Asus R9 290x DCII (Guru3D Review)

[moonbogg]

http://www.guru3d.com/articles_pages/asus_radeon_r9_290x_directcuii_oc_review,1.html

I'm sure you've seen it, but this looks really nice. When the 290's like these come out I'm going to have a really hard time not pulling the trigger on those sexy beasts.

 

[Termie]

Discussed somewhat in this thread: http://forums.anandtech.com/showthread.php?t=2355359

Also, check out Hardware Canuck's review of the same card:

http://www.hardwarecanucks.com/foru...64694-asus-r9-290x-directcu-ii-oc-review.html

I'm somewhat surprised they are reviewing these cards before they're actually on the market. That can always lead to changes after the card is reviewed, but I think there's enough pressure on AMD to release these that it (or its partners) are making an exception here.

 

[Aikouka]

17C difference between this and the retail cooler at load, and the ASUS card doesn't throttle.

 

[Lonyo]

17C difference between this and the retail cooler at load, and the ASUS card doesn't throttle.

Comparing temperatures is stupid and pointless.
When you have a card designed to operate at a specific temperature at the lowest noise level it can, comparing temperatures is stupid.

The R9 290X reference runs at 95c because it's been told to run at 95c.
The comparison reviews should make is what noise level does a card make while running at X temperature.

The R9 290X with stock cooler could run cooler, if you turn up the fan, but it's designed not to. Any HS/F setup could make the card run cooler if it was told to.

 

[nitromullet]

Comparing temperatures is stupid and pointless.
When you have a card designed to operate at a specific temperature at the lowest noise level it can, comparing temperatures is stupid.

The R9 290X reference runs at 95c because it's been told to run at 95c.
The comparison reviews should make is what noise level does a card make while running at X temperature.

The R9 290X with stock cooler could run cooler, if you turn up the fan, but it's designed not to. Any HS/F setup could make the card run cooler if it was told to.

I think the bottom line, no matter how you compare it, is that the Asus cooler is better than the reference AMD cooler.

 

[krumme]

Comparing temperatures is stupid and pointless.
When you have a card designed to operate at a specific temperature at the lowest noise level it can, comparing temperatures is stupid.

The R9 290X reference runs at 95c because it's been told to run at 95c.
The comparison reviews should make is what noise level does a card make while running at X temperature.

The R9 290X with stock cooler could run cooler, if you turn up the fan, but it's designed not to. Any HS/F setup could make the card run cooler if it was told to.

I agree. If you get the asus card run it at low noise setting. Then you use the new technology that can handle 95c and what the cards are build for. As always excellent card from asus especially regarding low noise. They keep getting better at it. Damn impressive.

 

[thilanliyan]

Comparing temperatures is stupid and pointless.
When you have a card designed to operate at a specific temperature at the lowest noise level it can, comparing temperatures is stupid.

The R9 290X reference runs at 95c because it's been told to run at 95c.
The comparison reviews should make is what noise level does a card make while running at X temperature.

The R9 290X with stock cooler could run cooler, if you turn up the fan, but it's designed not to. Any HS/F setup could make the card run cooler if it was told to.

I think the bigger issue is the throttling with the reference card due to the higher temps. With the lower temps of the non-ref card, it can keep clocks up.

 

[moonbogg]

This Asus card only has one noise setting: NONE!

 

[Attic]

Now you just know that I can throw all kinds of superlatives at you guys, but I can be brief as well. This is the Radeon R9-290X we all expected it to be, period.

Amen to that.

And not to beat a dead horse,…

but,

Amen to that!

 

[krumme]

I think the bigger issue is the throttling with the reference card due to the higher temps. With the lower temps of the non-ref card, it can keep clocks up.

Nonsense. Its only some reviewers thst have throttling cards. Most users seem to have the ability to push the small bottom or move a slider. But at the expense of noise. The reference cards is very noisy to some but defenetively not throttling in real users computers right now.

I would never personally buy a ref card because of my noise preferences. But the throttling is overblown bs. Creating problem that nobody will ever get. Spin.

 

[krumme]

This Asus card only has one noise setting: NONE!

I unplug my gfx entirely. And turn off all the 500rpm fans in my case. Thats my low noise setting. 17db is noise in my world. Lol.

 

[thilanliyan]

Nonsense. Its only some reviewers thst have throttling cards. Most users seem to have the ability to push the small bottom or move a slider. But at the expense of noise. The reference cards is very noisy to some but defenetively not throttling in real users computers right now.

Sorry, I meant they throttle at reasonable noise levels. I have 3 290s...and they can get very loud at above 60% fan speed. My idea of noise may be different to others though as I water cool most of the time. The noise above 60% fan speed on my 290s is unacceptable for me. Thankfully the cards are sitting outside where I don't have to hear them.

 

[Stuka87]

Nonsense. Its only some reviewers thst have throttling cards. Most users seem to have the ability to push the small bottom or move a slider. But at the expense of noise. The reference cards is very noisy to some but defenetively not throttling in real users computers right now.

I would never personally buy a ref card because of my noise preferences. But the throttling is overblown bs. Creating problem that nobody will ever get. Spin.

Why are you trying to defend the reference card? They DO throttle. If the fan reaches its max allowed speed, and the temp hits 95C, the card throttles and performance drops off. There is no arguing this. AMD tried to spin that they wanted it to be this way, but ultimately it was a cost cutting move to use the cheap coolers from the 79x0 cards.

This card shows that it doesn't take much work to make it perform like it should have 1.5 months ago.

 

[hjalti8]

I think the bottom line, no matter how you compare it, is that the Asus cooler is better than the reference AMD cooler.

The Asus card has horrible VRM temps, significantly higher than reference cards. Poor job ASUS :thumbsdown:

 

[blackened23]

Nonsense. Its only some reviewers thst have throttling cards. Most users seem to have the ability to push the small bottom or move a slider. But at the expense of noise. The reference cards is very noisy to some but defenetively not throttling in real users computers right now.

I would never personally buy a ref card because of my noise preferences. But the throttling is overblown bs. Creating problem that nobody will ever get. Spin.

LOL. Say what now. What did I just read. "Nonsense" you say - but you would never personally buy reference. You don't like noise but I thought the noise issue was all spin? I sense some cognitive dissonance going on here. The statement that 100% of ref 290X users aren't getting throttling is outright nonsense as well. Even in uber mode, with crossfire "sandwiched" cards you have to up the fan threshold past 55% in overdrive settings to prevent throttling in most chassis. And 100% of users aren't using uber mode even with single GPU configurations, and i'm sure you're aware of the throttling issues there.

That said, I do like the asus card. It fixes all the issues with the terrible reference design, and performs great. Maybe it will inspire AMD to up the ante with a revision B blower reference design, which would be more ideal for mGPU users.

 

[monstercameron]

it is funny because some biased individuals on this forum have just been spanked...this custom card ties with the 780Ti, loses in nvidia sponsored but wins in amd sponsored and costs less...this certain poster^ should now stop posting BS and start facing facts...

Warning issued for callout.
-- stahlhart

 

[Silverforce11]

Reference cards will throttle at their default fan, even uber mode 55% isn't enough for single cards in some games and thats shown by reviewers who test in ideal situations, open bench or high airflow cases. It's just a horrible cooler and I would never buy it without the intention of going water cooling.

We all knew these aftermarket cards will fix the issues tho, so no real surprises.

And it does trade blows very well versus the much more expensive 780 ti, wins some, lose some. If SKYMTL is correct and he is able to squeeze 1.3ghz on it, then its an awesome card, despite the VRM temps (asus uses different VRM components than reference).

 

[SolMiester]

it is funny because some biased individuals on this forum have just been spanked...this custom card ties with the 780Ti, loses in nvidia sponsored but wins in amd sponsored and costs less...this certain poster^ should now stop posting BS and start facing facts...

Are you doing a member call out on false accusations?, because a certain poster has spouted BS at!
Oh, and seriously, do you understand what spanked means..

 

[blackened23]

it is funny because some biased individuals on this forum have just been spanked...this custom card ties with the 780Ti, loses in nvidia sponsored but wins in amd sponsored and costs less...this certain poster^ should now stop posting BS and start facing facts...

Facing facts? The 780ti is still faster especially when it's overclocked and even at stock - aftermarket 780ti's are faster.

780TI OVERCLOCKED :

DC II 290X OVERCLOCKED :

Let's not get into that though. I think the 290X DC II is a great card. It isn't faster or better than the best custom 780ti cards, but it is a better value if you can find it at MRSP. Don't misunderstand me here because I really like the 290X DC II - it is everything the reference SHOULD have been.

Now you can stop with the attacks and actually do a proper rebuttal to my points, which were all factual. Even the HardOCP crossfire 290X tests, done on an open test bench, Brent stated that the fan cap had to be raised past 55% in overdrive to maintain clocks. This is an open test bench mind you. And i'm sure you're 100% aware that silent mode will throttle.

Now this is a tangent and I don't feel like going down this road, because the DC II 290X is a great card. But you're the one being rude with personal attacks, and everything I stated is factual. Therefore, next time you can just think for a second before you hit post.

 

[Enigmoid]

Looks nice. Temperatures are good. Surprised by the small OC though.

 

[SPBHM]

the new cooler transformed the card, silent, run at 1GHz all the time, and OC to 1.2GHz... there's hardly anything negative worth saying about it, great card, hopefully many other custom R9 290s are going to be available soon,

 

[Silverforce11]

Facing facts? The 780ti is still faster especially when it's overclocked and even at stock - aftermarket 780ti's are faster.

IF your measuring stick is Crysis 3, then sure. Overall, its gonna be a lot closer to the 780ti than to the 780.

 

[blackened23]

IF your measuring stick is Crysis 3, then sure.

More like....anything. You can view benchmarks of the overclocked GHz edition 780ti Gigabyte card at hardwarecanucks in which it is faster than the 690 and the DC II 290X (based on extrapolation) even in AMD titles such as hitman absolution. I should point out that Hitman generally favors AMD hardware by a large margin as well.

http://www.hardwarecanucks.com/foru...gigabyte-gtx-780-ti-ghz-edition-review-8.html

This wasn't even the point of my post, though, read it in the entirety. It was more about a response to monstercameron who suggested I "face facts".

The asus DC II 290X is a great card. I don't disagree with you in that respect - it fixes all of the issues that were present with the reference design, and it's fast, therefore I cannot say anything negative about it. It's a great little card. It is also a better value (at MSRP) than the 780ti.

 

[monstercameron]

To be clear, not a member callout but calling out a certain few that pollute threads with charts and walls of text, now that a proper R9 card is out and when the comparisons come in from various sources all their claims of superiority will be even less than valid.

aside a 50mhz core and 100mhz gpu clock bump combined with a new cooler gave the 290x such a boost, so much so as to surpass the 780ti in crysis 3, what have the watercooling guys been seeing in terms of performance?

To be clear, that WAS a member callout.
-- stahlhart

 

[Imouto]

Been said in other thread as well but the difference between reference and custom is ~10% when OCed.

Nvidia loyalists are preemptive about Mantle. AMD is at a ridiculous short distance of Nvidia, if BF4 delivers with Mantle things are going to get really ugly. Even a mere 20% performance boost can start a trend in the whole lineup for both brands.