[H] XFX 290x DD Crossfire vs 780ti SLI Review

[wand3r3r]

[H] pointed out that the GTX 780TI cards throttled back to their core clock speed and stayed there during the tests. Brent makes it clear why this is happening in the post below.

http://hardforum.com/showpost.php?p=1040581044&postcount=49

GTX780 (and Ti) are set at 80c target temperatures and in these tests they were hitting 83c.
http://www.hardocp.com/article/2014...tion_edition_crossfire_review/10#.UuoKl_RdWq0

We all know R9 290/X has a much higher target temp of 95c so the cards in the test don't throttle as they get nowhere near that temperature.

Raising the target temperature in the 780Ti SLI would reduce the deficit in performance.

Interesting. The performance deficit was pretty large though, I don't know if it could be made up with that.

 

[Silverforce11]

[H] pointed out that the GTX 780TI cards throttled back to their core clock speed and stayed there during the tests. Brent makes it clear why this is happening in the post below.

http://hardforum.com/showpost.php?p=1040581044&postcount=49

GTX780 (and Ti) are set at 80c target temperatures and in these tests they were hitting 83c.
http://www.hardocp.com/article/2014...tion_edition_crossfire_review/10#.UuoKl_RdWq0

We all know R9 290/X has a much higher target temp of 95c so the cards in the test don't throttle as they get nowhere near that temperature.

Raising the target temperature in the 780Ti SLI would reduce the deficit in performance.

Toms also posted that they had to set their 780s at 80% fan speed to maintain its normal to tier boost, so their data for its performance is from manual fan. [H] did the right thing though, if its a test with default cards then leave it default, if its an OC bench then sure, go max fan who cares.

 

[3DVagabond]

Toms also posted that they had to set their 780s at 80% fan speed to maintain its normal to tier boost, so their data for its performance is from manual fan. [H] did the right thing though, if its a test with default cards then leave it default, if its an OC bench then sure, go max fan who cares.

So did hardware.fr. Not surprising really that custom cooled Hawaii cards are beating them.

We need to see a review comparing custom vs. custom since neither company's reference coolers can handle these cards. I think it's a real disservice for more sites to not make this better known. Especially the ones who simply run short benchmarks. The numbers they show aren't representative of real world performance. Although it has been pointed out on these boards before and people denied it. Where were the big expose' reviews? :\

 

[guskline]

The XFX DD 290X is $699 at Newegg. Compare it to a EVGA GTX780TI SC at $729 and see if the results are different. From what I read in this article they pitted one of the top end custom cooled R9 290x cards against a stock 780TI. I agree if the MSRP of the 290X was $569 it wins hands down but at the present price I don't know because a custom cooled 780TI is about the same.

 

[wand3r3r]

The XFX DD 290X is $699 at Newegg. Compare it to a EVGA GTX780TI SC at $729 and see if the results are different. From what I read in this article they pitted one of the top end custom cooled R9 290x cards against a stock 780TI. I agree if the MSRP of the 290X was $569 it wins hands down but at the present price I don't know because a custom cooled 780TI is about the same.

Yeah, that's true.

 

[Spidre]

So did hardware.fr. Not surprising really that custom cooled Hawaii cards are beating them.

We need to see a review comparing custom vs. custom since neither company's reference coolers can handle these cards. I think it's a real disservice for more sites to not make this better known. Especially the ones who simply run short benchmarks. The numbers they show aren't representative of real world performance. Although it has been pointed out on these boards before and people denied it. Where were the big expose' reviews? :\

The 780ti stock cooler definitely can, they just need a custom fan curve. I have mine set up to not go above 70%, and it prevents throttling with an overclock during extended games.

This was, however, a stock comparison so I guess keeping everything on default was what they were going for.

 

[tweakboy]

WTF is 5760. How much you pay for those cards... thx for reviews..