Sapphire Radeon HD 4890 Vapor-X 2GB review

Wow I cant believe it! A 4890 is beating a 285!

And that's with ATI drivers 4 months old which was a serious shame imo, one month or even 2 i could forgive but 4? That's just bad form

I'm glad you pointed that out Gremlin, I can't help but get really annoyed at reviews (not just BTs) when they use out of drivers. Enjoyable read though.

:D

Originally Posted by Jack_Pepsi
I'm glad you pointed that out Gremlin, I can't help but get really annoyed at reviews (not just BTs) when they use out of drivers. Enjoyable read though.

:D

Thanks for the comments. PLEASE CHECK YOUR PMs!!!

WTF is happening with the multi-GPU solutions on COD5? On the higher resolutions, they all go down HARD on the minimum framerates, dropping under 25, which is definately unplayable! Yikes. the single-cards perform much better :?

Originally Posted by Xtrafresh
WTF is happening with the multi-GPU solutions on COD5? On the higher resolutions, they all go down HARD on the minimum framerates, dropping under 25, which is definately unplayable! Yikes. the single-cards perform much better :?

Yep, not smooth.

Hmm, lots of connectivity here, but there still should be no reason to even include D_sub.

Still an awesome card, and I would get one if it wasn't for the fact that DX11 cards are literally coming around the corner, and I need to save up for a LCD2080UXi.

Nice review, well written, but on relevance EPIC FAIL.

Pull your fingers out of your behinds and update your damn drivers. Cheap-ass websites will end up with cheap-ass readers.

Originally Posted by Action_Parsnip
Pull your fingers out of your behinds and update your damn drivers. Cheap-ass websites will end up with cheap-ass readers.

Actually, I haven't seen much "rabbit-out-of-hat" drivers from ATI for quite a while. That said I do agree, although in this case it might've been a bit harsh.

Originally Posted by Bindibadgi
Yep, not smooth.

Is the situation fixed with newer drivers considering the age of those tested and the big gains since then? Cat 9.8 had HUGE gains for multi gpu setups iirc

I too had noticed drivers are somewhat out of date when these tests are performed!

If you look at the gainward gtx295 single pcb review you will see what I mean. (comments section)

Andy

Well spotted on the drivers guys.

Is there a reason why the 9.4 ATi drivers were used instead of this months 9.8 or the 9.7 if you tested these cards before the 15th August?

good review and a great card that said wouldn't a sapphire 4890 toxic with vapor-x be a better choice with manufactured clock speeds
of 960 that is under warranty?

Why Isn't there a picture, or description of the bundled goodies??

Originally Posted by TTmodder
Why Isn't there a picture, or description of the bundled goodies??

1st page, 4th paragraph, last sentence ;)

Hmmm, 4-6wks from now you will be able to buy a 5850/5870 DX11 class card, both of which should be quicker than this, and for the same amount of money, in the 5850's case anyway......

Silly to buy a new CPU or GPU at the minute......

Originally Posted by Elton

Quote:

Originally Posted by Action_Parsnip
Pull your fingers out of your behinds and update your damn drivers. Cheap-ass websites will end up with cheap-ass readers.

Actually, I haven't seen much "rabbit-out-of-hat" drivers from ATI for quite a while. That said I do agree, although in this case it might've been a bit harsh.

Its got nothing to do with rabbits. Its good practice to keep forceware and catalyst up to date. 4 revisions out of date is just epic fail. Not even xbitlabs or legionhardware pull this kinda crap. Fail.

The Nvidia 185.85 drivers better than any of the 190 series drivers I've used especially in Crysis.

Would of been good if they used the Nvidia 185.85 drivers vs ATI/AMD 9.8 Cats.

Originally Posted by Combatus

Quote:

Originally Posted by TTmodder
Why Isn't there a picture, or description of the bundled goodies??

1st page, 4th paragraph, last sentence ;)

Whoops, guess I'm used to seeing pictures of the goodies :P

cf results?

Originally Posted by Nature
cf results?

Not accurate as the drivers are outdated.

where are gta4 or arma2???
the 2 games that could make use of more ram are not even tested
this review doesnt answer the "Is there any benefit to more memory?" since it doesnt test scenarios where it is relevant

As for the question of the drivers, we don't - and never have - used every single revision of the drivers, especially for ATI, who are committed to monthly driver updates regardless of whether they bring big performance increases. Sticking with one driver allows us to compare a card under review to a wider selection of other cards, and we feel that makes the coverage more useful to more readers.

Na thats just the standard we didnt want to run the benches again reply. You are far more tardy than ever regarding drivers, and 4 revisions is just a joke and bad practice. If usefulness to readers was a concern then the test games would have been revamped to something more current. GTA 4 and arma 2 for instance. Furthermore sticking to one ancient set of drivers makes meaningful comparisons to other competing cards and multi-card setups quite impossible. You are not expected to use every revision, but more than 2 revisions out smacks of idleness. This is not a point that should need discussing as it should never have arisen.

I mean what happened to being on the cutting edge? Dang its a bad crowd around here.