AMD Radeon 7990 reviews thread

[blastingcap]

Ladies and gentlemen, let's stop the personal attacks and instead address the merits of the reviews.

Super Moderator BFG10K.

Despite the catfights, I bet this thread has been 100% male so far.

The 7990 frametimes even with the early-beta frametime-optimized drivers are apparently still not on par with Kepler according to Tom's:

http://www.tomshardware.com/reviews/radeon-hd-7990-review-benchmark,3486-8.html (you can click on other games too, the results are the same: 7990 microstuttering is higher as measured by frametimes)

This is troubling.

 

[Will Robinson]

oh no!...not Folding @ Home too

 

[Chiropteran]

Anyone else notice this part of the AT review- "With Battlefield 3 we ended up encountering a terrain issue on NVIDIA’s latest drivers, so we’re using our old NVIDIA scores for the time being until we get the issue sorted out"

Whatever happened to nvidia's "superior" graphic drivers? And TBH it all seems kinda biased, nvidia gets the benefit of it's new beta driver in the games where it works, but instead of just failing the test by default anandtech reverts back to previous drivers.

Pick and choose the best driver for each game, fair much?

 

[Lonbjerg]

Fine....Anandtech delivered a useless Multi-GPU review.

It dosn't included frametimes or runt frames, instead it use a dumbed down single FPS number.

That number is useless..and false...and it will only lead to false posts, using the useless metric as "evidence".

Good job Anandtech...tro-lol-lol-lol...fast becomming irrelevant in GPU reviews.

 

[cmdrdredd]

Anyone else notice this part of the AT review- "With Battlefield 3 we ended up encountering a terrain issue on NVIDIA’s latest drivers, so we’re using our old NVIDIA scores for the time being until we get the issue sorted out"

Whatever happened to nvidia's "superior" graphic drivers? And TBH it all seems kinda biased, nvidia gets the benefit of it's new beta driver in the games where it works, but instead of just failing the test by default anandtech reverts back to previous drivers.

Pick and choose the best driver for each game, fair much?

What's wrong with posting scores that render correctly vs one that could potentially inflate scores??

 

[Chiropteran]

What's wrong with posting scores that render correctly vs one that could potentially inflate scores??

So you would be fine with using a broken driver that only works correctly in one game as a valid measure of that games performance?

If a driver is broken in even a single game, it is a useless driver for anyone who plays that game, and shouldn't be used in the review at all.

 

[3DVagabond]

Fine....Anandtech delivered a useless Multi-GPU review.

It dosn't included frametimes or runt frames, instead it use a dumbed down single FPS number.

That number is useless..and false...and it will only lead to false posts, using the useless metric as "evidence".

Good job Anandtech...tro-lol-lol-lol...fast becomming irrelevant in GPU reviews.

It doesn't matter how many times you post the same thing.

 

[cmdrdredd]

So you would be fine with using a broken driver that only works correctly in one game as a valid measure of that games performance?

If a driver is broken in even a single game, it is a useless driver for anyone who plays that game, and shouldn't be used in the review at all.

They did not use the driver that was broken. That's the whole damn point.

It didn't work in one game so the threw that out and gave numbers from a working driver for that game. They explained why they did too. I don't see any thing wrong with that. They are comparing performance on as level a field as they can, not trying to find the best driver for everything.

 

[parvadomus]

Where are frametimes and runt frames?
You now understand why anandtech makes USELESS GPU reviews?
This tell the TRUE story:

http://www.pcper.com/reviews/Graphics-Cards/AMD-Radeon-HD-7990-6GB-Review-Malta-Gets-Frame-Rated

The title gives an important clue:

"AMD Radeon HD 7990 6GB Review - Malta Gets Frame Rated"

Using a single FPS number about multi-GPU is retarded....thx anandtech for the useless review....you only gave people false "info" to use...ffs.

Nice that you point to the most NV biased site out there (ask yourself why they didnt use the driver provided by AMD). Why not talk about for example, tomshardware finding? The beta driver removes runt frames on every game. :awe:
But I know biased sites are the favourite for trolls like you :biggrin:

 

[Silverforce11]

Despite the catfights, I bet this thread has been 100% male so far.

The 7990 frametimes even with the early-beta frametime-optimized drivers are apparently still not on par with Kepler according to Tom's:

http://www.tomshardware.com/reviews/radeon-hd-7990-review-benchmark,3486-8.html (you can click on other games too, the results are the same: 7990 microstuttering is higher as measured by frametimes)

This is troubling.

Except its not perceived, such tiny variations. Blind testing in their "pepsi challenge", all the games except for 2 had complete smooth gameplay. Tom also did a radeon pro review and that utterly made it smoother than NV's solution. This doesnt change much, CF is for ppl willing to tweak a bit to get faster perf for less. The problem is, the 7990 isn't less expensive. Its not justified to equal a 690 in price when all these negatives are factored in.

 

[BallaTheFeared]

At 40 fps in CF there is no way it was smooth just because the frames delivery was.

 

[cmdrdredd]

Nice that you point to the most NV biased site out there (ask yourself why they didnt use the driver provided by AMD). Why not talk about for example, tomshardware finding? The beta driver removes runt frames on every game. :awe:
But I know biased sites are the favourite for trolls like you :biggrin:

They didn't use the driver because it would not be indicative of the expected performance and usage that someone buying the card gets. You can't get that driver, so when you buy it you don't get the fixes.

This is a case of you not even bothering to read. They actually showed the new drivers performance in their testing on a separate page and explained why they didn't use it. Put simply they didn't want people to think that's how it works out of the box, it doesn't.

 

[Chiropteran]

They did not use the driver that was broken. That's the whole damn point.
.

Yes they did, in every other game. In reality, you can't pick and choose and use a different driver for each game. You install one driver and use it for everything. If BF3 won't work with the new nvidia driver, than the new nvidia driver shouldn't have been used for any of the benchmarks.

It's hilarious because it isn't even new nvidia hardware, it's just total ineptitude that they put out a broken driver.

 

[cmdrdredd]

Yes they did, in every other game. In reality, you can't pick and choose and use a different driver for each game. You install one driver and use it for everything. If BF3 won't work with the new nvidia driver, than the new nvidia driver shouldn't have been used for any of the benchmarks.

It's hilarious because it isn't even new nvidia hardware, it's just total ineptitude that they put out a broken driver.

Are you even listening/reading? The driver only has issues in one game. That's it. Just one. So they didn't use it for that one game. Its not that hard to understand.

If they used that driver for battlefield 3 you would have a hissy fit over it. Then if they didn't use it at all someone would say "but...but...new driver"
So they did the right thing, leave that driver out for that game and explain the issue. It really is that simple. Your on a witch hunt when there is none to be found.

 

[BallaTheFeared]

Yes they did, in every other game. In reality, you can't pick and choose and use a different driver for each game. You install one driver and use it for everything. If BF3 won't work with the new nvidia driver, than the new nvidia driver shouldn't have been used for any of the benchmarks.

It's hilarious because it isn't even new nvidia hardware, it's just total ineptitude that they put out a broken driver.

:hmm:

 

[Chiropteran]

Are you even listening/reading? The driver only has issues in one game. That's it. Just one. So they didn't use it for that one game. Its not that hard to understand.

Yeah okay, you install a driver just to play one game, and then when you are done you uninstall it and install the other driver? Nobody does that.

In reality you can't pick and choose your driver on a game by game basis. If they wanted to be fair, they should have tested the 7990 with 2 different drivers and used the best score for each game. They did not.


It's a flawed testing environment.

 

[BallaTheFeared]

Why not? You wouldn't even have to reboot.

By your logic though you assume people play 17 different games all day long?

 

[cmdrdredd]

Yeah okay, you install a driver just to play one game, and then when you are done you uninstall it and install the other driver? Nobody does that.

In reality you can't pick and choose your driver on a game by game basis. If they wanted to be fair, they should have tested the 7990 with 2 different drivers and used the best score for each game. They did not.


It's a flawed testing environment.

Lol...no its called explaining all aspects of a situation to the reader. If battlefield 3 was your main game and it mattered to you more than any other game, that information would be super important would it not?

Unlike most sites, they are assuming your not the average GameSpot reader.

 

[Chiropteran]

Why not? You wouldn't even have to reboot.

By your logic though you assume people play 17 different games all day long?

No, but I'm assuming some people who play battlefield 3 might play some other game occasionally. Also, if the driver is broken in BF3 it's probably broken in some other games as well, just because their specific benchmark tests didn't reveal the broken doesn't mean it isn't there.

Why does nvidia even have a beta driver in this test? It's a 7990 review.

Lol...no its called explaining all aspects of a situation to the reader. If battlefield 3 was your main game and it mattered to you more than any other game, that information would be super important would it not?

Uh, they don't explain "all the aspects". Like, what if you actually play battlefield 3 and can't use the broken nvidia driver, then how is your performance going to look for the other games? Nobody knows apparently, because they decided to use the broken driver for testing the other games instead of an actual working driver.

 

[BallaTheFeared]

Why does AMD have a prototype driver in the test?

 

[wand3r3r]

I believe I've seen reviews where the cards drivers didn't work and that game was left empty for that card.

 

[Chiropteran]

Why does AMD have a prototype driver in the test?

Because it is a review of an AMD video card? And because the driver actually works in all the tested games?

I believe I've seen reviews where the cards drivers didn't work and that game was left empty for that card.

Yeah, I have also. Why is nvidia given special treatment here, reverting to an old driver? They should get a big fat F for failure if the driver used doesn't actually work.

 

[BallaTheFeared]

Because it is a review of an AMD video card? And because the driver actually works in all the tested games?

Lol? Are you kidding right now? I see other cards in the review, though please correct me if I'm wrong.

Where were you the last year and a half when CF didn't work at all and was still in reviews and AMDs cards were stuttering up and down the charts for months, but now because of one issue in one game you're completely dumbfounded?

I bet you lost sleep when AMD sent out half working boost bios to reviewers after a tech site leaked a review weeks before the NDA dropped, must have been a rough time for you.

Really?

 

[Chiropteran]

Lol? Are you kidding right now? I see other cards in the review, though please correct me if I'm wrong.

Where were you the last year and a half when CF didn't work at all and was still in reviews and AMDs cards were stuttering up and down the charts for months, but now because of one issue in one game you're completely dumbfounded?

Really?

Uh, that is the whole point. If an AMD driver has issues causing micro stutter, the review points this out. It doesn't go and switch drivers to one that doesn't have the issue. Here is the same situation for nvidia, except nvidia gets a free driver change to best possible possible driver for that game.

 

[BallaTheFeared]

They pointed it out, said so and switched to another driver.

What's your issue really, why not just say it? You're mad they used a performance increasing beta driver from Nvidia, this upsets you because it was a review to showcase AMD and drum up sales.

Things are a lot easier when you don't try to hide your real agenda.