Crossfire: AnandTech vs. [H] reviews

O

orkan

[H]ard|Gawd
Joined
Jul 9, 2002
Messages
1,488
What gives?

Am I reading something wrong here? [H]'s review of crossfire seemed to have it labeled as weak sauce, while Anandtech's review showed it to be a contender.

Normally the reviews of both sites come out about the same with minor differences. Something I missed?
 
It shows great potential for use with R520's and ATI's next gen but for the X800's it's pretty weaksauce.

That's the basic gist I've got from the reviews so far anyway.
 
If you base it off the actual technology then it looks very promising with great performance gains and features. If you base it off the x8x0 versions then it isn't anything special (although I still think the performance gains are impressive)
 
It looks the same as SLI at increasing performance with 2 cards, but SLI has an extra year of maturity and seems to be a better thought out solution. I just don't like the master/slave card & dongle thing.

I'm sure Crossfire, could turn out to be a real contender with the next gen cards that will probably be planned a little better and obviously perform better. (maybe matching cards - no master /slave crap or limited 1600x1200@60hz)

But IF I was in the market for a videocard right now. THe clear choice would be a 7800GTX.
 
tornadotsunamilife said:
If you base it off the actual technology then it looks very promising with great performance gains and features. If you base it off the x8x0 versions then it isn't anything special (although I still think the performance gains are impressive)
Click to expand...

currently the ACTUAL technology is the x8x0 line of cards, and that's it. If you specualte and extrapolate to it's POTENTIAL performance in a card not yet released, it might be cool.

At the moment, it is not actually as good as SLI, and it is hard to make the argument for EITHER technology as being a good value to anyone other than the person willing to plunk down the cash for two cutting edge cards to get the highest performance possible. For everything else, after you take into account motherboards, PSUs, and buying a second card, it pretty much is a cheaper or equal cost decision to go with a card further up the performance scale.
 
I can't speak wrt their testing methodology. I can only speak on behalf of ours, and to that what you see in our review is exactly what I experienced while playing those games on the cards tested. That was as high as I could set the quality settings on each game and still have smooth gaming performance.

I would urge the point that timedemo "benchmarks" and apples-to-apples "benchmarks" are not an accurate means to gauge the gaming experience delivered by each setup.

They may guage something else, but it isn't the true gaming experience IMO.
 
It looks promising, but it's not a contender just yet. When the R520 cards come out, then it may be a real heavyweight contender, but not right now as you can only use X800 or X850 cards with it.
 
The [H] review made the cards look pretty good really. They scaled exactly the same as the single X800 line did between the Geforce 6 series. Seems pretty accurate to me.

The biggest limitation with Crossfire is the resolution limit. R520 may or may not overcome this.
 
I initially thought the [H] review made i look wweak as well... then went to look at it again after reading the Anand one..

The changes in settings between the cards seems very subtle, and I didnt notice a lot of the adjustments on the crossfire vs the sli in [H]'s benchmarks
 
The performance is in line with what I expected, better than 6800ultra SLI in most cases and AA quality is better than anything under NV kingdom. Unfortunately its too late for the x800 series, its much better and even cheaper to upgrade to a 7800gtx than getting a xfire mobo+ xfire master card. The question is if its worth waiting for R5x00 crossfire.

Whats really appealing is the crossfire chipset, very overclockable and performs on par or better than the nforce4 SLI. Having Azalia audio is a great plus.
 
I don't know why they're even making this retail. Even the most seasoned ATI fan boy will admit that it's much wiser to spend 500$ on a 7800GTX than to spend 900$ on a Crossfire set-up for the same performance and a 1600x1200 60hz limit.
 
Brent,

In your opinion (emphasis on opinion), Anandtech was probably using "canned" benchmarks with like settings accross both platforms? IE the age old battle between [H] "gaming experience" review style vs. the standard apples to apples approach?

Even that usually shows like performance among different review sites. This time however, ATI's crossfire seemed to come in pretty low on the totem pole at [H], where other sites (not only anandtech) were raving about it.

I'm going to go read your review again. Maybe it will make more sense to me this time.
 
I do think that [H]ard kinda ignored most of CF's good points and just didnt mention them.
I dont think that they skewed any benches because while the H may leave some points out here and their (We are only Human) they never skew the Benchmarks like Toms of HA.

I know what Im saying is kinda offensive since I love this sight and most every review has been great so dont get mad at me Brent or Kyle! :(
I've found that normaly most ATi Features are kinda shruged off. Its never really been all that bad here and your not biased. I mean you had a review of SM3.0 that totally put it in a realistic light with out all that marketing crap. I dunno maybe its just me being over zealous. Agian Sorry! :D
 
orkan said:
Brent,

In your opinion (emphasis on opinion), Anandtech was probably using "canned" benchmarks with like settings accross both platforms? IE the age old battle between [H] "gaming experience" review style vs. the standard apples to apples approach?

Even that usually shows like performance among different review sites. This time however, ATI's crossfire seemed to come in pretty low on the totem pole at [H], where other sites (not only anandtech) were raving about it.

I'm going to go read your review again. Maybe it will make more sense to me this time.
Click to expand...

It is a point of view is what the [H] does. And honestly...I like it. It is a great second view point beyond the apple to apple.

What the [H] does is look at data a bit more in detail other canning. Sorta like the 50% of all marriages end in divorce. That is true...but do they never clarify that point by saying 50% of all total marriages end in divorce. Remember that people who usually get divorced once are repeaters in that field. The actual number is 32% of first time marriages fail.

I think my favorite one is the amount of premature babies in the 20's....oh yeah...I forgot, people back then never had sex before marriage :D

-tReP
 
When you read the reviews ignore the comparisons with the 7800** line of cards

look soley at the 6800 / x8* and the system looks good as mentioned.

What i didnt like was the "cons" of hard to install BS - come on - how hard is it to remeber you need to put one card on top, and one on bottom - not like people complained about dual channel ram and neededing to use other slots. - that was lame and to me seemed like the reviewer "needed" to find something wrong with it.
 
orkan said:
Brent,

In your opinion (emphasis on opinion), Anandtech was probably using "canned" benchmarks with like settings accross both platforms? IE the age old battle between [H] "gaming experience" review style vs. the standard apples to apples approach?

Even that usually shows like performance among different review sites. This time however, ATI's crossfire seemed to come in pretty low on the totem pole at [H], where other sites (not only anandtech) were raving about it.

I'm going to go read your review again. Maybe it will make more sense to me this time.
Click to expand...

I don't think AT was "raving" about Crossfire?

http://www.anandtech.com/video/showdoc.aspx?i=2541&p=7
The final verdict on CrossFire is very mixed. It is clear from our Game tests that SLI has a worthy competitor with parts in its price range at 1600x1200@60Hz and below. However, we have a hard time buying the idea that many gamers are going to shell out the money necessary for CrossFire with that kind of limitation.
Click to expand...

It is very hard for us to support Super AA having seen the abysmal performance scaling we have shown here.
Click to expand...

we have to strongly recommend against the purchase of an X800/X850 series CrossFire card.
Click to expand...

Could they have slammed it worse? We have different definitions of "rave"?
 
That is kind of odd...

They show it coming out ahead in like 90% of their benchmarks... and then slam it like that. I never noticed it I guess. :(
 
orkan said:
That is kind of odd...

They show it coming out ahead in like 90% of their benchmarks... and then slam it like that. I never noticed it I guess. :(
Click to expand...

They slam it because it will cost a HUGE amount, when you could just get a GTX for like half the price and get about the same performance.
 
orkan said:
That is kind of odd...

They show it coming out ahead in like 90% of their benchmarks... and then slam it like that. I never noticed it I guess. :(
Click to expand...

It's coming out ahead of dual 6800-Ultras - which are already being phased out. The criticisms about Crossfire havent been about performance - it scales really well, but the timing, cost and technical limitations make it unattractive at this point.

If you had $700 today you could buy 2xX850XT or 2x7800GT. Do you think it will come out ahead in 90% of benchmarks then?
 
trinibwoy said:
If you had $700 today you could buy 2xX850XT or 2x7800GT. Do you think it will come out ahead in 90% of benchmarks then?
Click to expand...

The whole reason for the res/refresh limits was the fact that you can use your existing cards...so you dont have to shell out$ $700 for xfire...more like $300 to $400. It still does not make it worth it...but you should consider all of its possible uses then make fun of it :)
 
Everything is hinging on r520, imho.

r520 has to be significantly better than 7800 in some significant area, either stellar performance, or equivalent performance at a much lower cost.

IF that happens, then we'll really see if Crossfire is a competitor to SLI.
 
Jbirney said:
The whole reason for the res/refresh limits was the fact that you can use your existing cards...so you dont have to shell out$ $700 for xfire...more like $300 to $400. It still does not make it worth it...but you should consider all of its possible uses then make fun of it :)
Click to expand...

$300-$400? You are forgetting the price of the Motherboard. The only possible use for X850 crossfire is to satisfy one's curiosity or waste money.

Say I have an Nforce4 PCI-e board and an X850XT. Do I :

(Option 1)
- Sell motherboard
- Buy unproven crossfire motherboard with non-existent community support
- Buy x850 master card with excellent ps2.0 performance and 16x12 60hz limitation

or

(Option 2)
- Sell X850XT
- Buy 2x7800GT with even better ps3.0 performance

There is no situation in which I could possibly find X850 Crossfire to be of any "use" as you put it.
 
Well
some flaws


Who says you will seel you x8*** card and motherboard at that with all the new tosy out :(
 
I'm not sure ATI was really planning on selling the x800 series with crossfire. I'd rather look at the xfire setup for scaling and features than actual performance at this time. I'm not sure anyone would even want to consider x800/850 based crossfire setups due to release times and associated costs.

R520 should have dual DVI from what i've read so the 1600x1200@60Hz should no longer be a factor. Now assuming a x800 = 6800 how well do their respective SLI/Xfire setups scale when that second card is added. If for example xfire gives a 60% boost and SLI gives a 40% boost you could see a large margin between the x1800 and 7800 SLI setups. Especially if the x1800s performs greater than or equal to the 7800s.

The biggest handicap I've seen so far is that AA performance when using multiple cards. I'm guessing it might be related to the bandwidth connecting the two cards but there's no real way of knowing. When the bandwidth increases for R520 it could very well minimize the problem. Otherwise the compositing engines they're putting on the master cards need to be upgraded to something a little more powerful.

Running 16xAA? + HDR(if it's even possible) on one of the yet to be released crossfire setups could produce a result that NV will have a hard time matching until a AA+HDR capable card is released.

As for the 512MB of ram it sounds like Vista is gonna chew up a ton of video ram so releasing even standard cards with 512mb could help out a lot. Sort of future proof the cards to a degree. Years down the road they might not be able to game well but at least basic windows performance would be reasonable.
 
You must log in or register to reply here.
Back
Top