Review AMD RX 5600XT Review Thread

[Hitman928]

Print Media

The AMD Radeon RX 5600 XT Review, Feat. Sapphire Pulse: A New Challenger For Mainstream Gaming

www.anandtech.com

AMD Radeon RX 5600 XT im Test

Mit der Radeon RX 5600 XT platziert AMD eine Grafikkarte auf Augenhöhe zur Nvidia GeForce RTX 2060. Die Effizienz ist sehr hoch.

www.computerbase.de

https://translate.google.com/translate?sl=auto&tl=en&u=https://www.computerbase.de/2020-01/radeon-rx-5600-xt-test/ [translated]

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

1660 Ti-like pricing and better-than-2060 performance!

www.tomshardware.com

Video Reviews

From early reviews it seems like the 5600XT and Nvidia 2060 are basically tied both in performance and power consumption although GN's 12V rail power numbers and Anandtech's system power numbers don't agree, not sure why that is.

 

[railven]

After watching the GN video, I'm starting to feel like AMD created a needless headache for itself while missing a big positive good will win.

As noted by Steve and other reviewers, that bios update is enough to create a new SKU. Had AMD just put out the 5600 with the old bios out at say $260, they'd have won a lot of good will, and if they waited a week (or two factoring the holiday for the region) they could have rolled out mouth watering OC versions for $280+.

Two markets catered to, AIB vendors aren't left holding a bag, and there is no russian roulette (is anyone else thinking this could qualify for bait and switch?) on whether you get a 12gbps or 14gbps card?

So much headache for what? There is a point in the video where Steve captures my exactly. "Just do better." AMD's worse enemy (in the GPU sector) continues to be AMD.

 

[Glo.]

After watching the GN video, I'm starting to feel like AMD created a needless headache for itself while missing a big positive good will win.

As noted by Steve and other reviewers, that bios update is enough to create a new SKU. Had AMD just put out the 5600 with the old bios out at say $260, they'd have won a lot of good will, and if they waited a week (or two factoring the holiday for the region) they could have rolled out mouth watering OC versions for $280+.

Two markets catered to, AIB vendors aren't left holding a bag, and there is no russian roulette (is anyone else thinking this could qualify for bait and switch?) on whether you get a 12gbps or 14gbps card?

So much headache for what? There is a point in the video where Steve captures my exactly. "Just do better." AMD's worse enemy (in the GPU sector) continues to be AMD.

AMD planned this SKU to be able to fight both ways: either with GTX 1660 Ti/Super, or with RTX 2060 if Nvidia would cut the prices.

The rushed BIOS was because they did not anticipated MSRP price cut so late.

And no. This is way overblown. The most screwed are the reviewers, not the actual users, unless they buy the GPUs without new BIOS, and have no idea, there is new BIOS for this GPU.

P.S. If it would be so, as everyone says, why RX 5600 XT boards in general are so overkill from cooling and power delivery perspective, if they would be not designed to handle the increased clock speeds?

 

[railven]

AMD planned this SKU to be able to fight both ways: either with GTX 1660 Ti/Super, or with RTX 2060 if Nvidia would cut the prices.

The rushed BIOS was because they did not anticipated MSRP price cut so late.

They could have done all this without shooting themselves in the foot. This is where I keep facepalming with AMD.

Roll out 5600 XT as planned, NV price cuts 2060, either A) price cut 5600 XT or B) leave be, tease NV/audience on twitter "just wait for the lunar festivities to quell...just wait...", launch 5600 XT OC's models with 10-12% more performance! for $280+

They could have handled this so much better, even you have to agree with that.

 

[Stuka87]

There will likely be an increase in RMAs for the 14gbs bios flashed cards (since no memory validations before shipping).

Unlikely. We are talking about a small number of cards that were already shipped before release date. And even then only the people that actually flash them might have an issue, as I am betting the majority of buyers won't. Because, the majority of PC hardware buyers really are not up to date on news. They walk into Best Buy/Fry's/MicroCenter, see a card with a fancy box that's in their price range, and buy it.

So we are talking a percentage of a percentage.

So sure, AMD handled this not the best. They could have done a 5600 and a 5600XT. But regardless of this, the issue has been blown way out of proportion.

 

[Glo.]

They could have done all this without shooting themselves in the foot. This is where I keep facepalming with AMD.

Roll out 5600 XT as planned, NV price cuts 2060, either A) price cut 5600 XT or B) leave be, tease NV/audience on twitter "just wait for the lunar festivities to quell...just wait...", launch 5600 XT OC's models with 10-12% more performance! for $280+

They could have handled this so much better, even you have to agree with that.

I do not see this as big of a problem as most people do. But I am the weird one here.

 

[GaiaHunter]

Keep hittin' the dead horse, man.

The problem does exist with some DX9 games at 1080p.
As you go up in resolution the card will get closer to the proper clocks.

The solution is using VSR and play at 1440p or 4K.
But I've noticed some improvements with new drivers.

Before while playing GW2 the core clocks would drop as low as 7MHz at 1080p, now it is 700ish MHz. At 7Mhz you could get some micro freezes. The performance was no better than with a RX480 sometimes worse.
If I bump the resolution to 1440p the lowest it downclocks is 1000ish MHz.
At 4K it stays above 1900MHz and often it would beat the 1080p performance. WIth the most recent drivers 1080p performance is better than 4K.

 

[railven]

I do not see this as big of a problem as most people do. But I am the weird one here.

The magnitude of the problem will be subjective. But I'm glad you admit you still see it as a problem.

 

[Hitman928]

The problem does exist with some DX9 games at 1080p.
As you go up in resolution the card will get closer to the proper clocks.

The solution is using VSR and play at 1440p or 4K.
But I've noticed some improvements with new drivers.

Before while playing GW2 the core clocks would drop as low as 7MHz at 1080p, now it is 700ish MHz. At 7Mhz you could get some micro freezes. The performance was no better than with a RX480 sometimes worse.
If I bump the resolution to 1440p the lowest it downclocks is 1000ish MHz.
At 4K it stays above 1900MHz and often it would beat the 1080p performance. WIth the most recent drivers 1080p performance is better than 4K.

There's no way it was actually running at 7 MHz, you wouldn't see micro stutters, you would see a slide show and performance wouldn't be anywhere near an RX480. Either there was a bug in the OSD software or something like Radeon Chill was on and at times the card was putting out the same frame over and over again and dropping to idle but in that case you should have had extreme lag in the controls.

Also, at 700 MHz, it's probably enough GPU horsepower to hit a CPU bottleneck in GW2. Sounds like (not having experienced it) there was a bug in the early drivers on some games that would cause the GPU to drop essentially into idle momentarily which caused a lot of dropped frames where the micro stutters would occur but that it has since been fixed and now it is just normal clock back-off to save power in the face of a CPU bottleneck. Possibly an over agrressive power saving algorithm on really low GPU demand games (hence why it showed up in some old DX9 games).

 

[VirtualLarry]

My OSD for my RX 5700 (XFX Reference Blower model), showed numbers like 7Mhz and stuff too, while at the Windows Desktop, and starting up mining. Once the mining actually "Got going", things changed up a bit, and it started reporting "real" clocks. Just really kind of eerie, seeing a 7Mhz GPU clock. How does it even paint the desktop at that clock? (I guess it can.)

 

[CHADBOGA]

Why did AMD feel the pressure to have a price/performance effective card with the 5600XT, when they did not for the 5500XT?

 

[CHADBOGA]

And no. This is way overblown. The most screwed are the reviewers, not the actual users, unless they buy the GPUs without new BIOS, and have no idea, there is new BIOS for this GPU.

P.S. If it would be so, as everyone says, why RX 5600 XT boards in general are so overkill from cooling and power delivery perspective, if they would be not designed to handle the increased clock speeds?

The boards can be overkill, but in many instances the memory won't be up to the increase in speed.

MSI for example, won't be putting out the bios for their launch card, as they don't believe the memory chips they are using can cope.

 

[Glo.]

The magnitude of the problem will be subjective. But I'm glad you admit you still see it as a problem.

For some people it might be. For me - whole launch is way better than I actually expected, because of the end result.

 

[DooKey]

AMD should have know Nvidia would have done something to try to spoil this launch. They were going to launch a 1660 Super Ti, or a 2060 LE, or just reduce the price of the 2060. Nvidia just can't let AMD have a clear win, their ego requires them to do something. What the something is may not be clear, but that something will be done is guaranteed. They should have had a viable response ready.

Now AMD has a mess with both board partners and customers. Steve is right, they keep shooting themselves in the foot.

LOL. Why would a business let another business have a clear win? I just don't understand why people are so tied to corporations.

 

[guachi]

The number of cards that launched that couldn't be upgraded was probably small. I agree that AMD should have just lobbed some incentives at the board partners and allowed them to clear inventory on a price cut and ALSO released a BIOS-updated 5650XT or something at the $290 level.

Any 5600XTs in the wild that could take the new BIOS could get flashed by the user and he gets FREE EXTRA PERFORMANCE! If it can't, you've not lost any performance (but you did gain a price cut).

That'd cause every card that could be flashed that to disappear from the shelves. People love feeling like they got something for nothing.

 

[insertcarehere]

Unlikely. We are talking about a small number of cards that were already shipped before release date. And even then only the people that actually flash them might have an issue, as I am betting the majority of buyers won't. Because, the majority of PC hardware buyers really are not up to date on news. They walk into Best Buy/Fry's/MicroCenter, see a card with a fancy box that's in their price range, and buy it.

So we are talking a percentage of a percentage.

So sure, AMD handled this not the best. They could have done a 5600 and a 5600XT. But regardless of this, the issue has been blown way out of proportion.

With the unfortunate development of the Coronavirus extending Lunar New Years for everyone in China, affecting production of Taiwanese/Chinese board partners (that's most of the AIBs). I suspect cards with updated VBIOS out of the box will be very thin on the ground for Q1 2020.

 

[Zstream]

With the unfortunate development of the Coronavirus extending Lunar New Years for everyone in China, affecting production of Taiwanese/Chinese board partners (that's most of the AIBs). I suspect cards with updated VBIOS out of the box will be very thin on the ground for Q1 2020.

Can we just stick with facts please? There have been very little so far, which goes to show you the state we’re in. There is very little insider information, desire to look at PCB’s, or good tech talk. It’s a bunch of tubers that quite frankly attract a segment of the population due to reasons I don’t care to elaborate on.

The card is a cut down version of the 5700, which by all accounts will have sufficient memory specs. We don’t have all the information, and it’s unlikely a board partner HAS to go through a process that will cause them financial loss. Unfortunately, that’s speculation as well. I’ll wait for Anand to do a deep dive... oh wait.

 

[lifeblood]

LOL. Why would a business let another business have a clear win? I just don't understand why people are so tied to corporations.

And I don't disagree with you. "Spoiling" a competitors big event has a long history and keeps things exciting for us the consumers/enthusiasts. The problem was not that Nvidia did something to steal AMD's thunder, I wasn't making a value judgement on that part, the problem was that AMD didn't plan for it and instead pulled a knee jerk reaction that will have unpleasant consequences for companies and consumers. How big of a problem is yet to be seen. I hope it turns out to be a minor problem that were blowing out of proportion.

 

[lobz]

Can we just stick with facts please? There have been very little so far, which goes to show you the state we’re in. There is very little insider information, desire to look at PCB’s, or good tech talk. It’s a bunch of tubers that quite frankly attract a segment of the population due to reasons I don’t care to elaborate on.

The card is a cut down version of the 5700, which by all accounts will have sufficient memory specs. We don’t have all the information, and it’s unlikely a board partner HAS to go through a process that will cause them financial loss. Unfortunately, that’s speculation as well. I’ll wait for Anand to do a deep dive... oh wait.

it's a new tapeout with erratas, but yes, essentially the same chip

 

[Stuka87]

it's a new tapeout with erratas, but yes, essentially the same chip

You sure about that? It looks as though this uses chips that were binned to not be adequate for the 5700/5700 XT. It would seem strange to tape out a new chip that is functionally the same. The only difference from a 5700 is the narrower memory bus and slower clocks.

 

[Shmee]

Interestingly, from what I am hearing the 5600XT is one of the best, if not the best, Ethereum miner now.

 

[VirtualLarry]

Interestingly, from what I am hearing the 5600XT is one of the best, if not the best, Ethereum miner now.

Now THAT's interesting. To me, at least. Better than the GTX 1660ti, finally, and RX 5700?

 

[Shmee]

According to this one article, yes. They were getting around 51MH/s at about 95W. So greater hashrate than 1660Ti, and more efficient than the 5700. Also, lower cost than the 5700.

 

[misuspita]

In my country, 5600XT is nowhere at the msrp, it was listed as such before you could buy one, but now it's at exactly the same price as 2060... which is aprox. 333 euros

 

[Stuka87]

In my country, 5600XT is nowhere at the msrp, it was listed as such before you could buy one, but now it's at exactly the same price as 2060... which is aprox. 333 euros

Does your country have VAT or anything that adds into the price? Out of curiosity, what does a 5700 go for? Since 333 euros is over the cost of a 5700 here in the states.

 

[misuspita]

Yes, VAT is included (19%) and all other taxes, that's the total price to pay. Price for lowest 5700 is ~384€