to knowing that AMD would have control of dev trends, but not knowing what they specifically needed to design their GPU's to do.

What is kind of shocking is how Nvidia has managed to beat AMD's "superior designs". I think the reality is that Nvidia has had superior designs, but AMD has the drivers seat with having the dev's designing specifically for their hardware. And yeah, that's a big deal, but I don't think that means Nvidia is doing a bad job, they are just at a huge disadvantage. As the shift in direction starts to settle down, we won't see old architectures have as much troubles with new content.

 

[bononos]

What makes the most sense, at least to me, is simply the fact that AMD predicted/engineered the current trends in dev development. The console wins likely are the biggest reason. As such, Nvidia has been playing catch up on these trends. They had no idea what was going to happen, while AMD did.

Now that many of these trends are clear as day, Nvidia has been redesigning their newer GPU's to over come these new changes.............

It wouldn't make sense for Nvidia to have "no idea" on what is going on. They have more than a small inkling on the rising trends. It must have been a conscious design decision on Nvidia's part for about 3 generations now. Post Fermi, Nvidia has castrated their fp64/compute part of their gpu to deliver better performance/watt for gaming. And they have been successful and rewarded by consumers for doing so. The most recent Pascal/Tesla showed a tripling down of their efforts. So its not like they only started "redesigning their newer gpus" to reflect changing trends.

If Nvidia feels that growing maturity of dx12 is going to impact markets, Volta would be a big change.

 

[bystander36]

It wouldn't make sense for Nvidia to have "no idea" on what is going on. They have more than a small inkling on the rising trends. It must have been a conscious design decision on Nvidia's part for about 3 generations now. Post Fermi, Nvidia has castrated their fp64/compute part of their gpu to deliver better performance/watt for gaming. And they have been successful and rewarded by consumers for doing so. The most recent Pascal/Tesla showed a tripling down of their efforts. So its not like they only started "redesigning their newer gpus" to reflect changing trends.

If Nvidia feels that growing maturity of dx12 is going to impact markets, Volta would be a big change.

With the original 600 series and 700 series, they weren't prepared for high compute. They started fixing that with the 900. With the 900 series, they weren't prepared for A-sync compute. With the 1000 series, they've taken big steps to fix that.

It seems pretty clear, that they were not prepared for these changes, or at least, were not expecting them to be a big deal so soon. Given how long it takes to design and produce a chip, they likely were just not ready for these changes, but these changes happened because dev's are designing their games to AMD's strengths, something Nvidia would be hard pressed to have known the details of, and which strengths dev's would favor.

I don't disagree that Volta will bring new improvements to further get on track, but I doubt we'll see anything as dramatic as Fermi's inability to keep up with modern titles.

 

[CHADBOGA]

New comparison from Hardware.fr:

Gainward GTX 1060 vs RX 480 Sapphire Nitro+: €280 match

- Power: Gainward GTX 1060 draws 83W less under load (TW3 Hairworks Off)
- Noise: 10.7 dBa less under load (!), 6.8 dBa less than Nitro+ at Silence Mode
- Overclock:
208 MHz GPU (+ 11%) and 2300 MHz memory (+ 15%) vs +1380 MHz (+ 3%) and 2250 MHz memory (+ 12.5%)
- Actual Gain from Overclocking (Overall): 12.3% vs 6.1%



www.hardware.fr/focus/118/rx-480-sapphire-nitro-vs-gtx-1060-gainward-match-280.html

Okay, so if I have read that website correctly, the 1060 when not overclocked, will on average be 10 - 12% faster, be quieter and use less electricity than the 480?

However the noise and electricity consumption difference isn't necessarily high enough to rule out buying a 480, with pricing being what it is likely to be.

Is that a fair assessment?

 

[Sonikku]

The pricing of the 480 has been a farce. If instocknow is any indication, the commanding majority of 480's to come in stock are almost exclusively the $240-$280 8gb variants. By the time you're paying $280 for a 480 all the jazzing up in the world doesn't make it a good value. Amd threw up this big $199 price tag with their big announcement, but I get the sneaking suspicion that they shipped a paltry token amount of 480's at $199 to technically meet the bare minimum of their promise when for all practical purposes this is a $240+ card.

Unless you burned the midnight oil and nabbed one bright and early on launch day chances are you aren't getting this sucker for $200, the price point at which it is the best value. The 1060 seems like a more reasonable alternative for its $250 asking price by the day. Now if only Nvidia can get their crap together with the api.

 

[YBS1]

Okay, so if I have read that website correctly, the 1060 when not overclocked, will on average be 10 - 12% faster, be quieter and use less electricity than the 480?

However the noise and electricity consumption difference isn't necessarily high enough to rule out buying a 480, with pricing being what it is likely to be.

Is that a fair assessment?

To be honest, the 1060 and the 480 seem to be pretty evenly matched. It's clear that for the most part the 1060 is the better DX11 card, and the 480 seems to be the better DX12 and Vulcan card as well as having a memory capacity advantage (not that that should come into play much as they are both suited to 1080p rather than higher resolutions, but some games do artificially limit your settings based on memory amount). I don't think someone could go wrong with either card, I just didn't want to fund nVidia so I went with the 480. I do think the 480 will end up being the faster card as they age based simply on recent history, I'm not sure I'd retain either card long enough for that to matter to me though. Right now the benchmarks bear out that the 1060 is indeed faster overall, the 480 counters by being less expensive. I think they are two of the better price/performance cards we've seen in a while (not counting fire sale cards 290X at crazy prices, etc.)

 

[CHADBOGA]

To be honest, the 1060 and the 480 seem to be pretty evenly matched. It's clear that for the most part the 1060 is the better DX11 card, and the 480 seems to be the better DX12 and Vulcan card as well as having a memory capacity advantage (not that that should come into play much as they are both suited to 1080p rather than higher resolutions, but some games do artificially limit your settings based on memory amount). I don't think someone could go wrong with either card, I just didn't want to fund nVidia so I went with the 480. I do think the 480 will end up being the faster card as they age based simply on recent history, I'm not sure I'd retain either card long enough for that to matter to me though. Right now the benchmarks bear out that the 1060 is indeed faster overall, the 480 counters by being less expensive. I think they are two of the better price/performance cards we've seen in a while (not counting fire sale cards 290X at crazy prices, etc.)

I've been using ATI/AMD cards for the best part of the last 10 years and have been happy with them, so I'm happy to stay with them again if I buy a new card this year, if they are competitive.

I'd give a weighting to performance in Doom, as that would be the main reason for why I would get a new GPU, probably around December.

I don't think the GPU prices have settled down yet in Australia as the 1060 is either the same price or cheaper than the 480, from one of the more popular retailers in this country.