** The AMD Navi Thread **

[Besty]

OH CRAP! That's from today's review. I completely missed that!
This invalidates the performance aspect of their reviews!!!

yeah and they better have some answers for it to as that is quite a serious mistake they made there.

 

[bru]

Sadly. These 2-month old Beta drivers leave significant amounts of performance, and since then AMD has released the 19.9.2 WHQL.

nvidia's cards are tested with some special "press release" driver, while AMD's cards with the oldest Beta

Screams nvidia bias and large cash payments incomes towards the wizzard...

Quite agree that it points to bias, not sure about screaming it though and where on earth did the large cash payments come from.

 

[EastCoastHandle]

Quite agree that it points to bias, not sure about screaming it though and where on earth did the large cash payments come from.

This is quite huge in the grand scheme of things though. Sure he might have over embellished about it. However, this type of taboo is something that most, if not all, reviewers avoid.

 

[Panos]

Sadly. These 2-month old Beta drivers leave significant amounts of performance, and since then AMD has released the 19.9.2 WHQL.

nvidia's cards are tested with some special "press release" driver, while AMD's cards with the oldest Beta

Screams nvidia bias and large cash payments incomes towards the wizzard...

Quite agree that it points to bias, not sure about screaming it though and where on earth did the large cash payments come from.

That's the same site pointed yesterday on another discussion, that shows the 2070S having 33db on Load!!!!! While every other reviewer shows 34db on IDLE!!!!!!!!!!

 

[Treetops422]

All said and done if you slide the ref cards up to 100% fan speed vs any aib it wins in thermals\OC. Loud yes, does it need to be 100% to beat aib no? At the same noise level as aib at any DB is it the best price vs performance? Yes. Something no reviewer cares to share.

GN is only human you would be hard pressed to find higher a quality of reviewer in all regards.

 

[4K8KW10]

Quite agree that it points to bias, not sure about screaming it though and where on earth did the large cash payments come from.

Why would they sabotage their entire activities with unnecessary unprofessional "reviews" if they don't receive some type of financial benefit/stimulus in order to offset the potential loss of business?
This is a recipe for disaster for them.

You know that we here agree that the global advice here is to avoid the website now?

Also, why doesn't AMD address the issues using some sanctions against techpowerup?

 

[Darujhistan]

Yeah I don't trust him any more either, he seems biased and he's turned into bit of a tool.

 

[Ross Thomson]

don't have issues with his testing. The XFX does have poor cooling just like the msi evoke. He is not the only one to mention the high memory temps. Wouldn't consider buying one over the much cheaper pulse.

 

[Treetops422]

If someone knows of a better reviewer then GN that speaks English lmk. Igorslab ftw, but that's in German sometimes cc in English. For what he does he does it well.

 

[4K8KW10]

OK bored of all the 5700XT stuff, give me some Navi 20 goodness

It's quite interesting if they indeed have Navi 12 as RX 5800 to tackle RTX 2080. The difference in performance between Navi 10 and Navi 12 would be minimal.

https://wccftech.com/amd-navi-12-radeon-rx-5800-and-navi-14-rx-5700-gpu-specs-detailed/

 

[TonyTurbo78]

OK bored of all the 5700XT stuff, give me some Navi 20 goodness

Is Navi20 going to be 2nd Gen RDNA? I've lost interest in GPUs these last 12 months.

 

[4K8KW10]

Is Navi20 going to be 2nd Gen RDNA? I've lost interest in GPUs these last 12 months.

Navi 20 should be Navi 10 v.2.0, like Polaris 10/20/30 being the same thing, maybe the difference would be revisions on different steppings.

 

[Panos]

All said and done if you slide the ref cards up to 100% fan speed vs any aib it wins in thermals\OC. Loud yes, does it need to be 100% to beat aib no? At the same noise level as aib at any DB is it the best price vs performance? Yes. Something no reviewer cares to share.

GN is only human you would be hard pressed to find higher a quality of reviewer in all regards.

What annoys me is that GN and few others is they shove a microphone to the back of the card, which is capturing the moving air hitting the microphone, not the actual noise of the card (fan, air) it would do half meter away at a usual position everyone's computer is.
And then you as card owner who sees and measures your card isn't doing over 32db (and the soft moving air sound not some high pitched screaming fan), you have to argue with morons who say the card is loud.
Because some muppet who makes money on youtube spread nonsense.

 

[nashathedog]

To be fair he did defend nvidia's NDA 2.0 and got a "lawyer" who completely contradicted him. The lawyer claimed that the NDA was nebulous among other concerns while Steve touted nothing was wrong with it more or less.
That is when I stopped watching him.

I'm sure with a bios update the card will do a whole lot better. But it wouldn't be a card I would want. I prefer 3 Fan design as they tend to be more quieter.

I tend to not bother with his videos as every video tends to be too long & too boring a bit like Tiny Tims,

As for the XFX I think the biggest issue is how boxed in the XFX card is, so it'll need fast fans under load regardless.
As for the Taichi there's too much going on with the shroud making it look ugly.

 

[bru]

What annoys me is that GN and few others is they shove a microphone to the back of the card, which is capturing the moving air hitting the microphone, not the actual noise of the card (fan, air) it would do half meter away at a usual position everyone's computer is.
And then you as card owner who sees and measures your card isn't doing over 32db (and the soft moving air sound not some high pitched screaming fan), you have to argue with morons who say the card is loud.
Because some muppet who makes money on youtube spread nonsense.

Surely if all the cards are measured the same it is a valid way to measure the sound?
Maybe a different way than what others do but no less valid.

Of course this is just my opinion and as we know opinions are like bottoms, everybody has one and often they are full of *@$#&.

 

[james.miller]

Valid, if somewhat misrepresentative.

 

[Panos]

Surely if all the cards are measured the same it is a valid way to measure the sound?
Maybe a different way than what others do but no less valid.

Of course this is just my opinion and as we know opinions are like bottoms, everybody has one and often they are full of *@$#&.

There is a small issue on the method. Non blower coolers distribute the heat all around them with small pressure. Blower coolers push all the air in a single direction with high pressure.
To give you an example why this method is the wrong way to measure a GPU noise here is an example.

If you ever had phone call during wind blowing, you always had the other person complaining about the noise, even if yourself couldn't hear a thing.
Similarly if you talked to someone who was against wind blowing, you always hear the extra high volume and pitch noise through the microphone but the other person is oblivious to it.

That's the point we discuss here, that's why the methods are wrong when measuring blower coolers, especially the 5700XT one.
Why especially the 5700XT? The Pascal/Maxwel/Kepler/Vega 64/290X blowers were noisy because of the different tone due to the fan was making most of the noise and no one argues here, however is unfair and idiocy to measure apples and oranges putting the 5700XT in the same category with cards like the Vega 64, 290X, GTX1080FE, 1080TI FE, 980Ti FE and the list goes on.
Every reference 5700XT owner will say the same, and I believe @LtMatt would agree on me also, as he has the AE like myself.

 

[EastCoastHandle]

I tend to not bother with his videos as every video tends to be too long & too boring a bit like Tiny Tims,

As for the XFX I think the biggest issue is how boxed in the XFX card is, so it'll need fast fans under load regardless.
As for the Taichi there's too much going on with the shroud making it look ugly.

Agree with you on that.

Overall all the AIB cards are better. However, a few need to iron out issues.
MSI with the thermal pads
XFX Bios fix

There is a small issue on the method. Non blower coolers distribute the heat all around them with small pressure. Blower coolers push all the air in a single direction with high pressure.
To give you an example why this method is the wrong way to measure a GPU noise here is an example.

If you ever had phone call during wind blowing, you always had the other person complaining about the noise, even if yourself couldn't hear a thing.
Similarly if you talked to someone who was against wind blowing, you always hear the extra high volume and pitch noise through the microphone but the other person is oblivious to it.

That's the point we discuss here, that's why the methods are wrong when measuring blower coolers, especially the 5700XT one.
Why especially the 5700XT? The Pascal/Maxwel/Kepler/Vega 64/290X blowers were noisy because of the different tone due to the fan was making most of the noise and no one argues here, however is unfair and idiocy to measure apples and oranges putting the 5700XT in the same category with cards like the Vega 64, 290X, GTX1080FE, 1080TI FE, 980Ti FE and the list goes on.
Every reference 5700XT owner will say the same, and I believe @LtMatt would agree on me also, as he has the AE like myself.

I really don't understand why reviewers cry that the sky is falling when they already know the characteristics of a blower style video card. It's function over form more or less. It will do the job however, it's going to be louder by default. And the actual noise has always been the air pressure not the actual fan itself. The blower is very quiet however the air pressure it makes generates the noise.

I would like to know if the RX 5700 AE is 7 phase or something else?

 

[Phixsator]

Are there any reviews out there where a proper tuned Vega is put up against a 5700XT? I don't trust the numbers from channels like Hardware unboxed as they have clearly shown they don't bother tuning their Vega card(I suppose you could make a case for why you shouldn't have to)

 

[Panos]

Are there any reviews out there where a proper tuned Vega is put up against a 5700XT? I don't trust the numbers from channels like Hardware unboxed as they have clearly shown they don't bother tuning their Vega card(I suppose you could make a case for why you shouldn't have to)

Vega 64 LC is comparable to a well tuned Vega 64 (if not bit better).
Here is a review
https://babeltechreviews.com/the-rx...cooled-rx-vega-64-and-vs-nvidias-super-cards/

I would like to know if the RX 5700 AE is 7 phase or something else?

7 phase with 8+6 pin connectors. But the chips are very well binned. I have downclocked (AMD these days ) and downvolted mine. Not only runs cooler but faster also than out of the box.