• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

4870 vs GTX 280 Devil May Cry 4 Benchmark

heatlesssun

Extremely [H]
Joined
Nov 5, 2005
Messages
44,154
Looks very impressive if this is a real benchmark, my guess is that it is.

So here's here's my single GTX 280 (one GTX 280 was bad and just got RMA'ed) clocked at 660/1420/1150 on sig rig results:

devilmaycry4trialdx1020ax7.jpg


And here's the 4870 I presume (source link: http://forums.vr-zone.com/showthread.php?t=292539)

dmc4dx1019x12gk0.jpg


My settings are all Super High 8xMSAA as this post claims so this should be an apples to apples comparison. Of course this game doesn't look to be very stressful but the demo does look good.

Time for you future 4870 owners to gloat!:p
 
For all we know the GTX 280 may not even be optimized for that. The game won't be out anytime soon so that makes sense. ATI may have went and optimized their drivers to good in that bench. If you got CS:S do that bench and see what happens.
 
For all we know the GTX 280 may not even be optimized for that. The game won't be out anytime soon so that makes sense. ATI may have went and optimized their drivers to good in that bench. If you got CS:S do that bench and see what happens.

?

The game is in TWIMTBP program and it's no secret Capcom has been working with NVIDIA for awhile to properly optimize their PC titles.
 
?

The game is in TWIMTBP program and it's no secret Capcom has been working with NVIDIA for awhile to properly optimize their PC titles.

yup thats right , and this demo say gains of around 30% with the 177 drivers, so nvidia definitely optimized their drivers for this game.

The result looks gr8 .
 
I can make my x1950 beat both of those given 5 minutes on Gimp, or even photoshop. I would wait until we have our own benchmarks.
 
For all we know the GTX 280 may not even be optimized for that. The game won't be out anytime soon so that makes sense. ATI may have went and optimized their drivers to good in that bench. If you got CS:S do that bench and see what happens.

Mmm. I smell an NV fanboy.

~S
 
Interesting. I'd like to see what a 4850 scores.

In crossfire it never went below 70-80fps running 1920x1200 w/ 8MSAA and everything else to the highest levels in game. Typically the FPS was running 80-100fps for the short time I watched the benchmark. This game clearly does not push hardware too hard.

That is with a Q6600 @ 3.5ghz and two 4850's stock speeds.
 
I'm going to give this bench a shot with my GTX280, I'm assuming the ingame settings were maxed, how about Nvidia CP settings?
 
Can't wait for all the reviews to start coming out. Now I gotta choose between a 4870 or a pair of 4850s. I may be willing to give up the extra speed of CF to not have to deal with any potential CF issues.
 
I have a new system being shipped with a GTX 280 (and a Gigabyte x48 mobo), I'm starting to wonder if I should consider putting a couple 4870's in there instead.

Whats the expected cost and release date of the 4870?
 
Thanks for that info.

I wonder if it would be better for a GTX 280 owner to switch to a 4870 crossfire setup (if it proves to be a lot better than the GTX 280) or wait out his Step Up time frame for an updated GTX 280 (maybe a GX2)?

Does anyone think a GX2 or similar Nvidia card will be released before the end of September?
 
Doubtful on a GX2 based on the current GT200 chip.
Possible GT200b later. (55nm)
 
Did anyone notice the "4870" benchmark at the top had a lower clocked Core2 while the 280 was paired with a higher clocked Extreme? That's funny. It still almost beat the 280. Another 340MHz and it might have.
 
I know it's a shot in the dark but anyone know what the 4th test is specifically stressing? Just the fact it's the only thing the 4780 beats out the 280 and it's the widest margin.
 
why is tazzmisshonx's scores much lower than the OPs GTX280?

His gtx280 is stock and the op's is overlcocked (which bodes well for the 4870 result if its true and stock) and his cpu is running slower than the op's.
 
His gtx280 is stock and the op's is overlcocked (which bodes well for the 4870 result if its true and stock) and his cpu is running slower than the op's.

Yeah but there's only a 10% difference in clocks and the scores are 50% higher...
 
I hope this is true. Being able to get a nice Intel board and have the option to use dual GPU's would make me sooooo happy.
 
Yeah but there's only a 10% difference in clocks and the scores are 50% higher...

I believe it is because the OP's CPU has a full amount of L2 cache and the first pic has the X9650. The former is extremely powerful and the latter has more performance than the cache crippled Q9300.
 
e8400@3.6
3870x2@stock

Had to use a camera for the pic, couldn't get one for some reason that would copy, kept getting a blank field, maybe a xfire issue I dunno. Anyhow..

I wouldn't trust this game as a benchmarking tool...



edit: wanted to add, all settings same as op, 8x, super high
 
Here's my benchmark @ 1600x1000 on a single 4850. 1600x1000 is the max resolution I can select because my monitor is native 1680x1050. I'm also using 8.6 drivers and my 4850 is overclocked to 690mhz Core, 1123mhz memory. It ran at around 85C during the benchmark.

Also, the 2nd and 4th scenes are very taxing. The 2nd scene is because of the amount of texturing and shading you get because its set in a jungle. I beat this game on Xbox 360 and the jungle area levels had occasional slowdown, so its just a very taxing area to render. The 4th scene is taxing because you start out with around 10 enemies all doing something different with complete independent actions and it slowly builds up until you have 30 enemies all doing something different.

dmc44.jpg
 
why is tazzmisshonx's scores much lower than the OPs GTX280?

I guess it's a combination of the overclock on the GTX 280 and the overclocked QX9650. I'm surprised that the CPU makes such a huge difference to the framerates in the game... I guess the 2.5GHz of the second GTX owner doesn't cut it? It's also got 6 MB of L2 cache... Even the q6600 has 8 MB... the QX 9650 has 12 MB.

Still, I'm amazed that an improvement in CPU can deliver such an impressive performance boost. Makes me excited about upgrading my Q6600 with Nehalem in a few months!
 
Something still does not sound right with my scores, can the OP post a screenshot of the ingame settings just to make sure that mine are set the same also?
 
If those scores are legit then ATI certainly has a winner on its hands. I've been a hardcore Nvidia supporter for some time, but if ATI has the goods then I have no reservations about switching over. Finally some real competition! If anything, it just means Nvidia will have to drop prices sooner rather than later. Its just all around a win-win situation (for the consumers at least).
 
Bill0527:

I'm on a 1680x1050 monitor, too, so I can't run any higher than that, either. Therefore, unless anyone wants to run the benchmark on lower settings, I can only compare to you. So, what settings did you run? Thanks!

EDIT: Also interesting, at least for me this benchmark does not report accurate processor speeds. Showed what was stock but I'm currently underclocking. *shrug*
 
Wait wait wait.

DMC4 is out on the PC?

Holy balls.

Haha, most of Capcom's bigger games are being released on PC now. Lost Planet, DMC4, Street Fighter 4, and obviously Resident Evil 5.
 
Can't wait 'till the 4870x2...
Someone knows anything about price?

Yup, I am waiting on a 4870 X2 as well, and now that I have a 24' widescreen, its going to be killer.. :eek:

Rumored release price of a 4870 X2 is going to be around 500 USD.
 
Here you are:
devilmaycry4trialdx1020vp7.jpg


Your performance does seem off.

Alrighty, I know what the issue is now, but I am at work so I'll rebench later tonight, the problem was that I had run my bench with msaa set to max also which was like 16xQ or something of the sorts, it still ran smooth though.
 
Back
Top