More information about R9 290X

[Elfear]

Rabbit with a pierced ear, creepy hands, and a thermometer in his mouth. Apparently I've really been limiting myself in the emoticon department.

 

[Arkadrel]

Rabbit with a pierced ear, creepy hands, and a thermometer in his mouth. Apparently I've really been limiting myself in the emoticon department.

Thats a horny rabbit, salavating/drooling at the thought of something sexy (the blushing cheeks),
honking a pair of air/imaginary boobs.

^-^

"thermometer in his mouth" you crack me up, you ve never watched any Anime series? like DBZ ect?



http://en.wikiquote.org/wiki/Shadow_Warrior

When seeing rabbits mate:

• "Having fun little Rabbit? Ha ha!"
• "Horny rabbits, hahaha."
• "Push in the bush there rabbit, haha."
• "Oh, queer bunny."
• "Oh you fa**ot rabbit."

Not sure what it is with asian culture, but to them rabbits are horny lil fellows.

If you played Shadow Warrior, youd remember all the fun referances/jokes he made about those guys

http://www.youtube.com/watch?v=aRJaZlgZRak&t=10m26s

 

[Elfear]

Thats a horny rabbit, salavating/drooling at the thought of something sexy (the blushing cheeks),
honking a pair of air/imaginary boobs.

^-^

"thermometer in his mouth" you crack me up, you ve never watched any Anime series? like DBZ ect?



http://en.wikiquote.org/wiki/Shadow_Warrior

Not sure what it is with asian culture, but to them rabbits are horny lil fellows.

If you played Shadow Warrior, youd remember all the fun referances/jokes he made about those guys

http://www.youtube.com/watch?v=aRJaZlgZRak&t=10m26s

Sorry. Only watched a couple anime and they didn't have any horny rabbits. I can see I've been missing out.

 

[GaiaHunter]

T
Not sure what it is with asian culture, but to them rabbits are horny lil fellows.

It isn't an asian culture thing.
Male rabbits will mate with a ton of female rabbits in a quick succession during mating season.

 

[DarkKnightDude]

I was unsure if this was a thread discussing horny bunnies or the r290x specs.

Anyways, I find it impressive they were able to cram all that considering the amount of space.

 

[sontin]

So about a 36% increase over 7970GE if true. That's actually faster than I was expecting.

3DMark11 is not really usefull because the tests are to short and the break between them is to long which helps the thermal boost clock.

A GTX780 is also 36% faster than the 7970GHz...

 

[Erenhardt]

3DMark11 is not really usefull because the tests are to short and the break between them is to long which helps the thermal boost clock.

A GTX780 is also 36% faster than the 7970GHz...

You comparing fan speeds or what?

 

[Skurge]

3DMark11 is not really usefull because the tests are to short and the break between them is to long which helps the thermal boost clock.

A GTX780 is also 36% faster than the 7970GHz...

It's not a useful benchmark for a number of reasons other than that. So I take nothing from it.

 

[wand3r3r]

3DMark11 is not really usefull because the tests are to short and the break between them is to long which helps the thermal boost clock.

A GTX780 is also 36% faster than the 7970GHz...

Yep, 3DMark11 isn't an accurate indication across different brands.

780 4511
7970 3304
Difference = 36% in 3Dmark11
Reality = 18%

Clearly 3DMark11 was optimized (or the drivers) for NV's architecture.
http://www.guru3d.com/articles_pages/radeon_r7_260x_r9_270x_280x_review_benchmarks,21.html
http://tpucdn.com/reviews/MSI/R9_280X_Gaming/images/perfrel_2560.gif

We can speculate anyway if they are ~equal (290x and 780) in 3DMark then the 290x will beat the 780 soundly as the 780 needs ~18% more in 3DMark to tie the 7970.

 

[Fox5]

Yep, 3DMark11 isn't an accurate indication across different brands.

780 4511
7970 3304
Difference = 36% in 3Dmark11
Reality = 18%

Clearly 3DMark11 was optimized (or the drivers) for NV's architecture.
http://www.guru3d.com/articles_pages/radeon_r7_260x_r9_270x_280x_review_benchmarks,21.html
http://tpucdn.com/reviews/MSI/R9_280X_Gaming/images/perfrel_2560.gif

We can speculate anyway if they are ~equal (290x and 780) in 3DMark then the 290x will beat the 780 soundly as the 780 needs ~18% more in 3DMark to tie the 7970.

So maybe it's not accurate for current games, but 3dmark's thing has always been "This is what the games of the future will look like."

And I'd wager that historically, the card that does better in 3dmark does do better in the long run. That said, it's probably at a point when that card is on the verge of obsolescence.
Some things also never come true, like 3dmark has been predicting heavily threaded games for a while. Maybe they should just license their engine to games.

Still, I guess it could be optimized for nvidia, there is a physx benchmark iirc, and there's enough difference between nvidia and ati that you could favor one or the other. I think Intel chips generally do above average in 3dmark too.

 

[SirPauly]

Still, I guess it could be optimized for nvidia, there is a physx benchmark iirc, and there's enough difference between nvidia and ati that you could favor one or the other.

Physics Test

A scene with a large number of rigid bodies is simulated and rendered. The rendering is done using light weight
techniques. The rigid bodies collide with each other and some of them are connected with joints. The simulation is divided to multiple threads by partitioning the simulated world to several isolated regions. The Bullet Open Source Physics Library C++ path is used as the physics SDK for this test, and compiled and linked statically into the test binary at development time.

http://www.3dmark.com/wp-content/uploads/2010/12/3DMark11_Whitepaper.pdf

 

[blackened23]

3dmark isn't a physx benchmark. It isn't nvidia optimized either. That said, it is a synthetic benchmark (synthetic means that both AMD / NV can optimize their driver for it) but that doesn't mirror real world performance in all cases. The only issue with 3dmark11 is that it is STRONGLY affected by CPU, overclocking, and memory overclocking. I used to hang around OCN's software/benchmarks forum and they have top 30 lists going around (or used to, probably still do) and all of those benchmarks are very much affected by CPU, overclocking, memory overclocking, so on and so forth. So it's really difficult to just simply look at a score and discern anything from it because the results vary widely depending on how much you're overclocking. The CPU physics score is COMPLETELY CPU based. So anyone that is overclocking gets a huge boost on their 3dmark score.

So, essentially, you can't simply look at a 3dmark score and learn anything from it. 3dmark is a useful benchmark if the testbed for benchmarking is consistent (I mean, completely consistent) when a reviewer tests various cards in the very same system. But when comparing multiple systems with different components? Eh? Not so much. Not useful. It is only useful if the testbed to compare cards remains consistent, as it would be for a reviewer testbed - as said CPU, memory, and overclocking dramatically affect the score (and especially the physics score).. So i'd wait a couple of days before reading too much in a 3dmark result.

 

[el etro]

So maybe it's not accurate for current games, but 3dmark's thing has always been "This is what the games of the future will look like."

And I'd wager that historically, the card that does better in 3dmark does do better in the long run. That said, it's probably at a point when that card is on the verge of obsolescence.

Good speculation here. Nvidia Fermi and Kelper really has ahead of AMD architectures in recent past games. And now we seen the game is gonna tie in those new games(Look at BF4, Saints Row, Rome II benchmarks).

 

[MrK6]

Wow, looks to be a mighty powerful card if those benchmarks are accurate.

 

[Elfear]

3DMark11 is not really usefull because the tests are to short and the break between them is to long which helps the thermal boost clock.

A GTX780 is also 36% faster than the 7970GHz...

I agree that 3DMark isn't all that useful in the long run. There are few synthetics out there whose performance can be extrapolated to real world gaming.

As wan3r3r pointed out though 3DMark can be useful for comparing cards from the same company.

 

[BallaTheFeared]

Only problem there is figuring out why A is slower than B, and how that will translate into real world gaming.

 

[Cloudfire777]

 

[csbin]

 

[Cloudfire777]

Nice score.

Looks like 290X is beating Titan by roughly 5% from what we have seen so far. Maybe it will even out when we get more games included in the scores.

 

[Face2Face]

Performance is looking good :thumbsup: I wanna see some higher overclocks on Hawaii though since we all know how well the GTX 780 overclocks.

 

[csbin]

800mhz base(1000mhz boost)

 

[Cloudfire777]

Performance is looking good :thumbsup: I wanna see some higher overclocks on Hawaii though since we all know how well the GTX 780 overclocks.

The good thing about R9 Hawaii series is that they have 512bit memory bus. That will greatly help with super bandwidth while running low memory clock to ensure a stable card.

Pretty sure that the chart is just scratching the surface of what R9 290X can do

 

[Tweak155]

The good thing about R9 Hawaii series is that they have 512bit memory bus. That will greatly help with super bandwidth while running low memory clock to ensure a stable card.

Pretty sure that the chart is just scratching the surface of what R9 290X can do

Did AMD send you a check or something?

 

[Cloudfire777]

Did AMD send you a check or something?

This is funny.

One moment I`m a Nvidia fanboy, the next I`m paid by AMD.

Anandtech forums, either black and white. Nothing in between

 

[Face2Face]

The good thing about R9 Hawaii series is that they have 512bit memory bus. That will greatly help with super bandwidth while running low memory clock to ensure a stable card.

Pretty sure that the chart is just scratching the surface of what R9 290X can do

I hope you are right. The Titan is sporting a 221Mhz OC and the 290X is only clocked 100Mhz higher... I am hoping Hawaii overclocks similar, if not better than Tahiti.