Spin Off: AT's Testing Methods & Uber Mode

[blackened23]

You can't honestly be suggesting that review sites should refuse to use certain features because they want to send a company a message.

Yes, that is what i'm saying. Having a BIOS switch that compromises the user experience is asinine - this should not be an issue, there should not be a BIOS switch. What there should be is a cooler that can handle quiet acoustics without throttling excessively - with the existing shroud, you pick and choose. This should not be the case and is not acceptable.

I remember that the Titan shroud cooler was quieter at 80% manual fan than the 7970 shroud at 45% manual fan. Which was mind boggling because the Titan shroud was spinning at a much higher RPM. What does this tell you? AMD is stuck in 2009 thinking that their 4 year old shroud is good enough. Sorry, that isn't the case.

 

[Silverforce11]

AMD needs to understand that this is not acceptable - I personally hope that other review websites send this message as well. The fact that the 290 and 290X cards require a 20% drop in performance for a quiet fan profile is downright RETARDED. Come on, are you kidding me? How can anyone defend this crap?

Because the consumer can actually CHOOSE which they want, less performance or more noise.

It's a good idea (more choice is good, no?), just not on a bios switch.. but its a separate issue.

 

[AnandThenMan]

Yes, that is what i'm saying.

:thumbsdown:

That means reviewers can pick and choose what features they deem appropriate. That is not objective AT ALL. Why stop there, refuse to bench certain games because you don't like them. Send a message to the game devs.

 

[blackened23]

Because the consumer can actually CHOOSE which they want, less performance or more noise.

It's a good idea (more choice is good, no?), just not on a bios switch.. but its a separate issue.

Why should this be a choice. AMD just should have put a more versatile cooler on the card which removes this issue entirely, and they could have done that. I cannot believe what i'm reading on this forum. After 500 web reviews with all of them complaining of uber mode noise and silent mode throttling, and countless forum pages with complaints of uber mode noise, here we are with people defending this. This should not be a choice, there should not be a BIOS switch. Again, there should just be a vastly improved shroud that removes this compromise altogether.

:thumbsdown:

That means reviewers can pick and choose what features they deem appropriate. That is not objective AT ALL. Why stop there, refuse to bench certain games because you don't like them. Send a message to the game devs.

Because the Joe Clueless consumer will buy a card based on the numbers they see on the pretty graphs, and let's face it - most people don't read reviews fully, they look at FPS charts. With that being the case they will buy a 290 card thinking they can get that level of performance and adjust the fan to their pleasing. OBVIOUSLY this is not possible with the 290 series. Want a quiet card? Sorry, you just lost 20% performance. Do you realize how ridiculous this is? There has never been a GPU *EVER* that required a 20% drop in performance for quiet operation. Kepler can throttle but even on reference cards I NEVER EVER saw one throttle more than 1-2 bins at maximum load, which is roughly 26mhz. Meanwhile, we have the 290 series cards throttling by 250+ MHz for quiet mode operation. And you're telling me this is acceptable. Are you kidding me, man? I like AMD GPUs but stop defending AMD on this matter and demand a better product as a consumer. The throttling and noise compromise is absolutely not acceptable, AMD needs to understand this.

 

[AnandThenMan]

Why should this be a choice. AMD just should have put a more versatile cooler on the card which removes this issue entirely, and they could have done that. I cannot believe what i'm reading on this forum. After 500 web reviews with all of them complaining of uber mode noise and silent mode throttling, and countless forum pages with complaints of uber mode noise, here we are with people defending this. This should not be a choice, there should not be a BIOS switch. Again, there should just be a vastly improved shroud that removes this compromise altogether.

Think about what you're saying. If reviewers refused to use uber mode, the only way we'd know about the ups and downs of it would be buy the card ourselves!

 

[Silverforce11]

I remember that the Titan shroud cooler was quieter at 80% manual fan than the 7970 shroud at 45% manual fan.

Now you're just exaggerating, because its clearly not: http://www.tomshardware.com/reviews/radeon-r9-290-review-benchmark,3659-18.html

Nv's % is different to AMD, because AMD blower peak RPM is higher.

80% fan speed to keep the 780 boosting its top tier, else it throttles. 47% fan speed to keep the R290 from throttling.

 

[USER8000]

:thumbsdown:

That means reviewers can pick and choose what features they deem appropriate. That is not objective AT ALL. Why stop there, refuse to bench certain games because you don't like them. Send a message to the game devs.

Agreed. An objective review would show ALL data points and let the user decide for themselves. In the case of the R9 290 series,the cards should have been tested in both quiet and uber modes,with links to an acoustic sample(with all recording conditions noted,so people can make their own comparisons too with their own cards).

Moreover,noise is very subjective too,as are things like coil whine,etc. It should be up to the reader to decide.

 

[krumme]

Well put, you should post this directly in the comment section of Ryan's explanation if you haven't done so already.

This has been my main point from first reading his 290 conclusions, why did he not just say something along those lines. "I don't like the acoustics, I think it's heading back in the wrong direction in terms of default fan speed and the resultant noise. But if you can deal with the noise..." Maybe an additional "I'm disappointed AMD did not provide a quiet mode for the 290." Instead we get almost a declaration of war. Is the "uber" thing an extension of this personal battle or is it something else?
.

Well he didnt refrase it that way. Why?

He said it had been some busy weeks

So very short time to make the review
Sensitivity to noise - and the delivery man
A hell bend beliewe in own judgement and interpretation

But i dont think that is enough. Loads of marketing bs from amd and nv surrounding him in short time. Nv probably started a campaing or what?

But he put his ear in their direction. And he took their wording and spin to large degree. And its his job not to do.

I read his piece as a little excuse. As close as you can get. Fine. But the uber weird reviews still stand as the idiotic decision not to include uber mode.

 

[blackened23]

Now you're just exaggerating, because its clearly not: http://www.tomshardware.com/reviews/radeon-r9-290-review-benchmark,3659-18.html

Nv's % is different to AMD, because AMD blower peak RPM is higher.

80% fan speed to keep the 780 boosting its top tier, else it throttles. 47% fan speed to keep the R290 from throttling.

It's not just about DB. I can assure you that anyone that has directly listened to the Titan shroud and the AMD shroud will attest that they sound "different", with the AMD shroud emitting a dust buster type of sound past 45% while the Kepler Titan shroud can be heard spinning, but doesn't have the same acoustical output in terms of noise. It doesn't sound like a dust buster (AMD does) - It doesn't sound as harsh.

 

[Abwx]

Yes, that is what i'm saying. Having a BIOS switch that compromises the user experience is asinine - this should not be an issue, there should not be a BIOS switch. What there should be is a cooler that can handle quiet acoustics without throttling excessively - with the existing shroud, you pick and choose. This should not be the case and is not acceptable.

I remember that the Titan shroud cooler was quieter at 80% manual fan than the 7970 shroud at 45% manual fan. Which was mind boggling because the Titan shroud was spinning at a much higher RPM. What does this tell you? AMD is stuck in 2009 thinking that their 4 year old shroud is good enough. Sorry, that isn't the case.

Whatever , this amount to make noise equalization as a mean
to allow a card to run at much higher turns/min , quite a first ,
why not an turns/min equalization and then one can complain
about the noise , instead we have tests done at quiet mode
and complains about the uber mode higher noise , in short
damned if they do and damned if they dont.

 

[Silverforce11]

Why should this be a choice. AMD just should have put a more versatile cooler on the card which removes this issue entirely, and they could have done that.

If you can't handle the noise, get an aftermarket card? Thats what I will do.

 

[Erenhardt]

GOOD. This needs to be done. I've said it before but noise is only part of the problem, the other MUCH BIGGER problem is throttling that results from that crap cooler.

AMD needs to understand that this is not acceptable - too much noise is not acceptable, and throttling with quiet acoustics is REALLY NOT acceptable. I personally hope that other review websites send this message as well. The fact that the 290 and 290X cards require a 20% drop in performance for a quiet fan profile is downright RETARDED. Come on, are you kidding me? How can anyone defend this crap? AMD should have made a better shroud. Period. I hope more websites make this message loud and clear to AMD that they screwed up, people do care about reference blower quality because not everyone is able to use an aftermarket design.

Good LORD. You people need to stop acting like AMD is the bullied victim here. THIS IS THEIR FAULT. PERIOD. I praised the 7970 back in the day because I thought it was a great GPU but it also DID NOT require an insanely high fan profile for good overclocked performance. I didn't lose 20% performance at 40% fan - I could overclock to 1100 *WITH* 40% fan. The GPU landscape has changed. This is not and should not be acceptable, and only if consumers and review websites hammer that point repeatedly will AMD get the message.

We had this posted by you in this topic a couple of times. We do not discuss cooler quality, not even noise levels.
I have to ask you to stop repeatedly OT already. Maybe you should create new thread "Hawaii reference cooler quality discussion", if you really feel the urge to repeatedly share your feelings about that cooler.

 

[Silverforce11]

It's not just about DB. I can assure you that anyone that has directly listened to the Titan shroud and the AMD shroud will attest that they sound "different", with the AMD shroud emitting a dust buster type of sound past 45% while the Kepler Titan shroud can be heard spinning, but doesn't have the same acoustical output in terms of noise. It doesn't sound like a dust buster (AMD does) - It doesn't sound as harsh.

[I disagree], listen to the videos.

Profanity isn't allowed in the technical forums.
-- stahlhart

 

[ICDP]

GOOD. This needs to be done. I've said it before but noise is only part of the problem, the other MUCH BIGGER problem is throttling that results from that crap cooler.

AMD needs to understand that this is not acceptable - too much noise is not acceptable, and throttling with quiet acoustics is REALLY NOT acceptable. I personally hope that other review websites send this message as well. The fact that the 290 and 290X cards require a 20% drop in performance for a quiet fan profile is downright RETARDED. Come on, are you kidding me? How can anyone defend this crap? AMD should have made a better shroud. Period. I hope more websites make this message loud and clear to AMD that they screwed up, people do care about reference blower quality because not everyone is able to use an aftermarket design.

Good LORD. You people need to stop acting like AMD is the bullied victim here. THIS IS THEIR FAULT. PERIOD. I praised the 7970 back in the day because I thought it was a great GPU but it also DID NOT require an insanely high fan profile for good overclocked performance. I didn't lose 20% performance at 40% fan - I could overclock to 1100 *WITH* 40% fan. The GPU landscape has changed. This is not and should not be acceptable, and only if consumers and review websites hammer that point repeatedly will AMD get the message.

Good grief man give it a rest, we get it you hate the R9 290X fan and AMD are evil/incompetent/cads for not making it better. This thread is not about the crap fan, it's about AT refusing to test what is a default mode of a GPU. Test both and let the users decide what setting they will use dependant upon their needs and preferences. Only testing the quiet setting is doing a disservice to those who watercool, game with Headphones or just don't care about the sound. Test both modes and tell us the advantages/disadvantages in an objective manner. By all means give a subjective conclusion, but don't leave out entire sets of pertinent data because they don't suit your own narrow set of preferences.

It's hypocritical to test Titan with the DP enabled despite the fact it is not an out of the box option. Why is it OK to open the Nvidia CP to turn DP on but it isn't OK to reach out and flick the switch on a GPU that is sitting within arms reach on his open test system? Seems like double standards to me and many others here.

 

[krumme]

This update is probably not aimed at this thread. We do not discuss here his noise sensitivity.

I am all in for harsh and nit-picky review style. Ask for more and we will see better products in the feature. True that there was a lot (maybe even to much) of personal opinion in the reviews, but still, mostly valid criticism.

The issue is consistency. While Hawaii reviews were focused on the downsides of the cards, 780ti review avoided that approach. One could go as far as to say: 780ti review silently passed by every shortcoming of the card.

The policy to test product as it comes "out of the box" is not something I would expect from enthusiast PC parts review - more likely from console/mobile review. But not even that is the issue. The issue is once again consistency. If they refuse to use "uber mode", why do they show Titan DP performance if it is off by default. Why do they show SLI performance if it needs hardware ingratiation mod to make it work? Why in one review we have both, out of the box only for one part and modified for other parts?

I would like to see what nvidia demanded from reviewers, what instructions nv made them to fallow. I think that sharing all the backstage action with readers could let them prepare for what they will see in the review. How to approach the data, and how to form they own opinion based on presented point of view (which seems to be influenced by card manufacturers).

Fully agree here. We want to see the reviewers guides now.

 

[Abwx]

It's not just about DB. I can assure you that anyone that has directly listened to the Titan shroud and the AMD shroud will attest that they sound "different", with the AMD shroud emitting a dust buster type of sound past 45% while the Kepler Titan shroud can be heard spinning, but doesn't have the same acoustical output in terms of noise. It doesn't sound like a dust buster (AMD does) - It doesn't sound as harsh.

Lolz , noise equalization is no more enough , now we have
ponderation of the noise content , it wont be long before
fourier transform is brought to the debate for "accurate"
noise distribution in function of frequency measurements....

 

[blackened23]

Good grief man give it a rest, we get it you hate the R9 290X fan and AMD are evil/incompetent/cads for not making it better. This thread is not about hte crap fan, it's about AT refusing to test what is a default mode of a GPU. Test both and let the users device what they like dependant upon their needs and preferences.

It's hypocritical to test Titan with the DP enabled despite the fact it is not an out of the box option. Why is it OK to open the Nvidia CP to turn DP on but it isn't OK to reach out and flick the switch on a GPU that is sitting within arms reach on hi open test system. Seems like double standards to me and many others here.

I already explained why. People (who read GPU reviews) will see uber mode numbers because the vast majority of folks just look at FPS graphcs without reading the text to understand the context - with that being the case, plenty of people will buy a card thinking they will get uber level performance without realizing that a quiet noise profile will drop performance by 20%. This is the problem. Noise is part of it, but throttling is a much bigger part of it.

Never before in the history of GPUs that i'm aware of, did quiet operation require a 20% drop in performance. Kepler has variance but it sure isn't 20% - in my experience, throttling is fairly minimal and doesn't impact performance. I have personally only seen 2 bin throttling (26mhz) which didn't affect performance in a perceptible manner - meanwhile, 290 loses 20%. How ridiculous is that? Now think of the buyer that purchases a card based on what they thought would perform on the uber level. Only when they make their card "quiet", it performs way lower. That is BS. Sorry.

 

[Silverforce11]

I already explained why. People (who read GPU reviews) will see uber mode numbers because the vast majority of folks just look at FPS graphcs without reading the text to understand the context - with that being the case, plenty of people will buy a card thinking they will get uber level performance without realizing that a quiet noise profile will drop performance by 20%. This is the problem. Noise is part of it, but throttling is a much bigger part of it.

Never before in the history of GPUs that i'm aware of, did quiet operation require a 20% drop in performance. How ridiculous is that? Now think of the buyer that purchases a card based on what they thought would perform on the uber level. Only when they make their card "quiet", it performs way lower. That is BS. Sorry.

If ppl read graphics without reading text, its their fault. duh. As a review site, you need to present all the data.

 

[Teizo]

Think about what you're saying. If reviewers refused to use uber mode, the only way we'd know about the ups and downs of it would be buy the card ourselves!

The most common sense thing would be do the review in regular mode against the stock of the other card. Then, at the end of the review do an additional review with the other feature with the caveat that this is not the default mode and this is why the scores are listed separate and alone, and maybe those scores should be best compared to OC 780 models.

It is pretty obvious that AMD was targeting the 780 in quiet mode and the Titan in Uber Mode...trying to kill two birds with one stone (which is why they tried to trump it up as the Titan Killer). The problem is Nvidia dropped the price of the 780 negating the advantage, so AMD then updated the driver of the 290 to try to put heat on the 780 because in it's regular profile it (the 290) was up against the $329 770 which would have had it over priced.

Basically, Nvidia just out played AMD with this, and even though AMD does have some great great cards at killer price points...they both come with very big drawbacks as the cards were made to operate at their threshold by default rather than have them run cooler with more potential to overclock.

I cannot emphasize enough that AMD fans trying to nitpick AnandTech and Ryan about this is one of the worst cases of sour grapes I've ever seen. Just admit that while AMD has some great cards, they just got outplayed by Nvidia because they did not have the foresight to put a better cooler on their card after deciding to allow it to run at it's peak threshold.

 

[SiliconWars]

This update is probably not aimed at this thread. We do not discuss here his noise sensitivity.

I am all in for harsh and nit-picky review style. Ask for more and we will see better products in the feature. True that there was a lot (maybe even to much) of personal opinion in the reviews, but still, mostly valid criticism.

The issue is consistency.

Snipped the rest of it because this was all that needed saying, perfectly summarised. AMD needs a beat-down for that cooler, but the methods used were inconsistent. "out of the box" is BS that Anandtech doesn't hold anything else they review to - It's only AMD graphics cards that have to follow this new rule, and that alone is the reason for this thread, and the accusations of bias.

 

[Erenhardt]

plenty of people will buy a card thinking they will get uber level performance without realizing that a quiet noise profile will drop performance by 20%. This is the problem. Noise is part of it, but throttling is a much bigger part of it.
How ridiculous is that? Now think of the buyer that purchases a card based on what they thought would perform on the uber level. Only when they make their card "quiet", it performs way lower. That is BS. Sorry.

What will happen when there is not uber mode shown in review and a person buys 290x expecting it to be a quiet card. During installation accidentally flips to uber mode and find out his card is much louder then in the review?
Tell me more about BS and ridiculousnes

 

[Leadbox]

Because the Joe Clueless consumer will buy a card based on the numbers they see on the pretty graphs, and let's face it - most people don't read reviews fully, they look at FPS charts.

I don't think anyone capable of buying and installing a card is as clueless as you're suggesting and if most ppl just looked at fps charts, then AMD should be doing much better than ~38% market share

 

[Nintendesert]

The most common sense thing would be do the review in regular mode against the stock of the other card. Then, at the end of the review do an additional review with the other feature with the caveat that this is not the default mode and this is why the scores are listed separate and alone, and maybe those scores should be best compared to OC 780 models.

It is pretty obvious that AMD was targeting the 780 in quiet mode and the Titan in Uber Mode...trying to kill two birds with one stone (which is why they tried to trump it up as the Titan Killer). The problem is Nvidia dropped the price of the 780 negating the advantage, so AMD then updated the driver of the 290 to try to put heat on the 780 because in it's regular profile it (the 290) was up against the $329 770 which would have had it over priced.

Basically, Nvidia just out played AMD with this, and even though AMD does have some great great cards at killer price points...they both come with very big drawbacks as the cards were made to operate at their threshold by default rather than have them run cooler with more potential to overclock.

I cannot emphasize enough that AMD fans trying to nitpick AnandTech and Ryan about this is one of the worst cases of sour grapes I've ever seen. Just admit that while AMD has some great cards, they just got outplayed by Nvidia because they did not have the foresight to put a better cooler on their card after deciding to allow it to run at it's peak threshold.

Sorry, but Anandtech has already done the debacle of comparing an OC card to a reference card in a major review. We're not going down that road again.

 

[SiliconWars]

Never before in the history of GPUs that i'm aware of, did quiet operation require a 20% drop in performance. Kepler has variance but it sure isn't 20% - in my experience, throttling is fairly minimal and doesn't impact performance. I have personally only seen 2 bin throttling (26mhz) which didn't affect performance in a perceptible manner - meanwhile, 290 loses 20%. How ridiculous is that? Now think of the buyer that purchases a card based on what they thought would perform on the uber level. Only when they make their card "quiet", it performs way lower. That is BS. Sorry.

This is bleeding edge technology. Most people just can't grasp how bleeding edge this is.

You want to put Titan or the 780 Ti to 100% fan speed, go ahead. The increases in performance will be tiny in comparison to the noise generated.

AMD's solution brings real performance benefits for that extra noise. Huge benefits that have never before been seen through simple thermal management. This is an extremely advanced turbo that is being denied the chance to show what it can do because Anandtech already decided it wasn't allowed.

 

[blackened23]

I don't think anyone capable of buying and installing a card is as clueless as you're suggesting and if most ppl just looked at fps charts, then AMD should be doing much better than ~38% market share

Believe me. It happens. Most GPU buyers dont' spend their times on forums like this and just want to know the frames per second. You guys need to demand better products as consumers. A GPU that requires the user to compromise between noise and performance isn't kosher - like I said, there has never been such a GPU, ever that i'm aware of. What GPU required a 20% drop in performance for quiet operation prior to the 290? Name one. Please. This is new territory and it isn't good territory at that.

A consumer demanding a better product would ask for a shroud that didn't require a BIOS switch for maximum performance and good acoustics - the Titan uses about 10 watts less than the 290X yet the acoustic quality is leaps and bounds better. If AMD had decided to put some effort into their shroud engineering, this would be a non issue. We would have uber performance without uber noise. We would also have quiet mode with uber level performance.

Basically it boils down to demanding a better product as a consumer. Apparently, some of you dont' want better products, and are happy that AMD is forcing this compromise. Now, I think the 290 is a heck of a card performance wise but the throttling and noise compromise is complete nonsense. This compromise has NEVER existed on any prior card. IMHO, AMD needs to get this message and if anything they should start work on a "B" revision of cards that has a much more versatile shroud. That would be a better product. That would be a more balanced product. That would completely eliminate the quiet mode throttling and the uber mode noise - this would give consumers a better product, which is what AMD should do.


Ok, its time to move on from this subject and get back on the actual topic.

-Rvenger