Spin Off: AT's Testing Methods & Uber Mode

[krumme]

As i can read:

In the 16 pages review of the 290 not a single time is it mentioned the 290x runs on 40% or quiet mode.

Am i reading right - have i overseen it ??? If;

How on earth is that possible? All that writing and not a single time is it mentioned directly.

There is absolutely no excuse for not mentioning setup conditions. Thats just pathetic.

 

[Janooo]

There's really no need.
R290X is priced to compete with the $500 GTX780 which it easily defeats.
GTX780 Ti is a $700 card and it would be a complete failure if it didn't beat the R9 290X at $550

... and 780Ti does not beat 290X in uber mode based on AT's own numbers.
I guess it is a failure

 

[raghu78]

AT's GTX 780 Ti review was bad. they flat out ignored the fact that R9 290X CF in uber mode was 20% faster than GTX 780 SLI at 4K and came up with a pathetic conclusion that both perform same at 4k. yeah a badly throttling R9 290X CF in quiet mode is 5.5% faster than the GTX 780 SLI while the R9 290X CF uber is 21% faster. the uber mode has been found to be definitely usable for 24/7 gaming by other reviewers like brent of[H] and scott of TR and has been totally ignored by AT. the noise issue was badly overblown.

 

[f1sherman]

NOW I get it. I was like WTF... what they mean no Uber mode

This 24 pages of complains is not because AT didn't include Uber mode. Because they clearly did include uber mode. For example:

It is because Andantech dared to test and comment 290X in default mode!!!!

OKIE... Understood.
#carry-on

 

[Subyman]

NOW I get it. I was like WTF... what they mean no Uber mode

This 24 pages of complains is not because AT didn't include Uber mode. Because they clearly did include uber mode. For example:

It is because Andantech dared to test and comment 290X in default mode!!!!

OKIE... Understood.
#carry-on

You are only half correct. The analysis did not include the 290X uber mode, so the charts did not align with the conclusion. The bigger issue is that Ryan said testing in uber mode was a one-time event. This leads us to believe all future reviews will not include the 290X's uber mode.

 

[Attic]

If anytyhing, that is the reason why reviewers shouldn't discard uber profile. They should bring viewer's attention to double bios switch. Show the features, so less "enthusiast" will miss them, or be confused about them.

Yep, have to agree with this. For my money [H]ardOCP and a few other sites definitely has a better approach than Anandtech on testing, evaluating, and reviewing the card and it's competition.

 

[felang]

Lets say you have 1 card with 100 modes.

One of them will be the lowest performance and one of them will be the highest performance.

Everything else will be in between.

Took me 1 second to realize this.

So obviously you test the lowest performance mode and the highest performance mode and claim that there is to time to review all the other modes.

Of course this is all looking at straws.

What is the point of different SKUs if one has a card with wildly variable modes?

Cards being release with 4 bios?
Yeah, right.

And what is this silly talk of aligning vectors?
We buy cards to play games - the vectors that matter are ability to play games, IQ, performance and price.
All the other are minor points that we take in account after the first big 4.

In reviews you align ability to play games (fortunately these days cards generally work fine), IQ and then get performance.

Excellent Post

 

[blackened23]

You are only half correct. The analysis did not include the 290X uber mode, so the charts did not align with the conclusion. The bigger issue is that Ryan said testing in uber mode was a one-time event. This leads us to believe all future reviews will not include the 290X's uber mode.

I think Ryan Smith would prefer that uber mode didn't exist. Why should there be a BIOS switch at all? Hypothetically speaking, would reviewers ever have to say anything to B revision cards with only one mode? To be clear, I mean one more that combines uber performance with quiet mode noise levels which are similar to competing Kepler GPUs. Would Ryan Smith have anything bad to say then? Would there be any question of which mode he benchmarks? I think not. That's what Ryan Smith and probably most consumers want.

AMD can still fix that. Nvidia got slammed for the GTX 480, just like AMD is getting slammed now. IMO, no matter how much some guys around here feel that AMD is the victim - a lot of that criticism is justified. That BIOS switch and compromise modes should not exist - and AMD can still fix it with B revision updated cards. Is this unreasonable? No, no I think not. Nvidia's B revision card was the GTX 580 and IMHO it was a freaking home run product at the time. AMD can do the same. The question is, will they? I think they should. I understand that some of you guys are content with how things are, though.

Come on dudes. This isn't an unreasonable request AT ALL from the number of reviewers hesitant to test uber mode (and there a lot of websites not testing it, actually), and the users who don't like the noise vs performance compromise. B revision cards!

 

[TerryMathews]

I think Ryan Smith would prefer that uber mode didn't exist. Why should there be a BIOS switch at all? Hypothetically speaking, would reviewers ever have to say anything to B revision cards with only one mode? To be clear, I mean one more that combines uber performance with quiet mode noise levels which are similar to competing Kepler GPUs. Would Ryan Smith have anything bad to say then? Would there be any question of which mode he benchmarks? I think not. That's what Ryan Smith and probably most consumers want.

AMD can still fix that. Nvidia got slammed for the GTX 480, just like AMD is getting slammed now. IMO, no matter how much some guys around here feel that AMD is the victim - a lot of that criticism is justified. That BIOS switch and compromise modes should not exist - and AMD can still fix it with B revision updated cards. Is this unreasonable? No, no I think not. Nvidia's B revision card was the GTX 580 and IMHO it was a freaking home run product at the time. AMD can do the same. The question is, will they? I think they should. I understand that some of you guys are content with how things are, though.

What does this have to do with the inconsistency in the testing policy?

 

[blackened23]

What does this have to do with the inconsistency in the testing policy?

What this has to do with the testing policy? I think I explained pretty clearly that without a BIOS switch and with B revision cards, this never has to be questioned. That's what i'm sure Ryan Smith wants, as well as the users who reads his reviews. He doesn't want to pick "quiet" or "performance".

..............just to be clear, there are NUMEROUS websites not testing uber mode right now. Does techspot have users complaining about the lack of uber mode testing? Pcper? Guru3d? No, not really. Only AT has users complaining about a lack of uber mode testing. But let's get to the actual issue at hand: if AMD fixes the actual issue, you guys won't have to complain about the lack of testing of uber mode. Right? I sure don't see anyone complaining to PCPer about their lack of uber mode testing. This seems like an ALL TOO OBVIOUS fix to me. B revision cards without that stupid freaking BIOS switch - and a new updated design that combines the highest performance with better acoustics. Maybe you disagree. Do you feel updated cards is an unreasonable request? To my knowledge, that's what Ryan Smith told AMD in his feedback to them. Update the design.

Like I said -Clearly, AT is one of MANY MANY websites that have expressed disinterest in testing uber mode - while it's great for them to clarify their policy on this, why can't AMD fix the actual, uhm, problem? That isn't an unreasonable proposition.

 

[Erenhardt]

That only makes me think nvidia has something to do with the issue.
Even if you slap titan cooler on 290X, it will still throttle with fan in quiet mode. Didn't change a thing.

What is next? Nvidia will release passive cooled reference cards, so we will disable fans on amd designs, so it is "fair"?

 

[Tleilaxu]

No reviewer should have to put their health at risk to review a GPU. AMD have already shown they are willing to push the limits of human audible decency to squeeze out a few extra Hertz.

If they aren't stopped now it won't be long until they release a card that actively deafens anyone in the surrounding vicinity all in the pursuit of performance.

 

[blackened23]

Even if you slap titan cooler on 290X, it will still throttle with fan in quiet mode. Didn't change a thing.

No, it really wouldn't. The titan shroud can dissipate over 250W while maintaining quiet sound levels. From my use of the 780 reference briefly, that shroud can be turned up to ridiculously high manual fan levels and still maintain good acoustics. The cut-off for AMD in terms of quietness right now is around 45%.

But back to my point. AMD needs to fix the issue, that way you guys don't need to complain about it. MANY websites aren't testing uber mode - the only difference is, AT is the only website with users complaining about it. Oddly enough. I dont' see users complaining at guru3d, pcper, or techspot. There are tons of other sites not testing uber mode either.

NOT UNREASONABLE to request an updated design "B" revision card. You know? Come on, you know that isn't unreasonable to ask. AMD can do that just like nvidia did it with the GTX 580 (to update the 480).

 

[GaiaHunter]

But back to my point. AMD needs to fix the issue, that way you guys don't need to complain about it. MANY websites aren't testing uber mode - the only difference is, AT is the only website with users complaining about it. Oddly enough. I dont' see users complaining at guru3d, pcper, or techspot. There are tons of other sites not testing uber mode either. NOT UNREASONABLE to request this.

What does loudness of quietness of the fan have to do with testing uber mode or not?

From what Anand and Ryan wrote it is based on of the box settings not on how loud or not it is.

 

[Erenhardt]

B revision cards without that stupid freaking BIOS switch - and a new updated design that combines the highest performance with better acoustics.

Even if you slap titan cooler on 290X, it will still throttle with fan in quiet mode. Didn't change a thing.

No, it really wouldn't. The titan shroud can dissipate over 250W while maintaining quiet sound levels. From my use of the 780 reference briefly, that shroud can be turned up to ridiculously high manual fan levels and still maintain good acoustics. The cut-off for AMD in terms of quietness right now is around 45%.

By some reviewers 40% is the default settings for 290X. Not 45%, not quiet-enough-but-above-40%, 40%! And it will throttle, just like titan/780 would.

cards without that stupid freaking BIOS switch

Now, how is dual bios a bad thing, stage is yours...

 

[Erenhardt]

What does loudness of quietness of the fan have to do with testing uber mode or not?

From what Anand and Ryan wrote it is based on of the box settings not on how loud or not it is.

Review sites no longer review loud cards I guess.

 

[BallaTheFeared]

Review sites no longer review loud cards I guess.

Evidently when given an option, reviewers take considerably quieter operation at the cost of a few percent of performance.

290 series looks like a stinker as an overclocker anyways, might as well enjoy decent performance without gale force wind noise.

 

[TerryMathews]

What this has to do with the testing policy? I think I explained pretty clearly that without a BIOS switch and with B revision cards, this never has to be questioned. That's what i'm sure Ryan Smith wants, as well as the users who reads his reviews. He doesn't want to pick "quiet" or "performance".

..............just to be clear, there are NUMEROUS websites not testing uber mode right now. Does techspot have users complaining about the lack of uber mode testing? Pcper? Guru3d? No, not really. Only AT has users complaining about a lack of uber mode testing. But let's get to the actual issue at hand: if AMD fixes the actual issue, you guys won't have to complain about the lack of testing of uber mode. Right? I sure don't see anyone complaining to PCPer about their lack of uber mode testing. This seems like an ALL TOO OBVIOUS fix to me. B revision cards without that stupid freaking BIOS switch - and a new updated design that combines the highest performance with better acoustics. Maybe you disagree. Do you feel updated cards is an unreasonable request? To my knowledge, that's what Ryan Smith told AMD in his feedback to them. Update the design.

Like I said -Clearly, AT is one of MANY MANY websites that have expressed disinterest in testing uber mode - while it's great for them to clarify their policy on this, why can't AMD fix the actual, uhm, problem? That isn't an unreasonable proposition.

Not a single word you said has a thing to do with AnandTech's testing policy. Please stop derailing the thread.

 

[piesquared]

nm

It's rather fascinating to see Anandtech move the goal posts at such an opportune time for NVidia...

 

[nurturedhate]

Slightly off topic here but it caught my eye.

Was just looking at this review on the gtx580 looking at noise data.
http://www.anandtech.com/show/4008/nvidias-geforce-gtx-580/17

Primarily this graph

Mainly this quote from Ryan "Where the GTX 480 set new records for a single GPU card, the GTX 580 is quieter than the GTX 285, the GTX 470, and even the Radeon HD 5870. In fact it’s only a dB off of the 5850, a card under most circumstances we’d call the epitome of balance between performance and noise."

The 5850 at 56dB is the "epitome of balance between performance and noise" but the 290 at 57.2db is "290 untenable as a purchase". 1.2db is the difference between "yea, we like that" and "oh dear god my ears, they bleed". It does not make sense to me.

Furthermore, this quote about the 57.1db 580. 0.1db difference. "the GTX 580 is not whisper quiet, but at no point in our testing did it ever get “loud”." 0.1db difference between never was "loud" and... well what else is there to say. These reviews were written by the same person. 3 years apart. 0.1db difference.

 

[VulgarDisplay]

Slightly off topic here but it caught my eye.

Was just looking at this review on the gtx580 looking at noise data.
http://www.anandtech.com/show/4008/nvidias-geforce-gtx-580/17

Primarily this graph

Mainly this quote from Ryan "Where the GTX 480 set new records for a single GPU card, the GTX 580 is quieter than the GTX 285, the GTX 470, and even the Radeon HD 5870. In fact it’s only a dB off of the 5850, a card under most circumstances we’d call the epitome of balance between performance and noise."

The 5850 at 56dB is the "epitome of balance between performance and noise" but the 290 at 57.2db is "290 untenable as a purchase". 1.2db is the difference between "yea, we like that" and "oh dear god my ears, they bleed". It does not make sense to me.

Furthermore, this quote about the 57.1db 580. 0.1db difference. "the GTX 580 is not whisper quiet, but at no point in our testing did it ever get “loud”." 0.1db difference between never was "loud" and... well what else is there to say. These reviews were written by the same person. 3 years apart. 0.1db difference.

He probably didn't realize that the gtx580 destroyed his hearing at that time. He's just guessing now.

 

[BallaTheFeared]

We all know different methods to measure were at hand, oh sorry, don't let me stop you.... Carry on, where is my torch!?!?!

 

[krumme]

AT can have different testing procedures than what we want. Thats okey. We can find they are way off and ridicolous, but that can happen. Fine.

But its deception, when testing methology is not specified.

In the long review of the 290, where it was blatently obvious that it was because of the fan mode selected of the 290x, that the results were so close.

I know Ryans review. He will be stating the methology used, and discussing it underway. The results beg for it. Its nearly impossible not to write about it, especially when so many pages is directed to issues that arise from it.

When he does not mention 290x is not run on uber in the 290 review, he is acting from direction of Anand.

Anand come here to go into dialogue. Fine. But a dialogue is based on trust. Its not trust, when review methology is not specified, and critical choises is not made clear to us and the readers.

 

[Z15CAM]

Just review in Umber Mode - Period - Then comment about the Noise.

I'm working on building an Arctic Accelero R9 290X-Xtreme III and don't care about AIB solutions.

Water Block will be my Last Resort.

 

[Erenhardt]

Just noticed that Overclocking is not out of the box aspect of the AMD cards, opposite to nvidia cards, which are overclocking within "out of the box" policy.:thumbsup:
http://www.anandtech.com/show/7503/the-amd-radeon-r9-270x-270-review-feat-asus-his