AMD Radeon RX 5700 and 5700 XT review

Posted on: 07/07/2019 04:30 PM

---

I personally am waiting for something that'll play games at 4K@60Hz for a reasonable price. Unfortunately, Hilbert reviews with AA enabled (and I don't care about AA at 4K) so it's hard to know what I really should get.
EDIT:
For the record, I don't blame him for enabling AA - I'm not asking him to change his testing routine just for me.

Posted on: 07/07/2019 04:32 PM

---

Wow, impressive.

Posted on: 07/07/2019 04:36 PM

---

Given the 5700xx price-performance differential vs Nvidia, and since (once again) no one has been disadvantaged at this stage, do we think Nvidia's might pull a fly one with their super stack reconfiguration and get a price cut before release on Tuesday? That would be, interesting.

Posted on: 07/07/2019 04:40 PM

---

The performance here is actually fantastic. Way to go AMD with navi!

Posted on: 07/07/2019 04:40 PM

---

I personally am waiting for something that'll play games at 4K@60Hz for a reasonable price. Unfortunately, Hilbert reviews with AA enabled (and I don't care about AA at 4K) so it's hard to know what I really should get.
EDIT:
For the record, I don't blame him for enabling AA - I'm not asking him to change his testing routine just for me.
Consider all reviews of 4k perf as only rough guidelines, as you will likely end up using different settings in the end. AA of course is the first to go off.

Posted on: 07/07/2019 04:44 PM

---

I guess to elaborate: I want something that can readily maintain 4K@60FPS with all max detail in most modern titles, except with AA off. But again, for a reasonable price. I have the money to buy a 2080Ti but I can't justify that price point. $400 is pretty much as high as I'm willing to spend on a single component for a gaming PC, and if that means I need to wait another year, so be it. Considering I'm still on 1080p, I'm obviously not in a major rush.

Posted on: 07/07/2019 04:49 PM

---

The RX 5700 (which I'm looking at) and RX 5700 XT cost way more where I live ($444/$509), but they're both trading blows with the RTX 2060 and RTX 2070.

However I do think this GPU release is a bit underwhelming (much like Vega), but thankfully they seem to have done an excellent job with the Ryzen 3000 series.

Posted on: 07/07/2019 04:52 PM

---

Given the architecture refresh, I'd give it a little more time to mature, and maybe to see the prices drop a little. That's what I'm doing, anyway.

Thankfully, the Linux drivers already seem to be in better shape (which are overall more important to me).

Posted on: 07/07/2019 05:03 PM

---

AMD made quite nice progress especially against their own older products. If you don't think about the prices much, and of course ignore the RT, it feels like Nvidia succeeded well with their Super versions, though. But as it is, a buyer just needs to ponder if they want to pay the extra for the Super, getting the RT as well of course. Based on this review, I couldn't say 5700 (XT) would absolutely be the way to go.

Nevertheless, it does seem like GPU prices may have permanently become a bit more expensive than they used to be. Nvidia could raise the prices due to lacking competition, and AMD would also like to charge a bit more to get money for their R&D, which they certainly need. It remains to be seen if Intel entering the fray can change anything. Typically Intel hasn't been the forerunner of cheap prices, though, haha.

Posted on: 07/07/2019 05:06 PM

---

Overall a litte slower than the new RTX Super, but also cheaper!

It will depend on the sale we catch on both brands and the importance or not that each one will give to the Ray Tracing factor.

As happened with Vega 56 vs. Vega 64, the 5700 XT seems to have been stretched beyond its sweet spot, the 5700 in performance/watt is a lot better!

Overall, solid cards from AMD and more or less matching Turing efficiency, despite the 7nm vs 12nm situation.

Interestingly, now, unlike what some people did with the RTX Super, no one complains about AMD showing the middle finger to the Radeon VII owners. A $699 card with only 5 months of market and now is virtually irrelevant to a $399 card.

Posted on: 07/07/2019 05:08 PM

---

Was about to dis it, but on 2nd hand it rivals 1080ti, and that is something for lower end chip.. Now the full Navi should easily match/beat 2080ti.. Im just wondering what the heat will be, if these small ones already heat a lot. Hopefully not another 290X.. xD

Posted on: 07/07/2019 05:09 PM

---

That's because the VII is a niche product. It has some solid gains in some places. It was mostly created as a placeholder. It's an older architecture. The Super series were made strictly to spite AMD. You can't really compare the two.

It's to be expected that a new GPU will outperform its predecessors. The Super series aren't new.

Posted on: 07/07/2019 05:12 PM

---

Interestingly, now, unlike what some people did with the RTX Super, no one complains about AMD showing the middle finger to the Radeon VII owners. A $699 card with only 5 months of market and now is virtually irrelevant to a $399 card.

The very expensive 16GB HBM2 had a bit to do with that. Together with this HBM2 memory, the VII also had further features directing it's use, in addition, away from gaming. As schmidtbag said, it's a uniquely niche product.

Posted on: 07/07/2019 05:18 PM

---

Respect!

This is by far the best (also fairest) review I've seen so far.

Posted on: 07/07/2019 05:20 PM

---

Good product and price in UK here is actually ok.

It's obvious that the drivers are good, but, haven't reached their full potential yet. Also, the move to GDDR6 is probably the best thing they've done here.

Posted on: 07/07/2019 05:29 PM

---

The problem is people in youtube-land actually think/thought the 16GB of ram will future-proof this product. Only content creators will be able to justify it, meanwhile everyone else will move over to XT and Super cards.

Posted on: 07/07/2019 05:43 PM

---

@Hilbert Hagedoorn, will you not use Anno 1800 as a benchmark for new GPUs?

Posted on: 07/07/2019 05:44 PM

---

Not a big issue with heat I think for 64CU card, if CUs have same arrangement.
64 CUs means 1.6 times higher performance at same clock. And would result in TBP of 335W at those same clocks. But limiting card to 300W would mostly result in 10% performance loss only.
So some 1.4 times higher performance than RX 5700 XT which would be around 2080Ti's performance and would have some 16B transistors.

Issue would be with bigger GPUs. Because there RDNA1 would have to throw away all that clock potential just to stay within 300W limit. This 1st iteration can go up to 2GHz for good chips. Gaming clock is around 1700~1800MHz to keep TDP. But bigger GPUs would be forced to go down to maybe 1450~1500MHz. That would be big waste.
And I think AMD is working hard on optimizing power draw so next iteration can use higher portion of achievable clock.

It is actually funny, because people with 5700 who raise TBP limit (and other limits) to like 240W may get stock 5700 XT performance via OC if their chip can clock that high. (Something like 7950 vs 7970)
But I think that it is pointless to even try to OC those cards till all power limits in BIOS are changed.

@Hilbert Hagedoorn : Do those reference cards have dual BIOS switch? I have seen some unusual rectangular hole on side of PCB close to cover which makes tiny access inside. Maybe there is switch. (Wishful thinking.)

Posted on: 07/07/2019 05:52 PM

---

Issue would be with bigger GPUs. Because there RDNA1 would have to throw away all that clock potential just to stay within 300W limit. This 1st iteration can go up to 2GHz for good chips. Gaming clock is around 1700~1800MHz to keep TDP. But bigger GPUs would be forced to go down to maybe 1450~1500MHz. That would be big waste.

This seems to be something nvidia does quite often and it seems to work for them.

RTX 2080 has less cores but higher frequencies then the 2080 ti, yet the 2080 ti performs quite a bit better then the 2080, same with the RTX titan. Same thing with the 1080 vs 1080 ti or titan X/XP, etc.

I'd rather have a much larger GPU with more cores and lower frequency and overall more performance options, then not have them at all.

Posted on: 07/07/2019 06:05 PM

---

This seems to be something nvidia does quite often and it seems to work for them.

RTX 2080 has less cores but higher frequencies then the 2080 ti, yet the 2080 ti performs quite a bit better then the 2080, same with the RTX titan. Same thing with the 1080 vs 1080 ti or titan X/XP, etc.

I'd rather have a much larger GPU with more cores and lower frequency and overall more performance options, then not have them at all.
True. But since AMD is going to hit power limit, making big 20B transistor GPU for example could maybe deliver 1.65 times higher performance than 5700 XT (10B transistors) due to need to stay within 300W limit.
Issue is price then. Twice as big GPU, much worse yields. As 5700 XT goes for $400, what would be price of card with GPU twice the size? $1000? $1200?

Sadly, even if AMD went chiplet way and managed to make card with 2 chiplets with same 20 Dual-CUs as 5700XT and therefore potentially $900 price point, it would still be power limited and therefore around that 1.65x multiplier of 5700 XT.

Had 5700 XT been 200W TDP card, I would be much more hopeful. Because like 60W out of that are memories, VRMs and blower. Then GPU would be lovely 140W and clock sacrifice on bigger/dual chip would be much smaller.
Maybe AMD will go with HBM2 again to save like 40W on 12GB card. But power is big limiter even for RDNA... for now.

Posted on: 07/07/2019 06:23 PM

---

True. But since AMD is going to hit power limit

Had 5700 XT been 200W TDP card, I would be much more hopeful. Because like 60W out of that are memories, VRMs and blower. Then GPU would be lovely 140W and clock sacrifice on bigger/dual chip would be much smaller.


Maybe i'm looking at this wrong, but the 5700 XT is only a 225 watt card. Sure there's the 50th anniversary edition, which is 235 watt. But that leaves them with 65-75 watts to play around with for the 300 watt limit.

To put that in comparison to nvidia, since that was my comparison before:

RTX 2080 was a 215 watt card, that's only 10 watts away from the 5700 XT, and they managed to increase the die size, add cores, decrease the frequencies, and have a 2080 ti at 250 watts.

Now, looking at this review, it appears the RX 5700 isn't even taking 235 watts, but rather 204, very close to the 200 watt TDP you wish it'd be. Now, that's just this test, could be variables that aren't taken, i get it. A 2080 takes 230 watts while a 2080 ti takes 266 watts, both above their rated TDPs, in this test.

Now from a cost perspective, i agree...ish, i still don't think the RX 5700 cards are that expensive to produce, even on 7nm, but i could be wrong. Either way, creating a twice the size GPU, which would realistically only be around 355mm2 (just for someone that is reading things that double 251mm2 of the RTX 5700 would be 502, that's not correct, that'd be more like 4 times larger of a die), i'm not sure that'd be required, and i doubt it'd increase the price of the GPU by twice (not the die, but the whole GPU, as memory prices, PCB, etc. wouldn't change, as long as the amount of memory, among other things, didn't change). I think they could create a large GPU with more cores/CUs and charge $500-600 and bring some really good competition to nvidias high end.

I'm sure there's a reason AMD didn't do it, high end is in itself a relative niche market, but i do believe the could have, within reasonable price points, brought out a large GPU to compete.

Posted on: 07/07/2019 06:34 PM

---

When you're financially constrained, like AMD is, you don't go for the smallest market first.... You go for the market with the greatest potential. These are mid-range cards.... That is the market with the greatest potential.

Posted on: 07/07/2019 06:38 PM

---

That's pretty much my guess as to why they didn't, hence the whole it's a relatively niche market. I feel like that's what they said they were going to do with their CPU market after bulldozer and that didn't work out so well for them. I hope by the time Intel releases their GPUs that AMD is fighting the full lineup again.

Posted on: 07/07/2019 06:39 PM

---

@Aura89 : Understandable point of view. Even agreeable mostly... Except maybe expecting much bigger chip to be happy with 256bit memory interface. AMD may have improved compression of some data types and remove need to completely read those compressed data blocks, decompress, alter, recompress, store... as they can now read and write specific chunks directly. But even that may not be good enough improvement to provide sufficient bandwidth for big GPU.

Actually, I would like to see memory downclocked and have measurement of performance impact.

Posted on: 07/07/2019 06:48 PM

---

I'll be honest, the whole 256bit memory interface and increase completely slipped my mind. I'm not sure it'd have a huge impact still though in overall price.

Memory and the bit interface has always been one of those areas that i find it's difficult to justify the increase if the cost puts it too much higher, as well as find out the bit increase cost.

By that i mean, RX 580, 4GB vs 8GB, most games show little to no difference between these. Not saying you can't find the difference, just can't find a huge difference.

And since we don't generally get exactly the same configurations of GPUs with different memory bits, it's fairly difficult to say "An RTX 2080 ti would have similar performance to an RTX 2080, if it had the same 256bit bus, even with more cores", or anywhere inbetween. At least, i haven't seen this kind of test being done.

So it's hard to say if a higher then 256bit memory would be required to have a better performing GPU by a decent amount, or how much more expensive it would make it if it did require it.