AMD Radeon RX 5700 XT

[W1zzard]

TPU used to test World of Warcraft before. Haven't seen them tested for a while.is there a specific reason for it? I know that game is hard to test as it is affected by many things but it is the one I play the most...

Always online, requires internet connection, no single player, patches can happen at any time, invaliding all data. and really not a lot of people complained. You're the third one so far

 

[efikkan]

I do believe the goal is to have a reasonable sized representative selection of relevant games. I don't think you can get everyone agreeing on which games this should be, but that shouldn't really matter as long as you have a decent amount of relevant games, enough to even out any outliers. So even if my personal picks may be missing, it shouldn't matter, as it still gives me a clear idea of the relative performance between products, and that is the primary thing buyers need to know.

 

[Turmania]

Thanks for the clarification. I understand the reasons for not reviewing it anymore, but those reasons stood the same when you were reviewing before, but please do not take it as I'm complaining or etc...I just wanted to end my curiousity.thanks.

 

[W1zzard]

Thanks for the clarification. I understand the reasons for not reviewing it anymore, but those reasons stood the same when you were reviewing before, but please do not take it as I'm complaining or etc...I just wanted to end my curiousity.thanks.

No worries, always appreciate questions and feedback. Back then WOW was much more popular, and I thought I could manage the reasons, but at some point got tired, and kicked it out

 

[Footman]

Nice upgrade for my Vega 64.
The high temps don't worry me as I'll be sticking this under water.
Now just have to find a good home for the Vega 64 and EK waterblock

 

[Prima.Vera]

What's with the crappy performance on the 10x0 cards from nVidia??? Are they pulling again driver slowdowns???

 

[EarthDog]

Are they pulling again driver slowdowns???

This was proven otherwise previously.

 

[INSTG8R]

Nice upgrade for my Vega 64.
The high temps don't worry me as I'll be sticking this under water.
Now just have to find a good home for the Vega 64 and EK waterblock

Yeah while I’m not disappointed I’m my Vega 64s performance this is definitely tempting.

 

[ratirt]

Yeah while I’m not disappointed I’m my Vega 64s performance this is definitely tempting.

Nice upgrade for my Vega 64.
The high temps don't worry me as I'll be sticking this under water.
Now just have to find a good home for the Vega 64 and EK waterblock

I think I'm going to wait for the higher end cards. It is an improvement but I'm going to end up buying the fastest NAVI AMD card so I don't think buying this one makes sense.

 

[INSTG8R]

I think I'm going to wait for the higher end cards. It is an improvement but I'm going to end up buying the fastest NAVI AMD card so I don't think buying this one makes sense.

Yeah I mean I game at 1440 and don’t really have any complaints so yeah hold on for “Big Navi” isn’t unreasonable

 

[tfdsaf]

I do wish the reviews would expand on the following conclusion.

"....beats RTX 2070 ...." yeah, it does outta the box, but not when both cards are overclocked.

From TPU graphs ....

@ 1080p ....

5700XT gets a 100% x 118.1 OCd / 111.8 Stock = 105.63
RTX 2700 gets a 99% x 138.7 / 128.3 OCd / 111.8 Stock = 107.02 ... 1.5 % faster the 5700XT

@ 1440p ....

5700XT gets a 100% x 118.1 / 111.8 = 105.63
RTX 2700 gets a 99% x 138.7 / 128.3 = 105.94 ... 0.2 % faster the 5700XT

@ 2160p ....

5700XT gets a 100% x 118.1 / 111.8 = 105.63
RTX 2700 gets a 100% x 138.7 / 128.3 = 108.11 ... 2.3 % faster the 5700XT

I'm not saying get rid of the stock numbers and use the OC instead. No review would be complete without the performance numbers out of the box. But, since the audience here overwhelmingly OCs their cards, the relativew performance with OC is actually what card selection will be based upon

As far as the 5% lead I saw several times in this thread, the TPU charts show 0%, 1% and 2%. , In addition , a) that's out of the box, b) cards OC'd where ?

1080p = 100 / 99 - h ttps://tpucdn.com/review/amd-radeon-rx-5700-xt/images/relative-performance_1920-1080.png
1440p = 100/ 98 - h ttps://tpucdn.com/review/amd-radeon-rx-5700-xt/images/relative-performance_2560-1440.png
2160p = 100 / 100 h ttps://tpucdn.com/review/amd-radeon-rx-5700-xt/images/relative-performance_3840-2160.png

... and c) the 5700 XT loses at all resolutions with both cards OC'd

Still, just reading this for curiosity's sake ... have to wait for the AIB cards on both sides in August to have their reviews out. It's not like non AIB cards have are something I'd ever recommend. But the temps and sound are untenable options at this point.... If AIB versions of the 5700XT can significantly improve on this and keep the performance deficit when compared to nvidia small when all cards are over clocked, they could cut into nvidias dominance in this segment if they they will remain $50 - $75 cheaper.

A for the sound thing... folks are confusing things ... while +3dB is effectively a doubling of sound pressure level + 10 dbA is a twofold in creases in perceived sound

+ 3 dbA = doubling of power level
+6 dbA = doubling of amplitude
+10 dbA = doubling of loudness

As for pricing ... speculation at this point is just that.

a) Is anyone actually planning to buy a "reference card ?
b) The AIB cards will have better numbers so why talk proce of cards you won't buy ?
c) The "I need to be the 1st one on my block to have the new shiny thing " crowd will keep prices high fpr at least 6 weeks and a lot of this "I only buy (red of green) team stuff".

Here, we won't be recommending any new cards till mid to late August at best.

Literally 99% of the population doesn't OC their graphic cards. Most don't even update drivers if windows doesn't do it for them. So to say that testing needs to be done on OC'ed cards, possibly with GPU samples that are hand picked and press drivers that allow for much better OC is absurd and stupid. If tech sites want to do OC reviews, they should buy 3rd party cards and test them on public drivers released for the masses.

But we know that RX 5000 series have issues overclocking and that the drivers are not mature enough, its absurd to base any conclusion on overclock results right now.

 

[ratirt]

Literally 99% of the population doesn't OC their graphic cards. Most don't even update drivers if windows doesn't do it for them. So to say that testing needs to be done on OC'ed cards, possibly with GPU samples that are hand picked and press drivers that allow for much better OC is absurd and stupid. If tech sites want to do OC reviews, they should buy 3rd party cards and test them on public drivers released for the masses.

But we know that RX 5000 series have issues overclocking and that the drivers are not mature enough, its absurd to base any conclusion on overclock results right now.

I agree. OC is an added value.
With that mature drivers I wouldn't go that far. Even if it's not matured yet, I'm pretty certain if the maturity of the drivers hit the roof, the performance wouldn't get as much gain as you think it would.

 

[HwGeek]

So yo can get 30% performance uplift without quality lose or double the FPS with slight Image softness thanks to Navis Image sharpening, all this with a clock of a button, unlike DLSS .
And you are right, sub 200$ GPU's will be able to get RTX 2060Super and 2070Super performance with this RIS upscaler.

 

[Agent_D]

Given the wildly varying temperatures in reviews around, it would seem the QC on the cooler mating surface is as poor as it seems to be on the R VII, hence the pad to sort of compensate for poor craftsmanship. Stumbled across this trying to find some other info on my R VII and could explain some of the 5700 variances - https://www.reddit.com/r/Amd/comments/arrxt2

 

[Super XP]

AMD finally has a GPU that competes very well against Nvidia and gives people another option. That said, the Blower Style Cooler needs to GO. Not sure what AMD is thinking but I ain't buying this, I am going to wait for the Custom Coolers by Sapphire and PowerColour.

No, it was fairly obvious from the start that the XT would perform more or less in line with the vanilla 2070 (and thus a tad faster than the 2060S). The way things stand now, I can't think of a single reason to go for the red card which is clearly inferior feature-wise and runs red-hot out of the box with literally no extra room for overclocking. Zen 2 CPUs, on the other hand, appear a really strong product, so it's quite likely my next build will be an AMD+Nvidia combo for the first time in my life.

I'm going for the 5700 XT with custom cooling by either Sapphire or PowerColour.
No way would I buy an Nvidia GPU after that GPP nonsense they tried to push forth and backfired for obvious reasons.

 

[Marecki_CLF]

92'c? wow.so much for 7nm , when you go lower nm , You will get more hard cap.I don't expect 5 nm or 3 nm to give miracle.

RX5700XT runs at 1.2V at default, which is insanely high. Dropping the core voltage to 1.0V results in 72C under load and fan speed at 2000-2400RPM. At this setting the card holds boost clock of around 1820-1830MHz and consumes roughly 155W.

 

[Dave65]

The card is great, the cooler BLOWS, literally

Id buy one or trade one of my 1080 tis for one just to play around with.

 

[Agent_D]

Given the wildly varying temperatures in reviews around, it would seem the QC on the cooler mating surface is as poor as it seems to be on the R VII, hence the pad to sort of compensate for poor craftsmanship. Stumbled across this trying to find some other info on my R VII and could explain some of the 5700 variances - https://www.reddit.com/r/Amd/comments/arrxt2

I'll still have to get my hands on a 5700xt for testing this out, but it is definitely a YMMV type thing. On my R VII I only saw drops of ~5-10c off junction temps with a lapping job on the stock cooler, however, I just did a fairly quick lapping on my girlfriends Vega 56 (flashed to 64) 2 days ago and saw the junction temps drop 25-35c depending on loads. If you're going to have the heatsink off anyways, worth the few bucks and 30 minutes of labor (probably a bit more if you got a really poor machining job).

 

[Durvelle27]

Mine should be here within a week

 

[Wonderer]

RX 5700 XT's temp :
Kitguru = 80'c
Computerbase = 85'c
overclock3d = 68'c
Techradar = 81'c
TPU = 92'c

hmm , TPU got worst card.

I had to lol, maybe with fan at 100%

You mean that Your card did not go fan 100% and temps were 92 degrees? Or what are You implying?

Well now, if I may ask, did You experience or notice something like Techspot in their review?

Temperature & Overclocking

Out of the box the Radeon RX 5700 XT runs at 84 degrees and maintains a core clock speed of 1740 MHz and a memory speed of 872 MHz which is basically 14 Gbps. The blower fan typically spun at 2100 RPM in our testing.
Overclocking the 5700 XT saw the core peak at just over 1940 MHz but at times would still drop as low as 1720 MHz and on average we saw a clock speed of 1860 MHz, which is a ~10% overclock. The massive fluctuation was due to the almost 90 degree operating temperature, at 88 degrees the GPU tends to throttle and with the fan running at 43% or 2100 RPM, there was room to move. As for the memory, we were only able to get the GDDR6 up to a transfer speed of 14.3 Gbps, so that was a bit disappointing.

With the core clock jumping all over the place we decided to crank the fan up to 100% -- where it’s incredibly noisy as it spun at 4600 RPM -- in this scenario the GPU peaked at 70 degrees and the 5700 XT was able to maintain a core clock speed of over 2 GHz at all times. That’s about a 20% overclock from stock, I reckon liquid cooling is going to be a popular option for these 5700 XT graphics cards.
As for the standard 5700, out of the box it ran at 75 degrees with a fan speed of 1900 RPM and maintained a typical operating frequency of 1670 MHz. Then when overclocked with the auto fan speed it hit 85 degrees, but the fan was only spinning at 1600 RPM. We can assume this is a bug with the early revision drivers.

I mean if You were also at their temps and fan did NOT crank up, then You had thermal throttling also and couldn't get everything out of the card?

As far as I can see in the fan speed did not go to 100% when You were testing it and it was throttling?

Why? Care to retest with proper fan speeds, if I ask kindly?

 

[W1zzard]

You mean that Your card did not go fan 100% and temps were 92 degrees? Or what are You implying?

look at the text that i quoted

 

[Super XP]

Mine should be here within a week

I'll be waiting for custom cooling solutions, from Sapphire as my #1 choice then ASRock and PowerColour as secondary choices depending on the price of course.
A blower Style cooler was a mistake by AMD unfortunately, not sure why they continue to come out with this type of cooling solution. Not to mention, the temps are high because the steel plate is not sitting properly and the chip is getting about 10C to 20C more hotter than it should. This was tested on another site. Don't remember the name ATM.

 

[Fluffmeister]

AMD seem to mess up launches time and time again, all this time waiting for Navi, and again the custom solutions aren't availble for weeks or even months later.

At least the team that creates the vast array of Powerpoint slides (with fake MSRP's) are kept busy.

 

[Eykxas]

I have a 5700XT with reference design. (Blower).

I have a case properly vented (NZXT S340) and after many hours of playing, the temps didn't exceed 72 / 73°. With a little noise (less than a Xbox One X).

If I oc my card (2120 MHz reported, ~2080Mhz real) the temps increase at 85°. I have manually tweak the fan curve. The noise increase too but it's acceptable.

So, the blower is not very good cooling but if you have a good batch for your card, it's correct with no real complain.

 

[Super XP]

Those are more than acceptable temps for hours of game play.
Hopefully moving forward, AMD never releases a blower style cooling GPU again.