Radeon RX 570 vs. RX 580 vs. GeForce GTX 1060 3GB vs. GTX 1060 6GB

[Nightfire]

I think the problem is you do not know the definition of presumably as if we did not expect Steve to put together a fair test.

Anyhow, here are some of the older RX 570 and 1060 3 GB comparisons:
https://www.techspot.com/review/1411-radeon-rx-570-vs-geforce-gtx-1060-3gb/page8.html

https://www.techspot.com/review/1660-geforce-gtx-1060-3gb-vs-radeon-rx-570-4gb/

Surprisingly, the 3gb 1060 is actually doing better against the rx570 than it was 2 years ago despite having less vram.

Still, most would agree the rx570 is the better buy.

 

[NightAntilli]

@ NightAntilli - Ok you made your point known. Nobody cares but you and you are now just trolling the thread. Way to show gratitude for the work done by Hardware Unboxed.

So you think because I have some criticism that somehow everything is being dismissed? Oh boy, aren't you rational...
My previous comments are at least on-topic and this exact comment by you is a lot closer to trolling than any of my previous ones, thanks.

In any case, out of all the GPUs, the one I would never advise anyone to get is the GTX 1060 3GB. Not only is it a deceptive product, the 3GB of RAM is simply not enough to use the card long-term.

 

[144hzGamer]

What annoys me of this article is;

1) The GTX 1060 6GB is put on top, as if it's the faster one, despite the RX 580 being faster in more games
2) Exactly that. The RX 580 is faster in most games (9 out of 12 to be exact), but the few outliers ruin the average framerate so both the GTX 1060 6GB and the RX 580 end up at 82 fps as if they're equal.
3) No chart was added with a direct games list comparison between the GTX 1060 6GB and the RX 580.

It almost seems as if the article is trying to hide that the GTX 1060 6GB is slower than the RX 580 in most games. At least the final recommendation is in line with the numbers.

What annoys me even more is not mentioning every gtx 1060 does 2000mhz on the core and destroys rx580 that way while still using less power and having better temps. Is almost like the article is trying to hide it. Ironic.

 

[Puiu]

What annoys me of this article is;

1) The GTX 1060 6GB is put on top, as if it's the faster one, despite the RX 580 being faster in more games
2) Exactly that. The RX 580 is faster in most games (9 out of 12 to be exact), but the few outliers ruin the average framerate so both the GTX 1060 6GB and the RX 580 end up at 82 fps as if they're equal.
3) No chart was added with a direct games list comparison between the GTX 1060 6GB and the RX 580.

It almost seems as if the article is trying to hide that the GTX 1060 6GB is slower than the RX 580 in most games. At least the final recommendation is in line with the numbers.

What annoys me even more is not mentioning every gtx 1060 does 2000mhz on the core and destroys rx580 that way while still using less power and having better temps. Is almost like the articlr is trying to hide it. Ironic.

No it is not. Not every 1060 OCs that well and very few OC their GPU to the max.
If you OC both the 580 and the 1060 you will maybe gain 2-3% in favor of the 1060 which still makes the 580 a better buy unless you have a specific game in mind that plays better on NVIDIA hardware.

People really need to stop with the stupid exagerations.

 

[144hzGamer]

No it is not. Not every 1060 OCs that well and very few OC their GPU to the max.
If you OC both the 580 and the 1060 you will maybe gain 2-3% in favor of the 1060 which still makes the 580 a better buy unless you have a specific game in mind that plays better on NVIDIA hardware.

People really need to stop with the stupid exagerations.

Every gtx 1060 does 2000mhz. Not all of them do 2100mhz. At 2000mhz goodbye RX580 while using less power. Facts. Check reviews.

 

[Puiu]

Every gtx 1060 does 2000mhz. Not all of them do 2100mhz. At 2000mhz goodbye RX580 while using less power. Facts. Check reviews.

I checked them and what I said still stands.

and again with the childish exagerations. the power draw of these two GPUs is low enough to be ignored. at most during peak usage the 580 will use between 30 to 50W more and no GPU stays at that peak usage more than a few moments.

in the end you are looking at savings of less than the price difference within a year.

You are literally arguing about a cup of coffe which is really weird. If this is your main argument, that you can gain a few % when you OC and a cup of coffee, then it kinda falls flat on its face. The average smart Joe doesn't care about any of those. You don't even save money on the PSU, you are fine with a 450W one.

FACT: real life usage beats arguments that rely on 2-3% and chump change.

Here are a few facts that you should use: esport titles generally play better with Nvidia, many productivity software support CUDA better than OpenCL, Nvidia GPUs are easier to resell, etc etc.
Use the ones above, not exagerations and other childish statements.

 

[Evernessince]

What annoys me of this article is;

1) The GTX 1060 6GB is put on top, as if it's the faster one, despite the RX 580 being faster in more games
2) Exactly that. The RX 580 is faster in most games (9 out of 12 to be exact), but the few outliers ruin the average framerate so both the GTX 1060 6GB and the RX 580 end up at 82 fps as if they're equal.
3) No chart was added with a direct games list comparison between the GTX 1060 6GB and the RX 580.

It almost seems as if the article is trying to hide that the GTX 1060 6GB is slower than the RX 580 in most games. At least the final recommendation is in line with the numbers.

What annoys me even more is not mentioning every gtx 1060 does 2000mhz on the core and destroys rx580 that way while still using less power and having better temps. Is almost like the article is trying to hide it. Ironic.

If every GTX 1060 did 2000 Mhz then Nvidia should have made them boost that high out of the box. Otherwise reviews are based on stock specs, not arbitrary silicon lottery.

 

[LogiGaming]

Every gtx 1060 does 2000mhz. Not all of them do 2100mhz. At 2000mhz goodbye RX580 while using less power. Facts. Check reviews.

I checked them and what I said still stands.

and again with the childish exagerations. the power draw of these two GPUs is low enough to be ignored. at most during peak usage the 580 will use between 30 to 50W more and no GPU stays at that peak usage more than a few moments.

in the end you are looking at savings of less than the price difference within a year.

You are literally arguing about a cup of coffe which is really weird. If this is your main argument, that you can gain a few % when you OC and a cup of coffee, then it kinda falls flat on its face. The average smart Joe doesn't care about any of those. You don't even save money on the PSU, you are fine with a 450W one.

FACT: real life usage beats arguments that rely on 2-3% and chump change.

Here are a few facts that you should use: esport titles generally play better with Nvidia, many productivity software support CUDA better than OpenCL, Nvidia GPUs are easier to resell, etc etc.
Use the ones above, not exagerations and other childish statements.

I´ve never seen a single GTX 1060 that doesn´t do 2000mhz on the core. In fact most even do 2100mhz, but to be on the safe side, 2000mhz is indeed a magic number.

At 2000mhz RX580 is obliterated. And power consumption is alwas a plus. Not the main selling factor, but if one card can be faster than the other while using less power, is always good. Anything else is just enter "denial mode".

 

[LogiGaming]

What annoys me of this article is;

1) The GTX 1060 6GB is put on top, as if it's the faster one, despite the RX 580 being faster in more games
2) Exactly that. The RX 580 is faster in most games (9 out of 12 to be exact), but the few outliers ruin the average framerate so both the GTX 1060 6GB and the RX 580 end up at 82 fps as if they're equal.
3) No chart was added with a direct games list comparison between the GTX 1060 6GB and the RX 580.

It almost seems as if the article is trying to hide that the GTX 1060 6GB is slower than the RX 580 in most games. At least the final recommendation is in line with the numbers.

What annoys me even more is not mentioning every gtx 1060 does 2000mhz on the core and destroys rx580 that way while still using less power and having better temps. Is almost like the article is trying to hide it. Ironic.

If every GTX 1060 did 2000 Mhz then Nvidia should have made them boost that high out of the box. Otherwise reviews are based on stock specs, not arbitrary silicon lottery.

Yes, because when an i7 8700k is officially only up to 4,7ghz means it can´t do 4,9ghz at all right? Oh wait...

That´s new to me. The fact the boost clock doesn´t go automatically to 2000mhz out of the box doesn´t mean most can´t reach it. I´ve yet to see a gtx 1060 that can´t do 2000mhz and a lot of them actually do 2100mhz. He was simpling stating the fact that GTX 1060 are great overclockers. Dismissing that on a hardware website review is just silly. Is as easy as moving a slider on MSI afterburner.

But if AMD cards could overclock almost up to 20% from their base peformance holy cow... everyone would brag about it everywhere. Usual business. I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes.

 

[Nightfire]

The Rx 580 mech 2 and 590s overclock pretty good, but I still think the GTX 1060 would top them with a good overclock, especially with the gddr5x.
Also, I hate everyone.

 

[Evernessince]

Yes, because when an i7 8700k is officially only up to 4,7ghz means it can´t do 4,9ghz at all right? Oh wait...

That´s new to me. The fact the boost clock doesn´t go automatically to 2000mhz out of the box doesn´t mean most can´t reach it. I´ve yet to see a gtx 1060 that can´t do 2000mhz and a lot of them actually do 2100mhz. He was simpling stating the fact that GTX 1060 are great overclockers. Dismissing that on a hardware website review is just silly. Is as easy as moving a slider on MSI afterburner.

But if AMD cards could overclock almost up to 20% from their base peformance holy cow... everyone would brag about it everywhere. Usual business. I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes.

Realize the fact that base and boost clocks are decided by silicon quality and how many good dies the company wants to get out of a product. The max boost clock is the clock that Intel/AMD/Nvidia decided on and is based off yields. It is indeed the number every unit sold can reach under the right conditions. If Intel could reach higher clocks on EVERY sample as you claim with zero loss in yields then they would have increased the max boost clock. Clearly they could not and thus we arrived where we are now. So yes, I think I will take Intel's word over your's. It's not like you could have provided evidence to such a ridiculous claim as EVERY GTX 1060 reaching that frequency when clearly Nvidia itself knew that EVERY GPU would not reach higher frequencies.

"I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes"

Preconceived bias. You come to this thread taking issue with the article and other members and lodged a claim with zero evidence and now you have the gull to blow off everyone else as "AMD bots"? That's a cheap cop out way to put off everyone who doesn't agree with you while providing nothing to the argument. I suggest you gather your evidence and field a comment that is supported by something first instead of trying to drag everyone else's down.

 

[CaptainTom]

The 570 is closer to a 1060 6GB in performance than the 1050 Ti.... and yet it is being sold for less than the 1050 Ti.

I pray that some day the general PC gaming public will grow a bloody brain and start buying good products...

 

[144hzGamer]

Realize the fact that base and boost clocks are decided by silicon quality and how many good dies the company wants to get out of a product. The max boost clock is the clock that Intel/AMD/Nvidia decided on and is based off yields. It is indeed the number every unit sold can reach under the right conditions. If Intel could reach higher clocks on EVERY sample as you claim with zero loss in yields then they would have increased the max boost clock. Clearly they could not and thus we arrived where we are now. So yes, I think I will take Intel's word over your's. It's not like you could have provided evidence to such a ridiculous claim as EVERY GTX 1060 reaching that frequency when clearly Nvidia itself knew that EVERY GPU would not reach higher frequencies.

"I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes"

Preconceived bias. You come to this thread taking issue with the article and other members and lodged a claim with zero evidence and now you have the gull to blow off everyone else as "AMD bots"? That's a cheap cop out way to put off everyone who doesn't agree with you while providing nothing to the argument. I suggest you gather your evidence and field a comment that is supported by something first instead of trying to drag everyone else's down.

1- they obviously ignored oc potential on the gtx1060 on this analysis. Thats why nvidia stopped sending GPUs to these reviewrs day 1, because they always bias their reviews.

2- Boost clocks on Pascal are not 100% related to silicon quality and you can manually clock some GPUs to 2100mhz when they dont even reach 1900mhz out of the box.

3- Yes, someone stating not all gtx1060 can reach 2000mhz is an AMD bot.

4- Ive seen you countless times on nvidia related news spreading toxicity to every user.

 

[144hzGamer]

Yes, because when an i7 8700k is officially only up to 4,7ghz means it can´t do 4,9ghz at all right? Oh wait...

That´s new to me. The fact the boost clock doesn´t go automatically to 2000mhz out of the box doesn´t mean most can´t reach it. I´ve yet to see a gtx 1060 that can´t do 2000mhz and a lot of them actually do 2100mhz. He was simpling stating the fact that GTX 1060 are great overclockers. Dismissing that on a hardware website review is just silly. Is as easy as moving a slider on MSI afterburner.

But if AMD cards could overclock almost up to 20% from their base peformance holy cow... everyone would brag about it everywhere. Usual business. I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes.

This! When AMD has some kind of banger no fanboy will shut up about it. Is just annoying

 

[Davron]

Where is the fact that RX580 consumes a lot more power? The TDP is higher in the AMD card at 185 watts while the GTX 1060 6Gb is only 120 watts. And ,in spite of that, it goes head to head with AMD RX580.
You only know how to complain, but praise, ufff... that's really hard to do, isn't it?
And then comes the conspiracy theories that nvidia buys people to have better reviews than AMD, because of course that's the only explanation...
They are both great cards and they do a great job at 1080p and 1440p.

Power consumption only matters when it's AMD. When nVidia had inferior power consumption for the same (or even worse) performance, everyone still bought nVidia. To me, power consumption is an excuse, not an argument. Not until the one complaining about it can show how it is indeed relevant.

I agree with you that at this power level it is basically moot, but when my two mildly overclocked and water cooled R9 295x2s with the accompanying 5930k took a 1.5kW power supply according to the power supply calculators, power did matter. That room got really hot in the summer when playing games or when the GPUs were used for other compute acceleration.

 

[CaptainTom]

1- they obviously ignored oc potential on the gtx1060 on this analysis. Thats why nvidia stopped sending GPUs to these reviewrs day 1, because they always bias their reviews.

2- Boost clocks on Pascal are not 100% related to silicon quality and you can manually clock some GPUs to 2100mhz when they dont even reach 1900mhz out of the box.

3- Yes, someone stating not all gtx1060 can reach 2000mhz is an AMD bot.

4- Ive seen you countless times on nvidia related news spreading toxicity to every user.

Man you don't know a lot lol. No Nvidia fan boy should want to bring overclocking into the equation. The 1060 is lucky to get 10% more performance even when you get a lucky yield at 2.1 GHz. However it is well known that Polaris can get a 15-20% boost when you overclock the ram and hit 1500Mhz on the core. Especially in the case of the excellent overclocking 570.

 

[Nightfire]

1- they obviously ignored oc potential on the gtx1060 on this analysis. Thats why nvidia stopped sending GPUs to these reviewrs day 1, because they always bias their reviews.

Damn, that is not going help you win any arguments here.

As it stands now, the Rx580 is 21% cheaper than the 6gb 1060. O/c vs o/c the 1060 may be ahead, but nowhere near that price difference. There is nothing bias about that.

And if you must know, Nvidia most likely stiffed Hardware unboxed due to the DLSS criticism. Like many others have found, 4k using dlss is closer in image quality to 1800p with TAA and close in performance as well. So really, nothing gained.

 

[Nightfire]

To keep the children from arguing further, I will do the work to settle this. I just tore my Achilles, so I really have nothing better to do.

Here are overclocking examples of the RX580 (mech 2) and GTX 1060 (9 gbps). Both are from MSI and both represent some great results for their respective gpus:

https://www.guru3d.com/articles_pages/msi_radeon_rx_570_and_580_mech_2_8g_oc_review,36.html

https://www.guru3d.com/articles_pages/msi_geforce_gtx_1060_gaming_x_plus_review,40.html

As you can see, both see about 10-15% actual game improvement over referance. Not exactly "good bye rx580".

The RX580 is a better value. The GTX1060 offers better thermals and efficiency among other things. Move on people.

 

[Evernessince]

1- they obviously ignored oc potential on the gtx1060 on this analysis. Thats why nvidia stopped sending GPUs to these reviewrs day 1, because they always bias their reviews.

2- Boost clocks on Pascal are not 100% related to silicon quality and you can manually clock some GPUs to 2100mhz when they dont even reach 1900mhz out of the box.

3- Yes, someone stating not all gtx1060 can reach 2000mhz is an AMD bot.

4- Ive seen you countless times on nvidia related news spreading toxicity to every user.

None of this disputes anything I've said nor does it provide any proof for your original claim. If you choose to ignore the facts on how Intel, AMD, and Nvidia decide clock speeds that's up to you. Once again, I'm waiting for you to prove every card can hit 2000 MHz...

"Boost clocks on Pascal are not 100% related to silicon quality"

and you just lost any last bit of credibility you might have had. Why don't you go ask steve or tim about that one. Boost clocks ARE 100% related to silicon quality.

You do realize that TechSpot published overclocking results with a 1060 multiple times in the past right? The scope of this article never included overclocking. Is it hard to google "Techspot 1060 overclock"? I fail to see how TechSpot is biased because you can't use a keyboard.

 

[RageofThunder3]

Yes, because when an i7 8700k is officially only up to 4,7ghz means it can´t do 4,9ghz at all right? Oh wait...

That´s new to me. The fact the boost clock doesn´t go automatically to 2000mhz out of the box doesn´t mean most can´t reach it. I´ve yet to see a gtx 1060 that can´t do 2000mhz and a lot of them actually do 2100mhz. He was simpling stating the fact that GTX 1060 are great overclockers. Dismissing that on a hardware website review is just silly. Is as easy as moving a slider on MSI afterburner.

But if AMD cards could overclock almost up to 20% from their base peformance holy cow... everyone would brag about it everywhere. Usual business. I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes.

This! When AMD has some kind of banger no fanboy will shut up about it. Is just annoying

Why do you have your two accounts "talking" to each other like this? Thinking this scam is gonna make your arguments more convincing?

Btw I can say the same thing to you:
Usual business. I have nothing against Nvidia, but Nvidia bots are the most annoying ones and make me hate Nvidia sometimes.

 

[144hzGamer]

Why do you have your two accounts "talking" to each other like this? Thinking this scam is gonna make your arguments more convincing?

Btw I can say the same thing to you:
Usual business. I have nothing against Nvidia, but Nvidia bots are the most annoying ones and make me hate Nvidia sometimes.

Two accounts?? Ahah how hilarious !! Why don't you ask mods about that statement?

Your logic is: no more than 1 guy can have a different opinion from mine, thus it is scam.

Reported!

 

[Puiu]

Realize the fact that base and boost clocks are decided by silicon quality and how many good dies the company wants to get out of a product. The max boost clock is the clock that Intel/AMD/Nvidia decided on and is based off yields. It is indeed the number every unit sold can reach under the right conditions. If Intel could reach higher clocks on EVERY sample as you claim with zero loss in yields then they would have increased the max boost clock. Clearly they could not and thus we arrived where we are now. So yes, I think I will take Intel's word over your's. It's not like you could have provided evidence to such a ridiculous claim as EVERY GTX 1060 reaching that frequency when clearly Nvidia itself knew that EVERY GPU would not reach higher frequencies.

"I have nothing against AMD, but AMD bots are the most annoying ones and make me hate AMD sometimes"

Preconceived bias. You come to this thread taking issue with the article and other members and lodged a claim with zero evidence and now you have the gull to blow off everyone else as "AMD bots"? That's a cheap cop out way to put off everyone who doesn't agree with you while providing nothing to the argument. I suggest you gather your evidence and field a comment that is supported by something first instead of trying to drag everyone else's down.

1- they obviously ignored oc potential on the gtx1060 on this analysis. Thats why nvidia stopped sending GPUs to these reviewrs day 1, because they always bias their reviews.

2- Boost clocks on Pascal are not 100% related to silicon quality and you can manually clock some GPUs to 2100mhz when they dont even reach 1900mhz out of the box.

3- Yes, someone stating not all gtx1060 can reach 2000mhz is an AMD bot.

4- Ive seen you countless times on nvidia related news spreading toxicity to every user.

"not 100% related to silicon quality" - I'm sorry but what? are you even trying to tell a coherent story? now I'm just confused by this kind of statement from someone who claims to be even a little bit tech-savvy O_o

 

[144hzGamer]

"not 100% related to silicon quality" - I'm sorry but what? are you even trying to tell a coherent story? now I'm just confused by this kind of statement from someone who claims to be even a little bit tech-savvy O_o

I mean the factory boost clocks. That user is implying that if gtx 1060 did 2000mhz they would boost to that value out of the box and thats totally not true. The default boost clock is always lower than what you can achieve manually. He just added up one more silly and wrong argument to the big list now.

 

[Peter Farkas]

Seeing this now I am very happy with my RX 580. Very good review on these cards especially taking the current prices into account. Cost per frame metrics really tell the truth! Good job!

 

[Dimitrios]

To the writer,

"Last Call: What's the Best Mainstream GPU?"

Probably one of the best headlines I have seen in GPU reviews. Very catchy.