AMD RX580 Reviews

XFX Radeon RX 580 8GB GTS Black Edition Review
http://www.overclockersclub.com/reviews/xfx_radeon_rx580_8gb_black_edition/

 

TechSpot just did what is probably the most extensive RX480 vs. RX580 vs. GTX1060 vs. GTX1060 9Gbps comparison, using no less than 27 games at 1080p and 1440p.



The higher-clocked memory seems to be doing pretty much nothing in the GTX1060, but they only benchmarked up to 1440p. The performance difference between a RX480 with 3rd party cooler (in this case it was a MSI Gaming X) and a RX580 seems mostly negligible to me.

Here is the chart that really matters:

 

Nice overall collection of titles to benchmark. Look forward to seeing the Ryzen version which is apparently coming.

 

need to say thatt i have not read the review in details, but the last graphic posted by TottenTranz is somewhat a bit depressive... you have nearly a half cut .. meaning if you choose games for review on the up part or the down part, you end with completely invert result. And for gamers, certainly everyone of thoses titles are somewhat valid.

 

You have 11 games that are 1% to 14% slower [6.9% slower on average] and have 15 games that are 1% to 18% faster [8.7% faster on average], meaning as an average you end up with better performance.

 

To me it is a split between them, but the prices of the new 1060 makes it less desirable to the 580 (where at least there is some chance to get 'moderate' priced custom 580s relative to latest custom 9Ghz 1060).
Benchmarks are a quirky business, especially when looking for best performance and testing uncapped/without frame limiting.
HardwareCanucks that also uses PresentMon (and also FCAT) shows Titanfall 2 actually marginally faster on the custom 1060 to the custom 580 where Steven W. at Techspot has it 10% faster for the 580.
Both sites using TSAA and highest settings.
And HardwareCanucks shows Gears of War 4 pretty even between custom 580 and custom 1060, yet Steven W. has 1060 over 10% faster.

The difference could be down to methodologies such as duration window and number of runs or down to different maps and chapters.
I assume the problem with greater duration tests is a higher likelyhood of inconsistent and loss of comparable benchmark runs between the various GPU review tests done by a site, but the problem is that games are far from consistent with dips or peaks lasting for a minute and a possibility to capture those rather than the actual overall performance.
It is a pain but these days seems to me the capture performance window needs to be around 3 minutes and due to the variation of whats on screen for such a window size needs to be repeated probably around 5 times minimum.
This shows a classic example where one can be caught either at the low or high point if duration is too short with uncapped frames, also shows the headache of reducing benchmark measurements just down to an fps score.



Suggests to me one needs at least 2 reliable sources-sites for correlating/validating results between manufacturers if reducing the numbers down to some kind of absolute number; more sites is better as less chance they all use same sequence/maps/etc for runs.

Cheers

 

Agreed, except for two titles: Warhamner is tested with the internal benchmark which gives the inverse picture of the actual gameplay from the game. And Ashes, no one really plays it or cares about it, there are more people playing the several years old Red Alert 3 than people playing Ashes. But I can understand it's relevance as one of the few titles with proper DX12 support.

 

I don't really see it that way. A typical difference of 5-10% with some 15-18% exceptions means that any current game will be similarly playable on either card from the same price point.
This is pretty comfortable for consumers, as there isn't an "obviously wrong" choice either way, save for some exceptions (e.g. being limited to a 250W PSU or owning a Freesync monitor). In this position, cards can just compete in price and that's great for us consumers.

There has been depressing days before, especially for people wanting to play the first generation of Gameworks titles, but this seems a lot better nowadays.

 

I wonder if there's an article comparing overclocked results? Since the 1060 has alot more headroom than the 580, I think it would make for a good competition.

 

Almost every graphics card review tests overclocking. There are plenty of articles like that.

This one covers some cards with factory overclocks and that's probably what the majority of consumers want to know. What's the percentage of people who overclock mid-end graphics cards?

 

The ones I've seen only benchmark a subset of games from the rest of the review for overclocked results. A review with a plethora of games like this but overclocked would be nice.

I can't speak for anyone else but I was planning to back when I was considering the 480/1060. Basically to ensure 60fps at 1080p since i don't have a gsync/freesync monitor.

 

Of course and I would too, but this is B3D, a forum for enthusiasts.

 

Review: Sapphire Radeon RX 580 Nitro+ in CrossFire
http://hexus.net/tech/reviews/graphics/107647-sapphire-radeon-rx-580-nitro-crossfire/

 

Interesting results, good to see Crossfire working well in some instances. Just 1 remark when I played TR (Crossfire RX480s) the terrible stuttering in the benchmark (especially at the start of Syria) wasn't present in gameplay.

 

http://www.phoronix.com/scan.php?page=article&item=16way-gpu-aug17&num=1

I thought AMD's drivers had terrible performance in Linux compared to nvidia, as so often people claim.
Turns out it's not the case Bioshock Infinite nor CS:Go, nor Deus Ex Mankind Divided (they made an OpenGL Linux version of this?!), nor Dawn of War 3, nor Civilization VI.
In fact, the only game they tested where AMD loses soundly is.. DOTA2..?

 

It might depend on the combination of tenses being used.
Following the chain of older reviews, progressively older driver stacks were increasingly far behind. It seems a number of notable performance and threading optimizations were not present until more recently.

It may help to cite specific claims and their dates to know whether they were off-base at the time.

 

Specific claims of what?
Claims of Nvidia cards traditionally punching above their AMD counterparts in Linux, particularly in the closed-sourced drivers?

And with dates?

 

It seems like AMD has improved significantly since the open source drivers have become the focus of driver development. The reputation it's had for falling short of Nvidia would have been earned over the time period prior.

To re-quote:

A literal interpretation would of that would be yes. In the past, AMD's drivers had poor performance.
The cited comparison is using recent Mesa Git drivers.
Going back several reviews, one can see how recently those improved versus the PRO drivers, and that is after AMD started investing more in development.
http://www.phoronix.com/scan.php?page=article&item=amdgpu-1730-radeonsi&num=1

It would be fair to remind people who may have stopped paying attention that AMD has made progress, but the claims that Nvidia traditionally did better are not without foundation.

 

NVIDIA is winning in Team Fortress 2, Dawn Of War 3, Dota 2 (Vulkan & OpenGL), and Metro Last Light.

Indeed. In fact, just a month ago, NVIDIA was well ahead in all Vulkan titles on Linux:

https://www.phoronix.com/scan.php?page=article&item=kblcpus-gl-vlk&num=2

 

And that's exactly the uphill battle AMD has to fight: Old(er) results left a lot to be desired and the tests' results are stuck in peoples heads now. Proves that, with peoples mentalities as a constant, you need to deliver from day 1.