The Official 3870X2 Reviews Thread

And whys that? till now CF has sucked for scalability- and some games wont even register CF-


If you scour through a bunch of the reviews you'll see alot of benches showing CF 3870's almost up par to 8800 GTX SLI; where as you'll see some games where the CF 3870 doesnt even beat a 8800 GT.....

So when you say you doubt it- I have my doubts too - which is why i made my statement.

Also if you had 7950GX2's youd understand also lol.

http://209.85.135.104/translate_c?h...ttp://www.pcpop.com/doc/0/266/266637_12.shtml

this review shows some interesting info- but if you review some 3870's theres more interesting reviews....

I also state the same for 9800GX2- They might be on par or beat some games than 8800 Ultras in SLI but you'll prob see some games get beaten with the SLI setup than the Quad SLI / Quad CF setup. its such a mixup that it makes no sense to really upgrade if you already have this current gen Video cards....

 

Yeah there is a boatload of wishful thinking and willful ignorance out there about how CF and SLI are perfect solutions... Basically every gamer who wants to play demanding games but doesn't want to learn about the hardware. There's lots of crazy frothing excitement over 9800GX2 and 3870X2 on quite a few forums out there and people claiming a dual card setup is much better than a more expensive 8800GTX or GTS 512. The problem is they ignore the games that don't respond to SLI/CF or don't respond very well. Sure some games really do benefit, but there's definitely a risk of a game you want to play only working well with one of your 2 cards/GPUs.

 

Yeah well SLI and CF are doing better, CF does really well where games register it, and SLI does good on alot of games too- one thing ive noticed is there arent many games where SLI doesnt work at this point but still some games where CF doesnt-

But thats more than the point- If you think SLI and CF arent the solution- Than think about Quad SLI and Quad CF.

 

Just looked at the HardOCP 3870 X2 review and to say this product is underwhelming is a bit of a understatement. Barely beating the 8800GTX and then a number of times not doing it at a similar price but at the cost of potential Crossfire issues, heat, and power consumption all being higher/worse? Not good. Consider that both of those "cards" are barely faster than $250 cards on the market and it's just bad. At least the 8800GTX has the advantage of being relatively aged, the 3870 X2 is brand new though and that just simply HURTS. Bad product...

 

Copied from XtremeSystems' forums http://www.xtremesystems.org/forums/showthread.php?t=174988:
Anandtech uses a EVGA nForce 780i SLI, ATI: 8-451-2-080123a
Computer Base using Asus P5W DH Deluxe (Intel i975X) with Cat 8-451-2-080123a
Hardware Canucks using Asus Blitz Extreme (P35) with Cat 8.451.2-080
Hot Hardware using a EVGA nForce 680i SLI, Cat ATI Catalyst Bets v8.451
PCPOP using ASUS Maximus Formula (X38) and Cat 8.1 (no indication of updated drivers)
TechPowerUp using a Gigabyte P35C-DS3R Intel P35, Catalyst 8.12 v8.451.2
Tech Report using Gigabyte GA-X38-DQ6 with Cat 8.451.2-080116a-057935E
Tom's Hardware using Asus P5E3 Deluxe (X38) with Cat Asus P5E3 Deluxe (X38)

 

Anandtech's Crysis results are interesting, consider my system out performs both their 8800GTS 512 and the HD 3870X2, one by 4 FPS the other by 1 FPS. 1 FPS is neligable, but 4 FPS difference for the same card? They're also using more memory than me and a much newer motherboard (mine is P965 based, their's 780i), and while I have a 200Mhz CPU advantage their's has much more cache, and Crysis doesn't even scale anymore at those frequencies anyway...

 

Yeah you can kind of tell whos using Canned Benchmarks and whos not.

Hard OCP's review is interesting. I also found PCPOP to be an interesting review lol. It seems most of the reviews now are using ATI's 8.45.12 drivers in which they said would be the best for framerates.

Im not surprised to say the least, I see half the reviewers get complete mixups in reviews. One will beat a certain card the other wont.

To me its just a mixup- but I can say this much i dont think this is a G80 Killer at all lol. I read the Techreport review which doesnt use a 8800 GTX but a Ultra and they get worse framerates in most of the games than any other Reviewer does.

 

As bad as H's review makes the X2 look, Anand's makes it look that good.

And I dont get the "canned benchmark" talk. My mind is very split on it, but I would think on this allegedly technical forum there would be a bit more rigorous take on it. I'm not sure why a "canned benchmark" would apparantly inherently favor AMD's card every time, and how exactly that would occur without "cheating". Case by case premeditated cheating in fact.

The alternative , in the absence of being GPU limited, is to suggest that somehow NVidia's cards suddenly handle err.."intense" scenes better. And that's a whole nother rather ludicrous notion, really.

As people said, H claimed to play the whole bog level for their COD4 benchmarks, well just depending on where the player looks or whether the smoke grenade goes off you could see vast differences between playthroughs. I think that's the point, it's not repeatable.

Canned benchmarks have served us in distinguishing cards all these years, will continue to indefinitely I'm sure, and I dont see why just now they're suddenly useless.

Not saying the X2 is a good card BTW. I wouldn't touch one of these dual GPU monstrosities from either company with a ten foot pole personally..

 

Looking at anandtech and computerbase reviews using the lastest beta drivers the x2 wins in most cases against the more expensive ultra. In some cases by huge percentage points it aboustely screams in older dx 9 games. In rainbow 6 in rapes the ultra. Yes the dual gpu version from nVIDIA wil be faster but much more expensive, not mention the x2 is faster then sli gt in some cases. I am getting one as soon as the price is closer to $400

 

ComputerBase link is wrong its a SLI/CF article, here is the X2 review.

 

Computerbase article has totally different results than others, this review the card looks really great in dx9 games, in the game Anno1701 CF scale 100% almost all resolution what tested (some cases X2 performance more than twice as fast as a single hd3870 ), but at dx10 not so good.
Something fishy with the COJ numbers ,the game never runs on any currently available HW playable framrate at DX10, not even at 1280x1024.

 

Anand's results have me not believing though simply because I get better results on my system by a sizable amount over their 8800GTS 512. 4 FPS is nothing to forget about when we're down to the 30~ levels. That is a canned benchmark, so why the large difference?

As for Call of Duty 4: Anandtech said they benchmarked a CUT SCENE. I'm sorry, but that's just... ugh. A cut scene? I mean really, they could have done a lot better than that. They could have at the very least recorded a demo of actual play and use that. Which I figured is what HardOCP did to get consistent results. Yet another game I'll have to look over myself...

Really the range of benchmarks and how they're done does get me. I can't stomach someone using a cut scene though for a game benchmark, that simply doesn't cut it and really calls into question the entire review for me personally. Especially when there are well known and useful tools for said game such as /record...

 

RADEON HD 3870 X2 2x512MB PCI-E@iXBT incl. HIS, MSI, Powerolor (TUL) and GeCube -->>>>>> http://www.ixbt.com/video3/rv670-3-part3-d.shtml#p18

Look at Crysis results.

 

Eh? Come again?

If a cutscene is rendered using the games 3D-engine and isn't a movie file then what's the beef?

In this case all a cutscene is...wait for it...wait for it... A recorded demo that's been edited for cinematic purposes.

In other words a cutscene rendered in realtime in the games 3D engine is no different than a user recording a demo of a playthough. Other than the fact that it's been edited with camera angles and cuts to enhance cinematic viewing.

Regards,
SB

 

Dunno why [H] seems to get worse results than TR and Anand. Maybe something to do with the "driver tango" AMD was apparently engaging in during the review period, which is never fun for anyone.

I'll be curious to see how mobo compatibility holds up as this thing actually gets into the wild on a bunch of platforms. Will it work just as well on Intel and particulary NV chipsets? I seem to recall GX2 having some issues.

Man, another 10.5" board. I'm not putting another 10.5" board in this case, it was too traumatic last time. And this case is only 2 years old, and pretty nice. I'll have it for several more years in all probability. Even 10.25" would be a big help. Ah well.

 

Bummer. That computerbase review used 5 different drivers for ati and 4 for nvidia cards.

 

Check out Tech Report's review http://techreport.com/articles.x/13967

It's interesting, the 3870x2 comes out significantly ahead of the 3870 Crossfire in Call of Duty 4 and Half-Life 2, it's about equal in Unreal Tournament 3, and gets beat by Crossfire in Crysis.

I wonder if it's solely due to the differences in clock/memory speeds between the two, or if there is something else going on here.