Custom R9 290/290X Reviews/Availability/Listings thread

[Silverforce11]

How about 780 Ti VRM temps? MSI Gaming:
http://www.guru3d.com/articles_pages/msi_geforce_gtx_780_ti_gaming_review,9.html

Same test for the ASUS DC2 R290X:
http://www.guru3d.com/articles_pages/asus_radeon_r9_290x_directcuii_oc_review,11.html

How about another custom 780 ti? EVGA SC SuperClocked ACX:
http://www.guru3d.com/articles_pages/evga_geforce_gtx_780_ti_sc_acx_superclock_review,9.html

AND Another custom 780ti? Gigabyte Windforce:
http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_780_ti_windforce_3x_review,9.html

There seems to be a trend actually. Open air design is worse for cooling VRMs than reference blowers. Why? Because the air blown down to cool the VRM is heated by the heatpipe/sink prior, its not blowing cold air like in reference cards with the VRM next to the blower, the first component to receive the cold air, then the GPU itself comes after.

 

[Imouto]

I love how you insert things like 100% efficient, like anything is 100% efficient. Of course you know that.

I don't think that he knows what efficiency means. Efficiency needs an input and an output to be measured and I really don't know what he's talking about.

 

[Skurge]

I guess if you have a crappy case then an open test bench with no active airflow will do better.

 

[f1sherman]

I guess if you have a crappy case then an open test bench with no active airflow will do better.

Hardly.
The only reason why Silverforce may have better temps with closed case is because his GPUs are cooled via combination of fresh air-intake-waterloop, and case air circulation.
(And also because of that air pocket between the cards)

In other words hot air inside his case does not affect his GPUs (and especially cores) as much as it would affect conventionally cooled GPUs in closed case.
Conventionally cooled GPUs, unlike WC-ed setup, are entirely dependent on air circulation inside the case.

 

[Silverforce11]

Gigabyte WF R290X: Such fail, hot and loud, barely better than reference card.. don't know why they bothered.

http://www.computerbase.de/artikel/grafikkarten/2013/gigabyte-radeon-r9-290x-im-test/4/

 

[ICDP]

Anyone looking at the DCII R9 290X VRM temperatures needs to read the figures provided by [H]. Just proclaiming the card as a failure is premature without looking at ALL the numbers.

In performance mode the VRMs are 91c for VRM1 and 78c for VRM2. You should also notice in performance mode the fans were only running at 49% max speed, by comparison quiet mode was 45% max fan speed. That is only 4% higher than in quiet mode yet that brought the temperatures down 17c on VRM1.

From my own experience with DCU II coolers they are still relatively quiet right up to around 70%. That will bring the VRM temps right down IMHO. It really pays to read all the information provided, don't just look at one figure and think OMG that sucks.

 

[Silverforce11]

Anyone looking at the DCII R9 290X VRM temperatures needs to read the figures provided by [H]. Just proclaiming the card as a failure is premature without looking at ALL the numbers.

In performance mode the VRMs are 91c for VRM1 and 78c for VRM2. You should also notice in performance mode the fans were only running at 49% max speed, by comparison quiet mode was 45% max fan speed. That is only 4% higher than in quiet mode yet that brought the temperatures down 17c on VRM1.

From my own experience with DCU II coolers they are still relatively quiet right up to around 70%. That will bring the VRM temps right down IMHO. It really pays to read all the information provided, don't just look at one figure and think OMG that sucks.

That's a good point, I hadn't read into fan speed much, noticed it was still in the quiet range so there's room for manual control to up the fanspeed.

I'm not impressed with any custom R290/X so far though, they need to offer better VRMs, more and better cooling on it to allow extra vcore and high OCs.

 

[3DVagabond]

That's a good point, I hadn't read into fan speed much, noticed it was still in the quiet range so there's room for manual control to up the fanspeed.

I'm not impressed with any custom R290/X so far though, they need to offer better VRMs, more and better cooling on it to allow extra vcore and high OCs.

We should get Matrix, Lightning, and Toxic versions.

 

[ICDP]

That's a good point, I hadn't read into fan speed much, noticed it was still in the quiet range so there's room for manual control to up the fanspeed.

I'm not impressed with any custom R290/X so far though, they need to offer better VRMs, more and better cooling on it to allow extra vcore and high OCs.

I bought an R9 290 that flashed to R9 290X and then added a Acellero Hybrid cooler. Both together cost less than a custom cooled R9 290 (non X). The VRM temperatures under load (20 minute Tomb Raider bench loop) at 1170 core +100mV are 89c for VRM 1 and 68c VRM 2.

 

[Skurge]

Gigabyte WF R290X: Such fail, hot and loud, barely better than reference card.. don't know why they bothered.

http://www.computerbase.de/artikel/grafikkarten/2013/gigabyte-radeon-r9-290x-im-test/4/

I think we should wait for other reviews seeing as how their Asus card performed poorly, but we also have to look at the possibly that Hawaii just does not OC very well unless you get a very good chip.

 

[ICDP]

I think we should wait for other reviews seeing as how their Asus card performed poorly, but we also have to look at the possibly that Hawaii just does not OC very well unless you get a very good chip.

I have tested 3x R9 290X and all got 1150 - 1180 at +100mV. It looks like the average OC on air is going to be similar to Tahiti (1170). My own R9 290X with +156mV reaches 1220, though that kind of volts is for water-cooling IMHO.

At 1170/1400 my R9 290X is ~18% faster than stock in actual games. So this is not a bad OC result at all for a reference design IMHO.

Tomb Raider

Stock ~ 60 FPS
OC 1170/1400 ~ 71 FPS

 

[Silverforce11]

I think we should wait for other reviews seeing as how their Asus card performed poorly, but we also have to look at the possibly that Hawaii just does not OC very well unless you get a very good chip.

All these announced custom cards so far are either reference PCB, slightly tweak reference or a weak custom board (ASUS's) with not much better mosfet setup than reference. Plus, they are open air design which blow warm air down onto the VRM cooling compared to reference which blows cool air onto that area first. It's not a good combination, if these cards already hit 90C VRM temps in gaming load, they will be at around 100 or 105C during mining.. and that would demand a much higher fanspeed, which gets noisy...

See where its heading? Not much better than reference for heavy load: ie. mining or big gaming OCs.

 

[Imouto]

I think we should wait for other reviews seeing as how their Asus card performed poorly, but we also have to look at the possibly that Hawaii just does not OC very well unless you get a very good chip.

I'm convinced these ComputerBase guys are doing something really wrong. Take a look at this screen:



This card is throttling at 84ºC!

And yet again I don't know why they're using the EVGA Precision software with AMD cards.

 

[bigi]

How about 780 Ti VRM temps? MSI Gaming:
http://www.guru3d.com/articles_pages/msi_geforce_gtx_780_ti_gaming_review,9.html

Same test for the ASUS DC2 R290X:
http://www.guru3d.com/articles_pages/asus_radeon_r9_290x_directcuii_oc_review,11.html

How about another custom 780 ti? EVGA SC SuperClocked ACX:
http://www.guru3d.com/articles_pages/evga_geforce_gtx_780_ti_sc_acx_superclock_review,9.html

AND Another custom 780ti? Gigabyte Windforce:
http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_780_ti_windforce_3x_review,9.html

There seems to be a trend actually. Open air design is worse for cooling VRMs than reference blowers. Why? Because the air blown down to cool the VRM is heated by the heatpipe/sink prior, its not blowing cold air like in reference cards with the VRM next to the blower, the first component to receive the cold air, then the GPU itself comes after.

The thermal imager sees Asus metal plate, that is why the temp is so even across its surface (all yellow).

The other cards posted, don't have metal plate and imager measures directly the PCB which is quite different.

 

[blackened23]

And yet again I don't know why they're using the EVGA Precision software with AMD cards.

Precision and Afterburner are essentially the same program written by the same person, Unwinder. He (the developer of both ) posts at Guru3d. Same program, different skin, both work with AMD and nvidia cards. MSI and EVGA pushes builds out at different rates, sometimes EVGA is faster and sometimes MSI is. But they're the same program based on the riva tuner core. Both work and are fully 100% functional with AMD cards.

In fact, anyone that has used both (which I suspect isn't you) would be blind to miss the striking similarities between the functionality of the two programs.

 

[moonbogg]

Wait, so these aftermarket cards suddenly suck because of VRM temps?

 

[Skurge]

I'm convinced these ComputerBase guys are doing something really wrong. Take a look at this screen:



This card is throttling at 84ºC!

And yet again I don't know why they're using the EVGA Precision software with AMD cards.

They say the Temp Target for the Gigabyte card is 85c

 

[Imouto]

Precision and Afterburner are essentially the same program written by the same person, Unwinder. He (the developer of both ) posts at Guru3d. Same program, different skin, both work with AMD and nvidia cards. MSI and EVGA pushes builds out at different rates, sometimes EVGA is faster and sometimes MSI is. But they're the same program based on the riva tuner core. Both work and are fully 100% functional with AMD cards.

In fact, anyone that has used both (which I suspect isn't you) would be blind to miss the striking similarities between the functionality of the two programs.

Precision X isn't pushing as much betas as Afterburner since they're Nvidia only. Would you update your program to support the competition? I don't think so.

Latest example is Afterburner trying to copy ShadowPlay. Why would EVGA even care about that? Why would EVGA update its software to support new GPUs they don't assemble?

 

[Borealis7]

[H] reviewed ASUS 290X and got good results: (i just noticed it's in Quiet mode, but still does not throttle)

 

[Imouto]

[H] reviewed ASUS 290X and got good results: (i just noticed it's in Quiet mode, but still does not throttle)

The new trend is telling people that these cards only have any good performance in open benches.

Proven wrong by the Sapphire 290 Tri-X in the very same site anyway.

 

[Imouto]

They say the Temp Target for the Gigabyte card is 85c

My point is why these cards reach the temp target in a high end case. And why proven good performers like the DCU and Windforce perform so badly while the Tri-X doesn't sweat a bit.

The retarded temp target for the Gigabyte card deserves a place in the Hall of Fame of failures.

 

[f1sherman]

I'm convinced these ComputerBase guys are doing something really wrong.

This card is throttling at 84ºC!

Could it be because... they are using enclosed case. A rarity these days.

+1 for Germans

 

[Lonyo]

Could it be because... they are using enclosed case. A rarity these days.

+1 for Germans

Um... no?
The temperature it throttles at has nothing to do with the case.
When it reaches the throttle temperature is determined in part by the case, but if the throttle temperature is set at 84c, it will throttle at 84c. Case is irrelevant... The card can be set to throttle at 95c, in which case in a closed case it might not throttle.

Case has nothing to do with what temperature it does throttle at, only whether it reaches the throttle temperature or not.

 

[f1sherman]

I'm sorry; what I meant is... of course it's going to throttle.

In enclosed case with 85C temp target.

 

[gorobei]

anand's review of sapphire 290 Tri-x is up. very impressive cooler on reference pcb. $450 msrp. quieter than gtx780. cooler temps help in reducing power draw.

http://www.anandtech.com/show/7601/sapphire-radeon-r9-290-review-our-first-custom-cooled-290