AMD Radeon RX 580: Prey You'll Upgrade

Manufacturer: AMD

Is it time to buy that new GPU?

Testing commissioned by AMD. This means that AMD paid us for our time, but had no say in the results or presentation of them.

Earlier this week Bethesda and Arkane Studios released Prey, a first-person shooter that is a re-imaging of the 2006 game of the same name. Fans of System Shock will find a lot to love about this new title and I have found myself enamored with the game…in the name of science of course.

View Full Size

While doing my due diligence and performing some preliminary testing to see if we would utilize Prey for graphics testing going forward, AMD approached me to discuss this exact title. With the release of the Radeon RX 580 in April, one of the key storylines is that the card offers a reasonably priced upgrade path for users of 2+ year old hardware. With that upgrade you should see some substantial performance improvements and as I will show you here, the new Prey is a perfect example of that.

Targeting the Radeon R9 380, a graphics card that was originally released back in May of 2015, the RX 580 offers substantially better performance at a very similar launch price. The same is true for the GeForce GTX 960: launched in January of 2015, it is slightly longer in the tooth. AMD’s data shows that 80% of the users on Steam are running on R9 380X or slower graphics cards and that only 10% of them upgraded in 2016. Considering the great GPUs that were available then (including the RX 480 and the GTX 10-series), it seems more and more likely that we going to hit an upgrade inflection point in the market.

View Full Size

A simple experiment was setup: does the new Radeon RX 580 offer a worthwhile upgrade path for those many users of R9 380 or GTX 960 classifications of graphics cards (or older)?

  Radeon RX 580 Radeon R9 380 GeForce GTX 960
GPU Polaris 20 Tonga Pro GM206
GPU Cores 2304 1792 1024
Rated Clock 1340 MHz 918 MHz 1127 MHz
Memory 4GB
Memory Interface 256-bit 256-bit 128-bit
TDP 185 watts 190 watts 120 watts
MSRP (at launch) $199 (4GB)
$239 (8GB)
$219 $199

Continue reading our look at the Radeon RX 580 in Prey!

For my testing I am using our Frame Rating capture-based performance method to see how much performance improvement we see in Prey. The target this time was to see if users going for this upgrade could reliably run at 2560x1440 resolution at the Very High quality preset.

View Full Size

Here are the results.

View Full Size

View Full Size

View Full Size

View Full Size

Clearly the new Radeon RX 580 8GB graphics card produces a better overall gaming experience in Prey than the R9 380 and the GTX 960. The difference between being able to run this new PC title at 60+ FPS at 2560x1440 or NOT, with high image quality settings enabled, is a mid-range graphics card upgrade.

Radeon RX 580 8GB, Average FPS Comparisons, Prey
  Radeon R9 380 4GB GeForce GTX 960 2GB
2560x1440 +74% +60%

The new Radeon RX 580 is 74% faster than the R9 380 and 60% faster than the GTX 960 in Prey, at Very High image quality settings at 2560x1440. That is not an insignificant delta!!

Prey is just one game, but this trend is something that AMD has been promoting since before the RX 500-series hit the street. Older GPUs and graphics cards based on them continue to fall behind in terms of both performance and power efficiency. If you are planning to run at resolutions higher than 1080p, or even if you want to run at higher refresh rates than 60 Hz at 1080p, then it may be time to start looking for that replacement hardware.

View Full Size

For the vast majority of consumers that purchase hardware in the $250 and below range, AMD’s new Radeon RX 580 family is a great choice.

May 17, 2017 | 10:36 AM - Posted by Kelamin

Why was this not tested against a Nvidia video card of a similar price to the 580 as a comparison? Or was that part of the arrangement with AMD?

May 17, 2017 | 10:42 AM - Posted by arbiter

I wonder as well why a gtx960 was used since that card is what 2 years old? Had a look a newegg just now and gtx1060's start at 180$ so that would be better pick to compare with.

May 17, 2017 | 10:44 AM - Posted by Kelamin

I figured the 960 as it might be considered a counterpart to the old AMD card that was tested.

May 17, 2017 | 10:57 AM - Posted by arbiter

Still odd without a gtx1060 since that is same price point, $160(3gb) and 220$(6gb) starting on newegg.

Other thing is why 1440p? Most market for these cards is 1080p.

edit: as I look back at some gtx1060 reviews that show results of a gtx960, i can see why it wasn't included.

May 17, 2017 | 02:28 PM - Posted by Bianchi4Me

Yeah, an 8gb card (as tested) is overkill for 1080p. Presumably, the driving force behind a desire to upgrade for a lot of users IS moving up from 1080p to 1440p displays. Honestly, if you're sticking with running 1080p there probably isn't much point in upgrading from a 390/960 level card. Doubly so for people running a 60 fps capped monitor, which is still the majority of 1080p gamers. Buying a new card to get theoretical frame rates way higher than your old monitor can display is obviously silly, so not surprised they did not test that usage case.

May 17, 2017 | 05:36 PM - Posted by Isaac Johnson

Most modern cards are able to keep 60fps at 1920x1080 if you're willing to mess around with the detail settings, but when you bring desired quality into the equation (which is a completely subjective thing) that goes out the window. I'm running a 390 at 1920x1080 and 70hz and many of the games I play, at the detail settings I wish to play them, can't keep a minimum of 70fps or higher, and several of them are a lot lower. In my opinion even a 1080ti isn't overkill for 1920x1080 gaming. But it all depends on what you play and how you play.

May 17, 2017 | 11:14 AM - Posted by Ryan Shrout

Guys, I love you but the point of the article is there several times:

"does the new Radeon RX 580 offer a worthwhile upgrade path for those many users of R9 380 or GTX 960 classifications of graphics cards (or older)?"

May 17, 2017 | 11:40 AM - Posted by Kelamin

Without comparison to other cards though it boils down to, are cards of the same tier now better than cards of 2 years ago? It just seems like a pointless premise.

May 17, 2017 | 04:04 PM - Posted by Bianchi4Me

Maybe it makes some sense based on AMD's history of re-branding the same cards through three successive generations. They might need to convince people there IS actually some difference between their current stuff and what they were selling years ago.

May 18, 2017 | 03:33 PM - Posted by Pinscher

I agree with this comment. Re-branding just waters down the product and causes confusion in the hierarchy. Additional it holds very little value to the brand itself to release cards with little performance difference. Also I'd mention that there is nothing wrong with training a product name beyond 1 year as cards have a multi year life cycle anyways.

If AMD took 2 or 3 years to release a wicked card, then I'd be far happier then being spammed year after year of rebrands

May 17, 2017 | 12:07 PM - Posted by Jtaylor1986

Yes because you have to create alternate reality scenarios where you competitor's current products don't exist when the 1060 is faster in Prey. I'm an AMD fan and I'm still kind of disappointed you let AMD use you like this. Is this that big a revelation that current generation GPU's on a smaller process node are faster than previous generation GPU's on a 5 year old process node?

May 17, 2017 | 03:00 PM - Posted by rock1m1

Agree with everything you said. I can't believe this just happened to one of most reliable sources of tech news right now.

May 17, 2017 | 11:22 AM - Posted by CNote

The point is to see if your old ass graphics card needs to be replaced... not unlike if you still drive a Ford Tempo.

May 17, 2017 | 02:15 PM - Posted by Alamo

because AMD s trying to grab the TAM of old GPUs, and move them to a more DX12 compatible cards.

May 17, 2017 | 10:54 AM - Posted by rock1m1

Look, I know you are paid. But this comparison makes no f**** sense lmao.

May 17, 2017 | 10:56 AM - Posted by arbiter

Well when you look back at AMD released slides you will see the trend of their comparison's like this.

May 17, 2017 | 10:56 AM - Posted by khanmein

@arbiter, This video sponsored by AMD. I hope this answered your question. ;p

May 17, 2017 | 11:00 AM - Posted by arbiter

Yea i know ryan said AMD didn't have a say in results but seems like they had a say in the hardware used which is very unfortunate.

May 17, 2017 | 03:44 PM - Posted by funandjam

Hence, why it was CLEARLY labeled at the very top of the article that it is AMD sponsored.

Take a few extra moments to reread the article. I think Ryan did his best to dance around the topic of "no 1060 results included", knowing full well that people like you are going to come out slinging conspiracy theories.
Want an example? the last sentence is a very good one:

"For the vast majority of consumers that purchase hardware in the $250 and below range, AMD’s new Radeon RX 580 family is a great choice."

By saying this, he has alluded to the idea that there are other video cards out there that are great too.

I think you and the others are actually have your panties in a bunch for the wrong reasons. If you want to be upset, then be upset that PCPer agreed to do sponsored content in the first place. Sponsored content is always going to be skewed one way or another because the company paying for the content has a vested interest to have the content look a certain way. Again, Ryan did a great job alluding to the idea that other sub $250 cards exist without specifically mentioning the gtx 1060.

Also consider, PCPer needs to get paid. Take a look around the comment section these days, the anon posts/ bot posts are gone and that actually helped drive return traffic(although I have no idea how much $ it made, but still)

Anyway, Ryan I think you did a good job, keep up the good work!

May 18, 2017 | 03:37 PM - Posted by Pinscher

As already mentioned, the hardware used in this article is indicative of the title of the article.

the purpose being to prove that the 580 is a valid upgrade path for those running older video cards.

May 17, 2017 | 11:04 AM - Posted by rock1m1

If the whole point of the article is meant to show why you should upgrade, however it does not show what you should upgrade to when you have a much better option for running this game on a slightly cheaper pricepoint - 1060.

May 17, 2017 | 11:09 AM - Posted by arbiter

AMD sponsored review, I looked at some history and if you looked you can see why gtx1060 wasn't included. Just by me saying that you can take a wild guess.

May 17, 2017 | 11:38 AM - Posted by Mr.Book

As an owner of an HD7950 running 25x14 resolution, I found this article very informative. I would be gaining double the frame rate of my existing GPU.

With VEGA to drop around the corner, it's difficult to justify the upgrade to the RX580 now. However, this may change once the pricing model for the VEGA is released.

Thanks for providing the RX580 for consideration.

May 17, 2017 | 12:41 PM - Posted by Jtaylor1986

Ryan I don't pretend to know the vision or business model of the site but with this arrangement you have effectively become a 3rd party marketing agency for AMD. I know the site evolved from an AMD fan site and I am personally ok with it going in that direction but if not this has probably damaged your credibility as neutral independent source of product reviews.

May 17, 2017 | 02:59 PM - Posted by rock1m1


May 18, 2017 | 03:39 PM - Posted by Pinscher

I don't agree. The results are factual and what's wrong with saying that the 580 is valid upgrade path from older GPUs?

May 17, 2017 | 03:27 PM - Posted by DOOFOO24

here steam videocard survey
i don't see r9 380x most people use GTX970...
just to spoil AMD VEGA is slower than GTX 1080 TI even in 4K sniper elite 4 DX12 by 10%

May 17, 2017 | 03:47 PM - Posted by funandjam

spoil what? you mean spoil rumors with even more rumors? GTFO here!

May 17, 2017 | 03:57 PM - Posted by DOOFOO24

well not rumors it base on amd demo yesterday you can see the result for yourself...
btw funandjam i well GTFO here! to stop by your MOM :)

May 17, 2017 | 05:19 PM - Posted by mAxius

can you include the 290/390 and the x variants

May 18, 2017 | 03:40 PM - Posted by Pinscher

Why don't you compare them yourself?

May 17, 2017 | 05:25 PM - Posted by Isaac Johnson

Rational person sees - "Testing commissioned by AMD. This means that AMD paid us for our time, but had no say in the results or presentation of them." - and then reads article in that context.

Irrational delicate snowflake sees article showing AMD in positive light and jumps straight to comments accusing author/website of being shills.

Some people's kids. Yeah, I know this is nothing new but that doesn't mean it's still not obnoxious and annoying.

Ryan (and crew), thanks for writing stuff on this website that I am currently viewing for completely free.

May 17, 2017 | 06:12 PM - Posted by Jtaylor1986

There is nothing unethical about what they are doing but they are essentially doing paid contract marketing for AMD now. AMD paid them to basically do a contrived review/product demo. I have no problem with this if PCPER is basically going down the path of becoming an AMD fansite but let's call it what it is a paid product showcase masquerading as a review. If this is the kind of thing that they need to do to keep PCPER financially viable I can live with it but you can't expect people to not say 'bias' every time they do an Nvidia review now given they are taking money from AMD to promote their products.

May 17, 2017 | 06:23 PM - Posted by Isaac Johnson

Wait, so not long ago they were accused of being paid by Intel to give Zen unfairly negative reviews. So which is it, are they working for or against AMD?

And what do you mean 'now'? They're already accused of being biased pretty much anytime they do a Intel/AMD/Nvidia review.

I think it's pretty cool how they are apparently taking bribes from all these companies at the same time to trash them all at the same time. It's such a perfect scheme it might actually work.

May 17, 2017 | 06:57 PM - Posted by Tim Verry

Indeed just a few weeks ago dang near every comment was about how we are NVIDIA shills :-(. I've come to terms with the fact that we will never win and that's okay. Being acused of being biased for/against every side might just mean we are doing something right after all ;-). It is clearly marked as sponsored content and people can repeat the tests themselves if they want to verify them.


I have been here 6+ years now and not once have I been told to write something a certain way to make something look good or bad when it isn't. If I had I would not still be here.

Look, if Ryan didnt take the sponsored content (which is going to help keep the site and its indepemdent content running) some other site would have and they might not have been as forthcoming that it was sponsored content. At the end of the day you can choose to trust PCPer or not, that is for you to decide. All I can offer is that we will continue to do our best to earn and retain the trust of our readers without whom we wouldn't be here!

May 17, 2017 | 07:50 PM - Posted by Isaac Johnson

Like the late great poet George Washington famously said, 'haters gon hate'.

May 17, 2017 | 07:53 PM - Posted by Jeremy Hellstrom

Dammit Tim, don't mention NVIDIA or Matrox will cut us off!

May 17, 2017 | 10:29 PM - Posted by Jtaylor1986

Did you even read the first sentence of the article, where they (to their credit) disclosed AMD paid them to write this article? This is not some tin hat conspiracy...

May 18, 2017 | 03:45 PM - Posted by Pinscher

All they did was a review on the upgrade path from an old card to a 580.

There is no biases involved as only two conclusions could be derived from the review:
1) yes it's a good upgrade bath
2) no it's bad upgrade path.

May 17, 2017 | 05:32 PM - Posted by CB

Great article.

I am currently considering upgrading from a 280x so this was right on point. I am also looking to get a new monitor.

I am considering both the 580 and 1060, but am waiting for info on Vega.

I am leaning 580 over 1060 because of FreeSync. The cards are comparable price and performance wise, but G Sync monitors are often $150+ more than an equivalent FreeSync. I may end up being able to get the cheaper Vega and a FreeSync monitor for what it would cost to get a 1060 and a G Sync monitor.

May 17, 2017 | 08:02 PM - Posted by pdjblum

great card and great value

thanks for the article

May 17, 2017 | 09:42 PM - Posted by goosegrease

I can't believe some of the complaints! Of course they're going to posit RX 580 in the best light possible. It's a sponsored post and they clearly said so. They need to pay bills somehow in order to bring us all the other exciting content we love. Would you refuse to eat meals prepared by your mother just because she paid for groceries by dancing at a strip club?

May 17, 2017 | 10:34 PM - Posted by Jtaylor1986

I actually lol'd at this :)

May 18, 2017 | 06:23 AM - Posted by DaveSimonH

"Such BIOS, much wow." Personally, clearly marked sponsored content doesn't bother me.

The article was interesting for me at least. I have a R9 280, bought in '14 as part of my summer build, so seeing where the 380 (close to 280) compares to the 580 is certainly interesting. A decent upgrade, but I think I'd rather wait to see what a 680 or 780 will look like and/or if we will see any x90 variant cards.

May 18, 2017 | 03:55 PM - Posted by Pinscher

I believe only the topic (key story line) was sponsored. The content was not, as the article explains.

May 18, 2017 | 06:59 PM - Posted by StephanS

So many comment for a sponsored article that only matter to people that own a GTX 960 or R9 380 and where also thinking of being a Rx 580, only to play Prey.... and playing only at 1440p

Do ANYONE actually fall into that category ?

Also, this leave question like... the 580,on paper, not tested, uses 65w more then the GTX 960 ? So do you also need to upgrade your PSU to go RX 580 ?

Also what CPU was used ? could it be possible that the GTX 960 owner have i5-2500k, and moving to a RX 580 . they are actually CPU limited at 1080p ?


May 23, 2017 | 05:58 PM - Posted by Southrncomfortjm

Look, this is really simple - it wasn't meant to be a fair fight, but that's okay.

All Ryan did, at the request (and payment) of AMD, is compare their current top card to cards from 2 years ago that cost the same amount. All that is meant to show is that we aren't talking about a few FPS difference, rather an upgrade would net the user over 50% more frames.

There's no point in comparing the RX 580 to the GTX 970 in this scenario since a person who owned a 970 2 years ago likely spent over $350 on that card so the comparison isn't right. This was about comparing todays $250 card to the $250 cards from 2 years ago.

Why did AMD approach PCPer for this? Because PCPer is an *independent* and *trustworthy* source for information like this. They wanted the results to be legit, so they came here. That's all there is to it.

Seriously, I doubt I've changed anyone's mind, but everyone who is going crazy over this, just chill out. This is a small, independent site, so they need to raise money to keep things going. This is one way. Want to keep them from ever having to do this again? Toss em a few dollars through Patreon. Otherwise, criticize the content without throwing around unsubstantiated conspiracy theories.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

This question is for testing whether you are a human visitor and to prevent automated spam submissions.