Retail Radeon R9 290X cards may be slower than press samples

[mars]

what?!?! 4% performance variation in silent mode? someone call the national guard!


all overclockers who run water, would build their system for 24/7 use, i wouldnt understand how someone setup their watercooling for the purpose of using the computer "not 24/7"... as it would be exactly the same.


and wtf buys a flagship high end card and runs it in silent mode?

[Sparky]

Anecdotally people I have talked to often don't like to leave their watercooled system on when they're not there, which is what I meant for 24/7 use (as in not shut down multiple times a day) because they worry about a leak happening while they aren't there to see it/take care of it. But, I guess it depends on the person, saying "most" is probably a little stronger than the truth.

Yes, it depends on the person with the loop. However I get the impression you are insinuating that I am stretching the truth or exaggerating from the comment. I am not.

What a surprise, coming from AMD... /sarcasm. Not that nvidia's perfect either but they don't pull shenanigans like this, framepacing, incorrect claims of when major driver releases are coming, etc. with regularity like AMD does.

nvidia's hands are far from clean. They've pulled plenty of their own stunts. However, nvidia isn't even the topic of this thread, so not sure why you're bringing them into it.

[vario]

Hardware.fr has updated the article with the same fan RPM :



Difference is lowered to 4% between press sample and retail one in quiet mode.

http://www.hardware.fr/articles/910-...riabilite.html

Thats no surprise, in uber mode it gets all the cooling it needs so results do not vary that much, when they manually increase fan to PROPER value it jumps up alot, now if they woud flash them to a review sample bios and reapply paste/cooling properly they would get to a almost same result + or - .

Goldentiger, im not picking a fight here , but , you are of course taking this thing to a bad place.Even when proper tests prove that there is no difference or difference is small.
Now, i understand that bad frame pacing was bad for crossfire (only) but i dont know if you followed it properly, after AMD started to fix their problem, it didnt lower performance much or at all, so it never was a dirty way to get numbers, read a in depth article on anandtech for example on that.Also, you do remember that throttling issues plagued a lot od nvidia cards too ? And they didnt have to put up with such a crappy cooler on this heat density.Nvidia has also some SLI issues.World is never black and white you know.

Im not gonna bore you guys anymore with the same things ;-) .It should be quite apparent to everybody what issues 290`s have and why.And if somebody is being adamant that its AMD`s dirty tricks , then well, nobody can change theyre mind at this point.

[BababooeyHTJ]

What a surprise, coming from AMD... /sarcasm. Not that nvidia's perfect either but they don't pull shenanigans like this, framepacing, incorrect claims of when major driver releases are coming, etc. with regularity like AMD does.

What are you talking about? Were you not around when Nvidia first released GPU Boost 2.0? There were driver updates for that too. Most people had to flash the bios on their Titan to prevent throttling. Of course you turned a blind eye to that.

At least this can be fixed with a driver update or by flipping a switch, or by using CCC.

Who is really worried about their loop springing a leak? Done correctly that shouldn't be an issue. I never even look at my loop after that first week or two.

[Hornet331]

Who is really worried about their loop springing a leak? Done correctly that shouldn't be an issue. I never even look at my loop after that first week or two.

Yeah after owing a water cooled pc for 4 years now i hardly look at it after I installed new parts in the loop. Just keep watching for the first day and its fine.

[Ace123]

Yeah after owing a water cooled pc for 4 years now i hardly look at it after I installed new parts in the loop. Just keep watching for the first day and its fine.

Watercooling for me has become such second nature that I dont even think twice after I install a new part. I drain the loop, then I tear out the part in question which is almost always a video card or video cards. I install the new waterblock deftly and install the card quickly into the system and fill it up with deionized water. I hold my hand under the card for a minute or so, if it doesnt leak, somehow I know it will not, and I go about agressively overclocking the card

[mars]

Watercooling for me has become such second nature that I dont even think twice after I install a new part. I drain the loop, then I tear out the part in question which is almost always a video card or video cards. I install the new waterblock deftly and install the card quickly into the system and fill it up with deionized water. I hold my hand under the card for a minute or so, if it doesnt leak, somehow I know it will not, and I go about agressively overclocking the card

if its the video card or just 1 component I don't even bother draining (this isnt good practice, i should be draining it), thank god for quick release fittings, makes it much easier to add extra blocks etc

[GoldenTiger]

Yes, it depends on the person with the loop. However I get the impression you are insinuating that I am stretching the truth or exaggerating from the comment. I am not.



nvidia's hands are far from clean. They've pulled plenty of their own stunts. However, nvidia isn't even the topic of this thread, so not sure why you're bringing them into it.

Er, no, I actually was referring to my own statement that "most don't run it 24/7" being inaccurate.

I already said nvidia's hands aren't clean either in the bold text you quoted , so I'm not clear on why your tone makes it sound like you think you're writing a correction when you're actually just agreeing. I brought that line up in the middle of the AMD discussion that I was writing because I was trying to pre-empt trolls going "but but but nvidia has done stuff too!", which clearly was smart since it's already now being brought up by you .

What are you talking about? Were you not around when Nvidia first released GPU Boost 2.0? There were driver updates for that too. Most people had to flash the bios on their Titan to prevent throttling. Of course you turned a blind eye to that.

At least this can be fixed with a driver update or by flipping a switch, or by using CCC.

See above. I already said nvidia isn't perfect either which you even quoted!... but their problems aren't on the same scale as AMD's tend to be. How is AMD going to fix an obscenely hot chip with inadequate cooler that barely keeps it under control (and often doesn't, resulting in severe throttling) without turning their leafblower into a wind turbine, with a driver update exactly?

[mdzcpa]

Been running water cooled 24/7 for years. If a loop leaks, it's going to do it within day 1. If you can't build a water system that doesn't leak, or you are fearful of leaking, put down the hose and barbs and just step away. I couldn't imagine having a system not designed for 24/7.

The 290X under water is going to be a killer piece of hardware for the money. It's hard to sneeze at $150 less per card. Once both the 290X and 780i are under water, much of the 780i advantage goes out the window. The 780i is a sweet card though. I'm on the fence....

[AliG]

Been running water cooled 24/7 for years. If a loop leaks, it's going to do it within day 1. If you can't build a water system that doesn't leak, or you are fearful of leaking, put down the hose and barbs and just step away. I couldn't imagine having a system not designed for 24/7.

The 290X under water is going to be a killer piece of hardware for the money. It's hard to sneeze at $150 less per card. Once both the 290X and 780i are under water, much of the 780i advantage goes out the window. The 780i is a sweet card though. I'm on the fence....

If I were you, I'd just get a 290, throw it under water (or LN2 if you can deal with constantly replenishing your tube lol) and overvolt + overclock the bejesus out of it. Then once it dies, buy a Maxwell whenever the hell they come out in the 2014.

Pricey? Yes. Extreme? TO THE MAX.

[Heinz68]

Interesting read

AMD?s Radeon R9 290 has a problem, but Nvidia?s smear attack is heavy-handed

Anyway there was not any Cherry picking binned golden chips as some people here insist on. All is about max fan speed preset, easy to fix with driver.

EDIT
I should add not only NVIDIA smear attack is heavy handed, same goes for few nvidia super enthusiast, thanks god not all.

[PedantOne]

Did anyone actually try the slower retail cards with the same, newer bios version which was used on the press sample cards?

no BFU will update bios in his card, especially with no official bios on manufacturers webpage

[Sparky]

Er, no, I actually was referring to my own statement that "most don't run it 24/7" being inaccurate.

I already said nvidia's hands aren't clean either in the bold text you quoted , so I'm not clear on why your tone makes it sound like you think you're writing a correction when you're actually just agreeing. I brought that line up in the middle of the AMD discussion that I was writing because I was trying to pre-empt trolls going "but but but nvidia has done stuff too!", which clearly was smart since it's already now being brought up by you .

Allow me to clarify, as I think there was a little communication issue.

1) I used the term "most people I know," so when you stated "saying most is probably a little stronger than the truth" it gave me the impression that you were calling into question the truthfulness of what I said. If I'm wrong, sorry, but that is how it came across.

2) You made it sound like AMD regularly tries to pull fast ones or always has lots of issues, but nvidia, while they have had some, do not usually. I just think your memory is short, or biased. Fermi was a flamethrower, nvidia mobile 7000 series cooked themselves to death regularly, the infamous wood screws lol, etc. I'm just saying the playing field on both of them is a lot more even as far as goofups and successes than you made it sound like. Again, if this is a misunderstanding, sorry, but that is how it sounded.

3) The only reason I mentioned nvidia in my post was because you did in the first place. I had no reason to otherwise and until you mentioned them others weren't either (unless I missed someone earlier).

Hopefully this clears it up a bit.

[FlanK3r]

Do you know whats TLB bug?...I think this story about "lower" performance of retails card R9-290 is the same. Only inflated bubble and boulevard story. And few people here are happy for it.

[Lanek]

Been running water cooled 24/7 for years. If a loop leaks, it's going to do it within day 1. If you can't build a water system that doesn't leak, or you are fearful of leaking, put down the hose and barbs and just step away. I couldn't imagine having a system not designed for 24/7.

The 290X under water is going to be a killer piece of hardware for the money. It's hard to sneeze at $150 less per card. Once both the 290X and 780i are under water, much of the 780i advantage goes out the window. The 780i is a sweet card though. I'm on the fence....

Clearly, my h2o system is running since the 7970 was out ( i let you count the time ), i remember have unbuild it when i have change my 360 radiator for a more powerfull ( this have allow me to remove my second 360 rad ( HW stealth ).... i just change water and just reput some additive time to time ( i have white tubing, so if you want verify it is easy, i can take a shoot, because after all this time, they are now yellow ( my bad addict to smoke ).. I just use connectors and tubing, i dont even use fixation on the tubing .. My system have never leak in nearly 2 years, i dont see why i could not let it run 24/7... this said, when im at my office, i have no reason to let it run in idle.. ( ok this do surely 11-12years i use Only watercooling ( never use air cooling since )

[spursindonesia]

This whole brouhaha is only because of fans variability + PWM fan control, the card instructed 47% fan speed, but because of fan quality variability (such as bearing, weight, etc) the fan actually spun less than that, resulting less thermal capacity of the cooler, hence much lower performance than reviewed. Now they put a firm RPM number limit on the driver & enforcing the optimal HSF's thermal capacity, BAM, problem's solved, no conspiracy or ill intention whatsoever.

[tajoh111]

This whole brouhaha is only because of fans variability + PWM fan control, the card instructed 47% fan speed, but because of fan quality variability (such as bearing, weight, etc) the fan actually spun less than that, resulting less thermal capacity of the cooler, hence much lower performance than reviewed. Now they put a firm RPM number limit on the driver & enforcing the optimal HSF's thermal capacity, BAM, problem's solved, no conspiracy or ill intention whatsoever.

Would people respond the same way, if nvidia had done the same thing and had the same excuse?

[mars]

Would people respond the same way, if nvidia had done the same thing and had the same excuse?

people will probably respond the same way.... considering AMD has done NOTHING in this instance except go cheap on a stock cooler.... jesus christ

[thatdude90210]

Would people respond the same way, if nvidia had done the same thing and had the same excuse?

Yes people would, just different people.

[Sparky]

Would people respond the same way, if nvidia had done the same thing and had the same excuse?

Same responses, just from different people

I think we've beaten the "stock cooler sucks" horse enough already. We all know it does.

[hennyo]

One thing that has me wondering on this card is "why" is it soo hot, is it strictly a graphical core unit issue, or is there something wrong with the power circuitry like the GTX 480 had.

Something many people do not know about the GTX 480, is that if you add active cooling to it, and get the core temps down under around 60 Fahrenheit you can cut around 50 to 100 watts off the load power usage at stock settings. I am curious if a similar thing happens with this chip considering how hot it runs.

[dasa]

better cooling can drop power use

here a better hsf on the 290 drops it from 290w to 260w max power use
http://ht4u.net/reviews/2013/r9_290_...est/index7.php

[vario]

Pretty much what could be expected, better HSF makes it quieter, better clockin and lowes power usage, all the bad points get a bit less bad.

And tajoh, even now, u still cant drop this,you are using word "EXCUSE" ,as if it was just a way to cover bad prior intentions by AMD...


Still, WHERE ARE THE UNDERVOLTING results ? Nobody with these cards here ?

[GoldenTiger]

Allow me to clarify, as I think there was a little communication issue.

1) I used the term "most people I know," so when you stated "saying most is probably a little stronger than the truth" it gave me the impression that you were calling into question the truthfulness of what I said. If I'm wrong, sorry, but that is how it came across.

2) You made it sound like AMD regularly tries to pull fast ones or always has lots of issues, but nvidia, while they have had some, do not usually. I just think your memory is short, or biased. Fermi was a flamethrower, nvidia mobile 7000 series cooked themselves to death regularly, the infamous wood screws lol, etc. I'm just saying the playing field on both of them is a lot more even as far as goofups and successes than you made it sound like. Again, if this is a misunderstanding, sorry, but that is how it sounded.

3) The only reason I mentioned nvidia in my post was because you did in the first place. I had no reason to otherwise and until you mentioned them others weren't either (unless I missed someone earlier).

Hopefully this clears it up a bit.

Hey, I know this is a little late a reply but, I wanted to reply anyway .

1) Sorry, I did mean that in regards to my statement, rather than yours, but it's very ambiguous looking back at it. I don't blame you for taking it that way.

2) What I was trying to say is that I feel like nvidia's slipups tend to be less severe for the most part, at least that's my opinion. Both definitely have had their issues.

3) It seems I outsmarted myself by trying to head off a possible discussion path rather than just seeing if it happened and addressing it if it did with my thoughts :P.

Regardless, I apologize for my overall rude tone there. I try to remain generally professional when I post but I absolutely slipped up there when replying to this topic.

[Ace123]

Hey, I know this is a little late a reply but, I wanted to reply anyway .

1) Sorry, I did mean that in regards to my statement, rather than yours, but it's very ambiguous looking back at it. I don't blame you for taking it that way.

2) What I was trying to say is that I feel like nvidia's slipups tend to be less severe for the most part, at least that's my opinion. Both definitely have had their issues.

3) It seems I outsmarted myself by trying to head off a possible discussion path rather than just seeing if it happened and addressing it if it did with my thoughts :P.

Regardless, I apologize for my overall rude tone there. I try to remain generally professional when I post but I absolutely slipped up there when replying to this topic.

This is the kind of attitude that will help keep out community strong. Kudos to you.