Review AMD Radeon RX 5600 XT Review: Look out, RTX 2060

King_V

Illustrious
Ambassador
This was actually better than I expected for the 5600XT, and, as the article stated, was exactly what AMD needed to do. Equaling or outdoing the 2060 while consuming less power was, I think, something Nvidia was not expecting. Nor was I, for that matter.

It's definitely still the target card for my son's upcoming GPU upgrade. More definitively than I would've said prior to testing.
 
This shows how efficient Navi is and with a bigger die like Big Navi, performance should surpass the 2080 Ti. In the other hand 7nm Ampere is around the corner so AMD won't have the victory for long if they release Big Navi now.
 
One thing that's probably worth noting is that all 5600 XTs currently available for sale will perform at least 10% slower out of the box than the results shown here, as they all include the old BIOS, since AMD didn't make that last-minute change until after they shipped to retailers.

So, unless one performs a BIOS update on the card, they are going to be missing out on a lot of performance, at least until new stock starts shipping with the new BIOS. And I suspect most people are not going to be flashing the BIOS of their card, as that's not normally a thing you do with a new graphics card, and is not without risk. Is the card even going to be covered if someone bricks it during the flashing process? Also, as I understand it, the new BIOS only applies to OC models, so models shipping at stock clocks may not receive a BIOS update to the higher clocks, effectively making two versions of the 5600 XT with a 10%+ performance difference between them. It won't be clear which is which just from the product names, and even the OC models may potentially come with the original, substantially slower BIOS, as that's what they all currently use. That certainly makes for a messy launch, and should have probably been mentioned in the review.

This review really should have included performance data for the original BIOS (that comes on the card), which is something the other 5600 XT reviews I've seen so far have done. It seems a bit bizarre that it wouldn't be included in the review, as this is not the kind of performance people will be getting from one of these cards right now, at least without flashing the card's BIOS.
 

King_V

Illustrious
Ambassador
One thing that's probably worth noting is that all 5600 XTs currently available for sale will perform at least 10% slower out of the box than the results shown here, as they all include the old BIOS, since AMD didn't make that last-minute change until after they shipped to retailers.

So, unless one performs a BIOS update on the card, they are going to be missing out on a lot of performance, at least until new stock starts shipping with the new BIOS. And I suspect most people are not going to be flashing the BIOS of their card, as that's not normally a thing you do with a new graphics card, and is not without risk. Is the card even going to be covered if someone bricks it during the flashing process? Also, as I understand it, the new BIOS only applies to OC models, so models shipping at stock clocks may not receive a BIOS update to the higher clocks, effectively making two versions of the 5600 XT with a 10%+ performance difference between them. It won't be clear which is which just from the product names, and even the OC models may potentially come with the original, substantially slower BIOS, as that's what they all currently use. That certainly makes for a messy launch, and should have probably been mentioned in the review.

This review really should have included performance data for the original BIOS (that comes on the card), which is something the other 5600 XT reviews I've seen so far have done. It seems a bit bizarre that it wouldn't be included in the review, as this is not the kind of performance people will be getting from one of these cards right now, at least without flashing the card's BIOS.

An excellent point - and what I'm seeing from the Anandtech review (not done reading) makes me wonder if the ideal for this card is the original reference clocks, but with the memory boosted to 14GB.

EDIT: I do admit I dislike the messiness. It seems somewhat akin to the RX 560 - 896 vs RX 560 - 1024 issue. Maybe THIS should've been the 5600 vs 5600XT, and if the OEM version has lesser performance than even the 5600 (non-XT), maybe call that 5600 LT?

EDIT 2: a brief look at NewEgg seems to indicate a lot of OC variants having slightly later release dates (Jan 23, Jan 27) with slightly higher prices (289 to 309, vs the 279 for the "base" versions). At least, that's what it seems like.
 
Last edited:
  • Like
Reactions: TJ Hooker
I see this article is the victim of cut and paste.... the how we tested header refers to the wrong card. How many other mistakes creapt in? Hopefully no more than in the headers used for formatting. Beyond that, THIS is a card to watch, or grab after a BIOS update, for right now.
 
An excellent point - and what I'm seeing from the Anandtech review (not done reading) makes me wonder if the ideal for this card is the original reference clocks, but with the memory boosted to 14GB.

EDIT: I do admit I dislike the messiness. It seems somewhat akin to the RX 560 - 896 vs RX 560 - 1024 issue. Maybe THIS should've been the 5600 vs 5600XT, and if the OEM version has lesser performance than even the 5600 (non-XT), maybe call that 5600 LT?

EDIT 2: a brief look at NewEgg seems to indicate a lot of OC variants having slightly later release dates (Jan 23, Jan 27) with slightly higher prices (289 to 309, vs the 279 for the "base" versions). At least, that's what it seems like.
I don't think it's like the Rx 560 issue. That was a fiasco with intentionally misleading advertising similar to GT 1030 crap that Nvidia did. This is just free performance if you do an update. It's a messy and unprofessional launch, but nowhere near the intentionally misleading products of rx560 and gt1030 variants.
 

TJ Hooker

Titan
Ambassador
I don't think it's like the Rx 560 issue. That was a fiasco with intentionally misleading advertising similar to GT 1030 crap that Nvidia did. This is just free performance if you do an update. It's a messy and unprofessional launch, but nowhere near the intentionally misleading products of rx560 and gt1030 variants.
The thing is we don't know if/when all models will get this update:
AMD said:
The updated VBIOS has been made available to our board partners for inclusion in select OC SKUs at launch.
https://www.anandtech.com/show/15422/the-amd-radeon-rx-5600-xt-review/2

The Pulse (which has the improved VBIOS) is what was sampled to all the tech sites for review, so that's what people are going to see in reviews/benchmarks, at least for now. But if they go out and buy an RX 5600 XT they could be getting a card that's slower than what they saw in those reviews by a not-insignificant amount. Seems like a pretty good chance people could be misled, whether intentionally or not.
 
The thing is we don't know if/when all models will get this update:

https://www.anandtech.com/show/15422/the-amd-radeon-rx-5600-xt-review/2

The Pulse (which has the improved VBIOS) is what was sampled to all the tech sites for review, so that's what people are going to see in reviews/benchmarks, at least for now. But if they go out and buy an RX 5600 XT they could be getting a card that's slower than what they saw in those reviews by a not-insignificant amount. Seems like a pretty good chance people could be misled, whether intentionally or not.
I guess we'll see. Like I said though, as of now, the situation is unprofessional, not misleading. It remains to be seem whether or not every model of 5600xt can receive this update. If every model can be updated, then there is no misleading at all, simply an unprofessional launch. If not, then we can say it is misleading and then debate if it was intentional or not.
 
  • Like
Reactions: TJ Hooker
TH published an article about the last minute vBIOS update for TDP increase just a couple days ago, but this writer clearly didn't get the memo.

Can't help but be disappointed that price creep (or lack thereof) has rendered "mid-tier" graphics power at $250-$300 where it's generally been $200-$250 in past generations.
 
Last edited:

jwcrellin

Reputable
I currently have a gtx 1070 that I purchased the month it launched, so like, 3'ish years ago. I've been thinking it's about time to replace but in my mind I thought that the 1660 and 1660 ti would be a step down. I'm fairly sure that a rtx 2060 would be a mild upgrade.

For snoots and gaggles I googled 5700 vs 1070. Turns out i'm below average now. AHHHHHH!!!

Any predictions so far as new ~$550 Nvidia cards Performance getting launched this summer?
 

Ncogneto

Distinguished
Dec 31, 2007
2,355
53
19,870
The only thing that is of a concern is the how the board partners implemented cooling on the 5600 xt. Common sense would dictate that it would be cheaper for them to just use the same cooling as they did on the 5700 as opposed to re-tooling for a different PCB and heat sink design. As such, hard to imagine very many boards not to fully utilize the updated vbios.
 

InvalidError

Titan
Moderator
In the other hand 7nm Ampere is around the corner so AMD won't have the victory for long if they release Big Navi now.
Nvidia could just cut the GTX1660S/GTX1660Ti/RTX2060 prices some more and call it a day since they're already so close in terms of perf/W.

AMD still got a long way to go if it needs 7nm to match Nvidia's performance and power efficiency on 12/14nm.
 

Olle P

Distinguished
Apr 7, 2010
720
61
19,090
... the only con is "Competing RTX 2060 has been on the market for a year".
My reflection also.
But I'm not sure how that's even a con...
The RTX 2060 has no more teething problems. That's its obvious advantage. There may be problems with the RX 5600XT that show up after a few weeks of use, just like there were with the early RTX 2080Ti.
But:
  • The RTX 2060's drivers are mature and you're most probably not going to get much more performance out of it.
  • Cards from AMD are usually getting more performance over time as drivers mature. Repeating the comparison two years from now is likely going to show more of a favor for RX 5600XT.
 
Last edited:

bramahon

Distinguished
May 4, 2010
33
1
18,540
Good review, excellent GPU - been waiting for this avatar of Navi! This could be the much hold back upgrade for my RX 470 but as Gandalf would've said - "The prices have to hold!" Also I miss Chris - he spent eons reviewing Polaris rebrand(s), lamenting how he could use something new and exciting which hardly Vega was.
 
  • Cards from AMD are usually getting more performance over time as drivers mature. Repeating the comparison two years from now is likely going to show more of a favor for RTX 5600XT.
We can see this in the Anandtech review as it has the RX580 and GTX 1060 6GB in the benchmarks. When the RX 580 was released, benchmarks had it generally about 7% slower than the GTX 1060 GB. Now we see the 580 about 7% faster than the 1060 6GB. That means the 580 performance grew by about 14% over the last 2 years, most likely due to driver updates. If the 5600XT does the same thing, we will see the OC versions outpace the 2060 by 15% in 2 years.
 
Missing the "normal" clocks for comparison seems like a glaring omission, but we're all smart here and can see through the smoke and mirrors.

The card it not bad, as it is just a RX5700LE instead of a RX5600XT, if that rings a bell to old dogs. I wonder how this will eat the RX5700's market as it's too damn close to it, even at 1440p! It's not "gimped" enough to make sense in the price bracket, but I guess AMD didn't want to go cheaper than 2060 territory? It's a weird placement, specially with the "have a new BIOS!" treatment that seems very "last moment decision".

All in all, I think AMD couldn't make Navi 14 close the gap to Navi 10, so had to come up with this half-baked solution. Still, the baking turned out fine and it seems like it'll perform well in the segment. I'm a bit salty about the 192bit downgrade though. I wonder if any of the OEMs will middle finger AMD with a 12GB 5600XT? That would be interesting :p

Cheers!
 
  • Like
Reactions: King_V

InvalidError

Titan
Moderator
I wonder how this will eat the RX5700's market as it's too damn close to it, even at 1440p! It's not "gimped" enough to make sense in the price bracket
192bits and 6GB is going to bite at some point down the road as games use more assets (need more VRAM to be comfortable) and require more VRAM bandwidth.

As far as I am concerned, the RX5600 and RTX2060 are still far more expensive than they should be, and so is everything above them.
 
  • Like
Reactions: jeremyj_83

TJ Hooker

Titan
Ambassador
I guess we'll see. Like I said though, as of now, the situation is unprofessional, not misleading. It remains to be seem whether or not every model of 5600xt can receive this update. If every model can be updated, then there is no misleading at all, simply an unprofessional launch. If not, then we can say it is misleading and then debate if it was intentional or not.
Looking hopeful that all cards will at least have the VBIOS available for download, even if they don't ship with it. Looked at a couple 'budget' 5600 XT models that are already available (ASRock challenger, PC Red Dragon, SVS Thicc II), all of them have new VBIOS versions available for download on their manufacturer's sites.

I do still question whether the average Joe is going to be aware of this sort of thing and actually flash a VBIOS, and therefore might lose out on a bit of performance if some cards ship with the lower powered BIOS. But if anyone who cares enough to look into it can at least update if they want that's not bad IMO.
 
  • Like
Reactions: TCA_ChinChin
Looking hopeful that all cards will at least have the VBIOS available for download, even if they don't ship with it. Looked at a couple 'budget' 5600 XT models that are already available (ASRock challenger, PC Red Dragon, SVS Thicc II), all of them have new VBIOS versions available for download on their manufacturer's sites.

I do still question whether the average Joe is going to be aware of this sort of thing and actually flash a VBIOS, and therefore might lose out on a bit of performance if some cards ship with the lower powered BIOS. But if anyone who cares enough to look into it can at least update if they want that's not bad IMO.
Different question is could these clocks be achieved just via over clocking it normally if you can't/don't want to flash the BIOS. Obviously AMD put 14Gbps RAM on the card so the RAM OC would be easy, the more difficult one would be core clock.
 
  • Like
Reactions: TJ Hooker
Looking hopeful that all cards will at least have the VBIOS available for download, even if they don't ship with it. Looked at a couple 'budget' 5600 XT models that are already available (ASRock challenger, PC Red Dragon, SVS Thicc II), all of them have new VBIOS versions available for download on their manufacturer's sites.

I do still question whether the average Joe is going to be aware of this sort of thing and actually flash a VBIOS, and therefore might lose out on a bit of performance if some cards ship with the lower powered BIOS. But if anyone who cares enough to look into it can at least update if they want that's not bad IMO.
I certainly hope AMD and the board partners both update all the cards they currently have, and then push to educate the consumer asap about the situation. It's in everyone's best interest that these bios updates are easy to access and pushed in a timely manner.
 
  • Like
Reactions: TJ Hooker
Cards from AMD are usually getting more performance over time as drivers mature. Repeating the comparison two years from now is likely going to show more of a favor for RTX 5600XT.
On the other hand, it's very possible that raytraced lighting effects may become the norm for "ultra" settings in the coming years, in which case, the 5600 XT may fall well behind with those enabled. The 2060 might not have great RT support (and that goes for the 20-series in general), but I fully suspect it can handle such effects a lot better than the 5600 XT. If that were not the case, AMD would have likely released drivers to enable DXR support on their cards by now, but they are likely holding off on that until their next generation of hardware is available that can actually offer useable levels of performance with those effects enabled.

We can see this in the Anandtech review as it has the RX580 and GTX 1060 6GB in the benchmarks. When the RX 580 was released, benchmarks had it generally about 7% slower than the GTX 1060 GB. Now we see the 580 about 7% faster than the 1060 6GB. That means the 580 performance grew by about 14% over the last 2 years, most likely due to driver updates. If the 5600XT does the same thing, we will see the OC versions outpace the 2060 by 15% in 2 years.
The RX 580 didn't see anywhere near that level of performance uplift (also it's nearly 3 years between those two articles). What you're seeing there is the result of changing the tests themselves. Looking at the games tested in those two reviews, only one of them remained the same, GTA V. And with only nine games being tested, the chances of a few outliers favoring one architecture over another and throwing off the results is very possible. And that almost certainly happened there, since just about any RX 580 review from the time of its release showed it performing right about on par with the GTX 1060 or slightly faster, so there was no 7% deficit to begin with. Their game benchmarks at the time likely simply favored the Pascal architecture more. Similarly, other benchmarks revisiting these cards haven't exactly shown anywhere near that level of change in relative performance between the RX 580 and the 1060, so I would suspect that the limited selection of games they are testing with now simply no longer favor Pascal, and now happen to be more favorable to the RX 500 series. That's why I prefer reviews that benchmark 20+ games, as it reduces the likelyhood that a few outliers will throw off the results to a significant degree.

We can look at the one game that remained the same though, GTA V. That game generally favors Nvidia hardware, but it might be indicative of any relative performance changes between the cards over time. In their RX 580 review, the 1060 FE was 25.1% faster at 1080p, and 18.7% faster at 1440p, compared to the reference-clocked RX 580. In the 5600 XT review? The 1060 is now 32.6% faster at 1080p and 28.6% faster at 1440p. If anything, that would be indicative of the RX 580 falling further behind in that game. Of course, it's also likely that their testing methodology of the game may have changed in those 3 years. In any case, it hardly supports the suggestion that the RX 580 has substantially improved in performance due to driver updates.

Yeah, it depends on how far you're able to increase power limit and voltage with the 'stock' BIOS, and how much these limits were changed for the new BIOS.
If I had to guess, they probably have that locked down on the original BIOS, since they likely didn't want the card to be able to perform so close to a 5700. It's likely a similar situation to the 5700 compared to the 5700 XT, where overclocking is limited, but by flashing the card with an XT BIOS one can unlock additional performance and overclocking headroom that is otherwise not accessible. See here, for example...