The initial performance of AMD's new Radeon RX Vega 56 may have looked a little lackluster at launch if you were expecting it to exceed the strides made by Nvidia's year-old GTX 10 series. At best, Vega 56 was a smidgen faster than the GTX 1070 for about the same amount of money, while consuming more power, so despite looking impressive in absolute terms, we felt it was going to be a tough sell.

We had the opportunity to give Vega 56 a good workout at launch, but we didn't receive a Vega 64 card on time for a review. In fact, our sample only arrived Monday, at which point we dropped everything and have been testing since.

With that said, this might be the biggest GPU benchmarking session in TechSpot's history, it's so large that we almost gave up after accepting the challenge. After about a week's worth of testing, we have an incredible amount of data to pour over. Our original plan was to add five games to the 25 titles we originally tested. In the end, after so many requests we felt compelled to add seven more games for a total of 32 titles benchmarked in this article.

Additionally, we've added results from the GeForce GTX 1070 and GTX 1080 Founders Edition graphics cards, which were compared alongside the custom board partner cards at resolutions of 1080p, 1440p and 4K.

We'll be using our Core i7-7700K test rig clocked at 4.9GHz but still plan to do extensive testing with Vega and the GeForce 10 series on Ryzen as well. Enough messing about, we have work to do...

RX Vega GPU Test Specs

Benchmarks

Getting things going, we have Ashes of the Singularity aka 'Ashes of the Benchmark' and this is a title I expected AMD's new Vega series to perform quite well in. Nvidia has put a heap of time into developing their drivers to make up for the GeForce 10 series' weak DX12 performance and we see it paying off here. The GTX 1070 roughly matches Vega 56 while the GTX 1080 matched best version of Vega 64 -- a mild result overall for AMD.

Assassin's Creed Syndicate has been brought back for this article after receiving quite a few requests for it, including from Tim -- not sure if he's a fan of the game or what. Unfortunately, this is a poor title for AMD and even the air-cooled Vega 64 card falls behind the GTX 1070. Not a great effort in our second title tested but AMD has another 30 to titles to recover.

Well that didn't take long, right into Battlefield 1 and already we're seeing much more competitive performance from Vega. Something worth noting though is how poorly the air-cooled Vega 64 card does in this title, only just managing to edge out Vega 56. The liquid cooled version is much more impressive with an average of 122fps, which is a good bit faster than even a custom GTX 1080.

Benchmarks: Civilization VI, Deus Ex: MD, DiRT 4

Civilization IV is another heavily requested title that I had ditched for my day one Vega coverage and you can see why from the results above: it's a bad game to test GPU performance. It's a great game to test CPU performance but as you can see here even with a heavily overclocked Core i7-7700K we are mostly finding the limits of the CPU. Again though, with 31 other games in this GPU comparison, adding Civilzation VI shouldn't skew the results.

Deus Ex: Mankind Divided is another DirectX 12 title but this one is far more GPU-bound and as luck would have it Vega does well here. This is of course an AMD sponsored title but even so it does show how well Vega can perform when a game has been optimized for it.

The Vega 56 came knocking on the GTX 1080's door here while the air-cooled Vega 64 model pulled ahead of the MSI GTX 1080 Gaming X 8G -- the liquid-cooled version advanced well ahead.

Dirt 4 left me confused after my initial Vega 56 review but I have since been able to confirm these results. Vega does well using 2xMSAA and even 4xMSAA but tanks with 8xMSAA, though when I say tanks, I mean it drops down to the same level of the competing Nvidia cards. Using CMAA the Vega cards absolutely trash the GeForce 10 series in what is a completely baffling result. I've asked AMD about the various performance anomalies in this title with Vega but I've yet to hear anything back.

Benchmarks: Dishonored 2, DOOM, F1 2016

Moving on from the rather spectacular Dirt 4 numbers we find some underwhelming performance in Dishonored 2. This title has never been kind to AMD so I'm sure no one was expecting anything overly impressive here. Still, seeing the air-cooled Vega 64 only match the GTX 1070 Founders Edition was disappointing to say the least. The liquid-cooled version trailed the GTX 1080 FE card as well.

Doom is a title where we always anticipate AMD to do well -- just check out the RX 580 beating the GTX 1060 by an 18% margin. In Vega's case, it's actually a lot less impressive. The Vega 56 is only 10% faster than the GTX 1070 FE while Vega 64 air-cooled is only able to match the MSI Gaming X 1080. So while strong performance overall, it's still not quite as impressive as you might expect in this title.

Despite Vega 56 racing ahead of the GTX 1070 in F1 2016, the air-cooled Vega 64 model was most unimpressive, only tacking on a few extra frames. The liquid-cooled version was much more remarkable and did manage to pull ahead of MSI's factory-overclocked GTX 1080. Overall solid performance from Vega here with the exception of the air-cooled 64 model.

Benchmarks: Far Cry Primal, For Honor, Ghost Recon Wildlands

Interestingly when testing with Far Cry Primal we see almost no difference between the air-cooled Vega 64 and Vega 56. For that reason the liquid-cooled version was 16% faster but even so only managed to match the MSI GTX 1080 Gaming X. Given how well the RX 580 does compared to the GTX 1060 in this title you have to think some driver works needs to be done here.

Performance in For Honor is okay, Vega 56 roughly matches the GTX 1070. Vega 64 is less impressive as the air-cooled model trails the GTX 1080 FE while it takes the more expensive liquid cooled version to match the MSI GTX 1080. Vega is more appealing if you focus on the 1% lows for this title, but either way the liquid cooled edition is much closer to the GTX 1080 than it is the 1080 Ti.

Vega 56 manages to match the GTX 1070 FE card in Ghost Recon Wildlands though again we see that Vega 64 really isn't much faster. The liquid cooled version did average 71fps though this only put it on par with the GTX 1080 FE and a few frames behind the MSI GTX 1080 Gaming X.

Benchmarks: GTA V, Hellblade: Senua's Sacrifice, Hitman

I once said Ryzen 5 1600 was a better value than the 7700K for gaming and last year I picked the GTX 1060 over the RX 480, but neither of those things stirred up the comment section nearly as much as dropping Grand Theft Auto V from the games we test with. Maybe that's a slight exaggeration but anyway you guys certainly weren't happy, so for this comparison at least GTA V is back.

Vega's less than impressed with that decision though, as even the liquid-cooled Vega 64 loses out to the GTX 1070 Founders Edition -- ouch! We're seeing GTX 1060-like performance in this one. Hopefully a driver update can improve that though I’m not sure how keen AMD will be to go back and optimize for such an old title, even if it's popular.

Hellblade's back for another round of benchmarks. Previously Vega 56 trailed the GTX 1070 FE in this title and of course that hasn't changed yet. Once again we find that the air-cooled version of Vega 64 just isn't able to pull away from Vega 56. The liquid cooled version does much better though it could only match the GTX 1080 Founders Edition.

Vega 56 matched the GTX 1070 when testing with Hitman but it again took the liquid cooled version of Vega 64 to match the GTX 1080 Founders Edition card. This is a disappointing result for a DirectX 12 title and an AMD sponsored one at that.

Benchmarks: ME Andromeda, Mirror's Edge Catalyst, Overwatch

Mass Effect Andromeda shows us what we've seen a number of times already, Vega 56 is roughly on par with the GTX 1070 while Vega 64 needs to be liquid cooled to catch, or at least come close to catching the GTX 1080. Another average result here for AMD.

Moving on we have Mirror's Edge Catalyst and for this one Vega looks quite competitive. The air-cooled Vega 64 model is still quite underwhelming but I was impressed with what the liquid cooled version pulled out this time. Vega 56 also does a good job against the GTX 1070.

Overwatch generally favors the green team and we see this here when comparing Vega to the GeForce 10 series. Vega 64 struggles to match the GTX 1070 in this title and even the liquid cooled version drops to the same 1% lows. I mean we are seeing well over 100fps at all times, but Vega 64 pales in comparison to the GTX 1080.

Benchmarks: Battlegrounds, Prey, Quake Champions

Vega sucks big time in PlayerUnknown's Battlegrounds right now and while many will no doubt argue this game shouldn't be included as it's still early access, I obviously disagree. In fact I believe most of you do given the overwhelming amount of requests I received to test Vega's performance in this title. Battlegrounds might only be in the early access phase but it's massively popular right now and it's a game AMD surely needs to optimize for ASAP! Hold off on releasing another optimized mining driver and sort out Battlegrounds please.

Looking at the average frame rate performance in Prey we see that Vega is competitive with competing GeForce 10 series graphics cards. That said, the 1% low results were always noticeably lower so I'm not sure if this is yet again another suspected driver issue or something else. I should note that the air-cooled Vega 64 model was pretty underwhelming while the liquid cooled version is decent.

AMD's recently released Crimson ReLive 17.8.1 driver improved performance in Quake Champions and this is the driver I used for all the Vega testing. Please note performance for the Fury X and RX 580 went unchanged. I saw between a 6-8% performance boost for the Vega GPUs in this title. That was enough to push Vega 56 ahead of the GTX 1080 in this title while the liquid cooled Vega 64 card wasn't that much slower than the GTX 1080 Ti. Impressive stuff in this game, even if it is a sponsored title.

Benchmarks: Rainbow Six Siege, Resident Evil 7, ROTR

Vega comes out to play in Rainbow Six Siege with the Vega 56 being 12% faster than the MSI GTX 1070 Gaming X. We even see a rare quality showing from Vega 64 as the air-cooled model eeks out a win over the MSI GTX 1080 Gaming X. The liquid cooled model pushes further ahead but still lags behind the GTX 1080 Ti by a 15% margin.

Here we see again that in select titles Vega can really turn up the heat on the GeForce 10 series. This time when testing with Resident Evil 7: Biohazard we see Vega 56 beating the GTX 1070 by a decent margin, though the air-cooled Vega 64 card does struggle against the GTX 1080, at least the MSI Gaming X version.

Vega 56 makes out well in Rise of the Tomb Raider though the air-cooled Vega 64 model is less impressive. Once liquid cooled the results aren't bad and we see the same 1% low performance to that of the MSI GTX 1080 Gaming X.

Benchmarks: Shadow of Mordor, Sniper Elite 4, The Witcher 3

Middle-earth Shadow of Mordor is an oldie but a goodie. We're testing with 32 games so why not throw this into the heap. As luck would have it, Vega's also quite competitive in this title. Vega 56 beats the GTX 1070 hands down and the liquid cooled version of Vega 64 takes out the MSI GTX 1080 Gaming X. Of course it costs a lot more but there's hope for the custom board partner models yet.

Sniper Elite 4 isn't a game I've tested with before but quite a few of you requested it and as a modern DX12 title I thought why not. This is an AMD-sponsored title so not surprising to see that Vega does well here. Even the air-cooled Vega 64 model has pulled up its socks and is looking respectable for once, though the Vega 56 is once again the star of the show and beats out a factory-overclocked GTX 1070 without much trouble.

The Witcher 3: Wild Hunt is another title that's getting on now and this one gives us some mixed results. Vega 56 looks good as it beats out even the MSI GTX 1070 Gaming X. However air-cooled Vega 64 falls short of the GTX 1080 and it takes the liquid cooled model before Vega is seeing 1080-like performance.

Benchmarks: Titanfall 2, The Division, Total War: Warhammer

The Titanfall 2 results are very competitive as we see Vega 56 match the GTX 1070. The air-cooled Vega 64 model can’t quite catch the GTX 1080 but the liquid cooled version gets the job done, albeit at a questionable price premium.

Tom Clancy's The Division was tested using DirectX 12 and here Vega 56 trounces the MSI GTX 1070 Gaming X in a true display of pixel dominance. Even Vega 64 does well, yeah the flipping air-cooled version actually beats the GTX 1080 by a handy margin and this means the liquid cooled model is hot on the heels of the GTX 1080 Ti.

Testing Total War: Warhammer usually plays into AMD's hands and we certainly see that here. Vega 56 is nipping at the 1080's heels while the air-cooled Vega 64 outpaces the GTX 1080. The liquid cooled card meanwhile was on the hunt for Nvidia's GTX 1080 Ti though it's not quite quick enough off the mark and ultimately gets left behind.

Benchmarks: Dawn of War III, Watch_Dogs2

When testing Vega 56 I was pretty shocked by how strong this GPU was in Warhammer 40,000: Dawn of War III. Since then other media outlets have also confirmed exceptional performance for Vega in this title so I'm confident in the results. If every game looked like this AMD would be laughing. Anyway Vega 56 sticks it to the GTX 1080 while the Vega 64 models pull even further head. The liquid cooled edition gets within spitting distance of the GTX 1080 Ti, get it, it's liquid coo... ahh never mind moving on.

Watch Dogs 2 hates AMD -- it doesn't matter if it's the company's Ryzen CPUs or Vega GPUs. Here we see the GTX 1070 Founders Edition model giving not just Vega 56 a hard time but also Vega 64. Liquid cooling Vega 64 pushes it ahead, but only just... just ahead of the GTX 1070 that is.

Power Consumption

Performance and pricing aside, power consumption is by far Vega's biggest hurdle. In all honesty, performance isn't bad and it might even improve quite a bit over time. Moreover, once pricing settles down I'm sure that will be fine as well. What won't change is the power consumption. Sure you can undervolt and I'll look at that in the future but for now this is what most Vega owners will be faced with.

Despite using a bit more power than the GTX 1080, Vega 56 really isn't that bad. Vega 64 on the other hand is a serious power pig. The already hungry air-cooled version makes the GTX 1080 Ti look like it's on a diet, while the liquid cooled Vega 64 behaved like a bear at a buffet.

Seriously though, pushing total system consumption 24% higher than the GTX 1080 Ti for on average 18% less performance is an issue. Not everyone cares as much as me about power consumption but it does come at a cost.

In its current reference card form, Vega is a 'hangry' graphics card that will take its problems out on your power supply.

Putting It All Together, Big Graphs

We kind of breezed through the 32 games tested and didn't go into the margins as much as we often do throughout the benchmark pages. With so many games we didn't want to bore some readers with a mass of percentages. Instead, we're about to hit you with all those percentages now as we look at performance on a per game basis in a single graph. Be warned: we have four comparisons here with 128 individual data points so grab your reading glasses and let's get into it.

Previously we found Vega 56 to be 2% faster than the MSI GTX 1070 Gaming X on average across 25 games. We've since downgraded slightly to the GTX 1070 Founder Edition which is a few percent slower and we've also removed Crysis 3 and replaced it with eight other games. So this has changed the margin slightly and now Vega 56 is 5% faster, though only about 3% faster than the MSI model.

Anyway as you can see the big wins for AMD here is again Dirt 4 and Dawn of War III. AMD does well in many of the newly released titles, particularly those that use modern APIs so that's well worth noting. Still troubling performance in massively and we mean massively popular titles such as GTA V and Battlegrounds is an issue and AMD need to work on this. We can probably turn the other way for games such as Assassin's Creed and Dishonored though. Anyway, overall a strong showing for Vega 56 and once pricing settles down and we get custom board partner models it should be a worthy alternative to the GTX 1070.

Moving on we have a much less exciting comparison to make with the air-cooled Vega 64 model -- what a stinker. Overall it was 5% slower than the GTX 1080 Founder Edition and of course it runs much hotter, louder and burns more power than a Skylake-X processor on steroids. Not much else needs to be said here. The reference card just isn't cutting it. We do feel like the board partner versions will be significantly better, especially for this model, but that's a story for another day.

Liquid cooling Vega 64 and allowing it to have those higher clock speeds while avoiding throttling really helps this complex GPU. The liquid cooled version was 4% faster than the GTX 1080 Founders Edition overall but in 10 of the games tested it was faster by a 10% margin or greater. Of course the liquid cooled edition is priced $100 above the GTX 1080 when looking at the standalone MSRP, so it's hardly "hot" in terms of price vs. performance.

Compared to the Aorus Xtreme GTX 1080 Ti, the liquid cooled Vega 64 was 18% slower and once pricing corrects it should only be 14% cheaper, so another sketchy result for Vega.

Wrap Up

At this point we've established a pretty good baseline of Vega's performance. Most other reviewers have concluded by saying that Vega 56 is slightly faster than the GTX 1070 and Vega 64 is slightly slower than the GTX 1080 while the liquid cooled version wins more than it loses but struggles massively against the GTX 1080 Ti.

That's pretty much the situation from a frame rate standpoint and we didn't need a 32 game benchmark for that conclusion, but it was interesting to see where Vega did well, and not so well.

It was also interesting to see where the air-cooled Vega 64 offered gains over Vega 56, to see which games have very little separation between the two GPUs, and to see where the Vega 64 dominated the GTX 1080 -- there is some hope for Vega yet.

Setting power consumption, pricing and even performance aside for a moment, I do believe one of Vega's biggest issues is AMD's own reference card design. This is yet to be proven, but if history has shown us anything it's that AMD's reference cards suck and they rob their own products of performance.

The liquid cooled model might very well be a best case scenario for Vega and that's kind of a good thing. If board partners can adequately cool Vega 64 using air coolers then we might see a significant improvement in performance for that model. That would also leave room for Vega 56 to improve and I can easily see that model beating the GTX 1070 by a comfortable 10% margin while remaining cool and quiet.

Shopping shortcuts:

On the subject of pricing, there is currently a supply issue with Vega that is reminescent of what we saw with the RX 580, RX 480 and even the GTX 1060, but this was always likely to happen and even more so with cryptocurrency mining as big as it is currently. I'm confident that Vega 56 and 64 models should start hitting the standalone MSRP once we see custom board partner cards arrive later next month. Of course, if there's still a serious supply issue due to miners then that might not be the case.

For now we'd recommend to avoid the reference cards, sit tight and wait for the much better custom versions and give supply a chance to catch up with demand before paying an inflated price. Overall, I'm still not 100% sure how I feel about Vega yet. There are some promising signs that it can be competitive, but I would like to reserve ultimate judgment until we have a few board partner models in hand.

80
TechSpot
score

Pros: Vega 64 is slightly slower than the GTX 1080, while the liquid cooled version wins more than it loses against the GTX 1080.

Cons: Poor availability and thus pricing. AMD's air-cooled RX Vega 64 reference card is ~5% slower than the GTX 1080 FE despite being hotter, louder and using more power (though we have high hopes for custom cards).