Desktop GPU Performance Hierarchy Table (Archive)

[cangelini]

Here is a resource to help you judge if a graphics card is a reasonable value: The gaming GPU hierarchy chart groups GPUs by performance.

Desktop GPU Performance Hierarchy Table : Read more

 

[adamovera]

Archived comments are found here: http://www.tomshardware.com/forum/id-2898695/desktop-gpu-performance-hierarchy-table.html

 

[rush21hit]

Wait, so it's really true HD5850 weaker than 750Ti? Haven't look for actual data but I'd just take what Tom's said here. And now I feel so old...
*I remember the times I had to do some serious fasting to buy it. Also did the same thing for HD4670 before that and X1650XT before that too.

 

[AndrewJacksonZA]

Hi guys

Thank you for keeping your chart updated, it's useful for me and I'm sure for many other people too.

I don't understand your reasoning for putting the GTX 1060 one tier above the RX 480. Please could you explain it to me? Based on benchmark evidence from your 1060 review http://www.tomshardware.com/reviews/nvidia-geforce-gtx-1060-pascal,4679-3.html as well as other reviews on the web (because you guys didn't include a Vulkan API in your benchmarks like, saaaaayyyyyy... Doom) I disagree. I would put them on the same tier.

If I had to only focus on DX11, I'd put the 1060 on a tier above the 480.
If I had to only focus on DX12 and Vulkan, I'd put the 480 on a tier above the 1060.
If I had to only focus on all three, I think that the same tier is fair.

In case any readers are looking for a Doom benchmark and somehow hasn't read one yet, go to [H]ardOCP: http://www.hardocp.com/article/2016/07/19/nvidia_geforce_gtx_1060_founders_edition_review/4
I'm pointing you to that site specifically because they seem, to me, to be a site that's biased towards Nvidia, but they do have the integrity to be fair and honest when the evidence demonstrates the opposite, as it is in this case.

@Toms team regarding Doom:
On 19 July 2016 Igor said that you guys were waiting on id to release a patch to enable Vulkan async compute on Nvidia hardware.

On 23 July 2016 Jason Leavey posted that Vulkan is now supported on AMD and Nvidia hardware with the appropriate drivers - it's just that async compute isn't supported on Nvidia hardware yet but there'll be an update "soon," just as Igor said.

That was over a month ago. Has id given any evidence that the patch is coming "soon?" Any date estimate whatsoever, perhaps? Is it ever coming out? Please add the Doom Vulkan benchmarks.

 

[Cavalierhero]

The 1060 is definitely not a tier above the rx480.

 

[dasper]

I am wondering the same as ANDREWJACKSONZA. Just today I was reading/watching reviews of Deus Ex and Doom on how the 480 is faster than the 1060 and yet it is put on a lower tier. Both cards were released close to the same time on the latest technology in the same price tier. What is the excuse for them not being on the same level?

 

[ffleader1]

How much Nvidia pay you to do this, exactly?
You put 390X and 290 on the same tier, 1060 is above 480, and 1060 is equal to a Fury?

 

[iispyderii]

http://gpu.userbenchmark.com/Compare/AMD-R9-390X-vs-AMD-R9-290/3497vs2171

The 390X is that much better than the 290 and they are on the same tier but the 1060 and the 480 ARE NOT!? You guys looking like major shills here.

 

[John_455]

What a joke, this ranking is clearly biased towards nvidia. The 1060 is junk compared to the 480.

 

[uglyduckling81]

The 1060 isn't junk. It's a little slower than a 980. It might be a spot too high on the list but then again it does outperform a 970 in every title which a 480 can't claim.
http://www.tomshardware.com/reviews/nvidia-geforce-gtx-1060-pascal,4679-3.html

 

[AndrewJacksonZA]

@dasper: I just read some Deus Ex:MD performance reviews and I agree.

@ugliyduckling81: While the 1060 can perform better than the 970 in all of those tests, it gets beaten by the 480 in others.

 

[AlexFromAU]

GTX 1070, GTX 980 Ti and R9 Fury X should all be on the same tier - as they all have same performance.

GTX 1060 is NOT better than a RX 480

 

[Retrogame]

I find this a handy reference. Reading various benchmark reviews (and waiting for an updated set of GPU charts!) you can see that your mileage varies depending on the application. When it comes to this chart, remember its purpose, it's to show you about how many models up the food chain you need to go before, in the editors' opinion, the upgrade is worth spending any money to do.

I would love to see this suggested addition: Beside the name of each model, if it could also show month/year of the release into the market of the chipset. It would help provide a bit more context.

For example, the computer I am on as I write this has a R9 380 GPU. It's in the same ballpark as the HD 6990.

So you have R9 380 (6/2015) with a Tonga rebrand chip, almost same as R9 285 (9/2014) almost same as dual gpu Antillies HD 6990 (3/2011) But whereas the 380 is a 2015 "mainstream" card launched for $200 or so in 2015, the 6990 was a dual gpu monstrosity that cost $700 four years earlier and provided about the same performance, but through single-card crossfire, plus the issues of more heat and power consumption. Quite a big difference in a 4 year gap. And beside it is GTX 770, a Kepler series chip updated and repackaged in 5/2013 for $400. I find such comparisons of generations and pricing (and the Red Vs Green war!) quite interesting and educational.

To circle back, this is why a table like this is very helpful. An upgrade is only an upgrade if actual performance is substantially better than what you're running after cutting through the marketing hype! I'm GPU and CPU agnostic, I build platforms that use anything.

 

[ffleader1]

I find this a handy reference. Reading various benchmark reviews (and waiting for an updated set of GPU charts!) you can see that your mileage varies depending on the application. When it comes to this chart, remember its purpose, it's to show you about how many models up the food chain you need to go before, in the editors' opinion, the upgrade is worth spending any money to do.

I would love to see this suggested addition: Beside the name of each model, if it could also show month/year of the release into the market of the chipset. It would help provide a bit more context.

For example, the computer I am on as I write this has a R9 380 GPU. It's in the same ballpark as the HD 6990.

So you have R9 380 (6/2015) with a Tonga rebrand chip, almost same as R9 285 (9/2014) almost same as dual gpu Antillies HD 6990 (3/2011) But whereas the 380 is a 2015 "mainstream" card launched for $200 or so in 2015, the 6990 was a dual gpu monstrosity that cost $700 four years earlier and provided about the same performance, but through single-card crossfire, plus the issues of more heat and power consumption. Quite a big difference in a 4 year gap. And beside it is GTX 770, a Kepler series chip updated and repackaged in 5/2013 for $400. I find such comparisons of generations and pricing (and the Red Vs Green war!) quite interesting and educational.

To circle back, this is why a table like this is very helpful. An upgrade is only an upgrade if actual performance is substantially better than what you're running after cutting through the marketing hype! I'm GPU and CPU agnostic, I build platforms that use anything.

The previous Hierarchy table was helpful. However, are you telling me that selling a Rx 480 and buy a 1060 is an upgrade, but selling a 290 to buy a 390X is not an upgrade?
For some reason, they messed this one up, with Nvidia bias.

 

[penn919]

I believe the rule of thumb was that you shouldn't upgrade unless the graphics card is at least two tiers above your existing card; therefore, they wouldn't recommend upgrading to a 1060 from an RX 480. I personally wouldn't upgrade unless the graphics card is at least three tiers above or more. ie, HD 7950 --> R9 Nano

 

[penn919]

I would like to see the AMD Radeon Pro Duo Added to this chart.

 

[Ahmedivx]

Dear Chris Angelini (the author of this article), you can never be more biased, the way you showed it in this article, some of the tiers are really messed up and it's only there to mess with people judgment about which cards are more powerful, needs serious adjustments to match real actual performance of the cards

 

[Ravi_Singh1900]

Hi, I have used gtx 960 earlier and I think it should be in league above gtx 760, r7 270x. In my opinion leaving a few games it doesn't perform as good as a r9 280 or r9 285 or gtx 670 or gtx 580.

 

[ozymankos]

Your GPU list is extremely biased against AMD...you put the GTX 1080 and 1070 on 2 different categories,yet you put Radeon 480 and 470 on the same category...
Radeon Pro Duo is not even mentioned and has 8192 graphical processors(stream units) while Titan X from Nvidia has only 3072 graphical processors(cuda cores)

 

[DSpider]

The RX 480 and 470 are correctly placed on the same tier. They perform almost identical, with a slight overclock the RX 470 beating the 480.

I'm a bit surprised that the R9 295X2 can keep up with the GTX 1070. I thought AMD didn't have anything to compete against the GTX 1070/1080 lineup. Hmph. *shrug*

 

[DSpider]

The reason why the GTX 1060 isn't on the same tier as the RX 480 is because the RX 480 isn't even close to the GTX 980. More like close to the 970. Hence, one tier down.

 

[Tomas_fireball]

Is this a joke? the R9 295x2 beats the GTX 1080 hands down. I have both and the nvidia is way slower. AMD should sue you guys for defamation

 

[PJ5301]

Where is AMD's ProDuo on this?

 

[bballbreit234]

I get that the AMD vs nVidia argument inspires a lot of anger and fanboy-ism, but the benchmarks are pretty clear. GTX 1060 beats the RX 480 in most DX 11 benchmarks, and the RX 480 beats the GTX 1060 in most DX 12 and Vulkan benchmarks. Most games today and for the next few years will be using DX 11. Hence, the GTX 1060 is a tier higher.

 

[PJ5301]

I get that the AMD vs nVidia argument inspires a lot of anger and fanboy-ism, but the benchmarks are pretty clear. GTX 1060 beats the RX 480 in most DX 11 benchmarks, and the RX 480 beats the GTX 1060 in most DX 12 and Vulkan benchmarks. Most games today and for the next few years will be using DX 11. Hence, the GTX 1060 is a tier higher.

I think you replied to the wrong person also they should be on the same tier due to the RX 480 performing much better in DX12 and not that far off in DX11 though personally I would go for the 1060 due to it performing much better in Arma 3