Ati Back on top again

[apoppin]

Originally posted by: McCartney
Anandtech is using canned benchmarks. Obviously when put in realworld gaming situations we see it's a different story. It's not the winner anymroe man learn to do some searching

i "get it" ... and i got it without "searching"

according to *you*

when AMD 'wins' - they are using "canned benchmarks"

when nvidia 'wins' - they are using "real-world benchmarks"
:roll:


and it appears that *all* the trusted HW sites - except for HardFUD - use "canned benchmarks"
:Q

ANYWAY, X2 is no r300 ... otoh, i'd say G80 is another r300

 

[Gikaseixas]

It's one card, one slot , best from ATI vs best from Nvidia so it competes directly with the 8800 GTX Ultra. I'm getting two GT's 512 in a month or so because that's the easiest way for me, that's what makes sense for me, if i had a non-SLI mobo i would get the new ATI for sure.

 

[Blacklash]

Taken from techPowerUp!-

http://img233.imageshack.us/im...8076/perfdollarts2.gif

The 8800GT 512Mb, HD3850 256Mb and 8800GTS 512Mb are better than the HD 3870 X2 "price vs performance". For most people the 8800GT 512Mb and 8800GTS 512Mb are definitely going to be the best choice. The HD 3870 X2 is certainly viable for folks on a 30" screen, however you can't always count on it.

I have yet to see a game where a 239usd 8800GT 512Mb couldn't put up perfectly playable FPS @ 1680x 4xAA|16xAF. Yes "Crysis" all very high will beat it down, and there isn't another graphics card out that can endure that well either.

 

[chinaman1472]

Uh, so what's the big deal? ATi hasn't churned out anything decent in a LONG time. It was bound to happen that they would release worthwhile products in the next generation.

The question, is how long can they hold it?

 

[Genx87]

Ati Back on top again
like the 9700pro good ole days

Not even close. This card is nice due to being a single slot solution. But when pitted against two 8800GTs it fails. If Nvidia didnt have a card coming down the pipe they would slap together a 2xGPU card by this summer and soundly whip this card into submission.

I guess it depends on what the next generation cards from Nvidia do. My guess is itshould be soundly whip this card.

 

[nemesismk2]

Originally posted by: Rollo
My thought is the 8800GTX is the only "9700Pro-esque" card since the 9700Pro. The 9700Pro had longevity, this card will be the fastest single slot card until the 9800GX2 launches (soon) and be trumped again in Q2 by the next gen cards if you can believe the rumors.

Being the fastest single slot card at some games for a month or two is not "9700 Pro like", IMHO.

Welcome back

 

[FormulaRedline]

I have to agree with nanaki333 about people being "fanboys." I see a lot of it on any hardware forum (even non-computer related hardware). There is no need to be so defensive about what you have. I just bought a G92 8800GTS, but I can still recognize that AMD has posted up some sweet numbers with this new card. Let's give them credit where credit is due. Could nVidia make a better card? Given what we all know about the G92 GPUs, they most likely could, but the important thing is they have not released such a card.

That being said, since we are all extreme hardware users here, I would like to see some overclocking results on these cards. If the 3870x2 is already hot and power hungry, can it really be pushed that much further? How would that compare to other OC'ed single card solutions? Anybody know if review site has run tests like these yet?

 

[McCartney]

Originally posted by: apoppin

Originally posted by: McCartney
Anandtech is using canned benchmarks. Obviously when put in realworld gaming situations we see it's a different story. It's not the winner anymroe man learn to do some searching

i "get it" ... and i got it without "searching"

according to *you*

when AMD 'wins' - they are using "canned benchmarks"

when nvidia 'wins' - they are using "real-world benchmarks"
:roll:


and it appears that *all* the trusted HW sites - except for HardFUD - use "canned benchmarks"
:Q

ANYWAY, X2 is no r300 ... otoh, i'd say G80 is another r300

Did I say that? No I didn't. I may be pro nVidia but clearly there's something wrong when the enthusiast reviewing community is being lazy and just using canned benchmarks. nVidia showed how easy it was to manipulate futuremark's 3dmark03 by making special drivers to pwn the 9700 in it even though they were equal cards.

You tell me, is it easier to make a card that will have the reviewers praising since they test with predetermined sequences or is it easier to test it in dynamic environments?

What I"m asking is: IS a test easier if you know the questions or if you don't?

If you think it's easier when you know the questions look at the 3870x2, and then look at all the reviews that use these benchmarks that are canned (meaning they are in-game but do not test dynamic conditions, they're very specific). Only people that hadn't canned the benchmarks were HardOCP and I'm not trying to start a forum war (unlike some people who came on the H forum and started a thread) I'm just simply stating the facts apoppin and you should expect them.

And don't fucking insult HardOCP. Last time I checked their whole staff (Inclusive of kyle) talk to us enthusiasts... Unlike Anand who can't even help people from his own demographic when they ask him questions bout his job (I'm East Indian). That's how good of a person Anand is. Yeah he's busy, so is Kyle. You have to realize ATT and H are the biggest sites out there... Act your age Apoppin, what are you? 45-50? It's about time you stop being a child and insulting a review that uses real conditions (which costs MORE money, takes MORE time) instead of canned benchmarks.

Why would Kyle have a tough time recommending a card if he spent money/time reviewing it? Seriously think about it.... Oh wait nevermind I saw your 29 000 posts in 8 years. That's roughly 3.6ish (estimation) thousand a year, and about 10 a day. I think this is how you do it. Am I right?

Get off the personal insult road McCartney. If you can't just talk about the tech in a civil way, don't talk at all.

Anandtech Moderator - Keysplayr2003

 

[Tempered81]

Originally posted by: Genx87
Not even close. .... If Nvidia didnt have a card coming down the pipe they would slap together a 2xGPU card by this summer and soundly whip this card into submission.

i think they do have a 2xgpu card coming.

 

[krnmastersgt]

Originally posted by: jaredpace

Originally posted by: Genx87
Not even close. .... If Nvidia didnt have a card coming down the pipe they would slap together a 2xGPU card by this summer and soundly whip this card into submission.

i think they do have a 2xgpu card coming.

That's what the 9800X2 is, nVidia's take on the 2 GPUs per card. There's been quite a few threads in this section already on it so I'm surprised people still don't know about it, I think it looks freaking awesome . Can't wait to get my hands on some, or the high-end 9 Series card that's just a single GPU.

 

[Stoneburner]

Originally posted by: McCartney

Originally posted by: apoppin

Originally posted by: McCartney
Anandtech is using canned benchmarks. Obviously when put in realworld gaming situations we see it's a different story. It's not the winner anymroe man learn to do some searching

i "get it" ... and i got it without "searching"

according to *you*

when AMD 'wins' - they are using "canned benchmarks"

when nvidia 'wins' - they are using "real-world benchmarks"
:roll:


and it appears that *all* the trusted HW sites - except for HardFUD - use "canned benchmarks"
:Q

ANYWAY, X2 is no r300 ... otoh, i'd say G80 is another r300

Did I say that? No I didn't. I may be pro nVidia but clearly there's something wrong when the enthusiast reviewing community is being lazy and just using canned benchmarks. nVidia showed how easy it was to manipulate futuremark's 3dmark03 by making special drivers to pwn the 9700 in it even though they were equal cards.

You tell me, is it easier to make a card that will have the reviewers praising since they test with predetermined sequences or is it easier to test it in dynamic environments?

What I"m asking is: IS a test easier if you know the questions or if you don't?

If you think it's easier when you know the questions look at the 3870x2, and then look at all the reviews that use these benchmarks that are canned (meaning they are in-game but do not test dynamic conditions, they're very specific). Only people that hadn't canned the benchmarks were HardOCP and I'm not trying to start a forum war (unlike some people who came on the H forum and started a thread) I'm just simply stating the facts apoppin and you should expect them.

And don't fucking insult HardOCP. Last time I checked their whole staff (Inclusive of kyle) talk to us enthusiasts... Unlike Anand who can't even help people from his own demographic when they ask him questions bout his job (I'm East Indian). That's how good of a person Anand is. Yeah he's busy, so is Kyle. You have to realize ATT and H are the biggest sites out there... Act your age Apoppin, what are you? 45-50? It's about time you stop being a child and insulting a review that uses real conditions (which costs MORE money, takes MORE time) instead of canned benchmarks.

Why would Kyle have a tough time recommending a card if he spent money/time reviewing it? Seriously think about it.... Oh wait nevermind I saw your 29 000 posts in 8 years. That's roughly 3.6ish (estimation) thousand a year, and about 10 a day. I think this is how you do it. Am I right?

You hate anand because he didn't help you get a job and you expected help because you are indian?

WTF?

 

[McCartney]

No. I didn't expect a job at all who told you that? I had questions about his CAREER. Please read. I don't believe in nepotism at all nor do I practice it. I just think that having an inside look at the industry would be a great thing.
Stop trying to use extrapolation because obviously you weren't very good at logical reasoning.

 

[LOUISSSSS]

i'm pretty sure the tests done at anandtech for GPU's are the same whether they are testing ATI or nvidia gpu's. i didn't read the review yet but i'm pretty sure the people conducting the tests on the 3870 x2 didn't do anything that they wouldn't have done for a nvidia gpu test.

so when nvidia cards win Anandtech's GPU performance tests, the tests are fair and Nvidia is God and when ATI wins GPU performance tests, they cheated. HA..

and please stop crying because anand didn't want to respond to your questions. nobody here needs to answer your career questions.

 

[Ackmed]

This is nothing like the 9700 Pro, "good ol days". There are way too many differences to even be mentioned as close.

 

[Tempered81]

yes agreed i should change topic to x1900xtx "good ol days"

 

[Ackmed]

This release much more not in common, than in common with the 9700 Pro launch.

 

[apoppin]

Originally posted by: McCartney

Originally posted by: apoppin

Originally posted by: McCartney
Anandtech is using canned benchmarks. Obviously when put in realworld gaming situations we see it's a different story. It's not the winner anymroe man learn to do some searching

i "get it" ... and i got it without "searching"

according to *you*

when AMD 'wins' - they are using "canned benchmarks"

when nvidia 'wins' - they are using "real-world benchmarks"
:roll:


and it appears that *all* the trusted HW sites - except for HardFUD - use "canned benchmarks"
:Q

ANYWAY, X2 is no r300 ... otoh, i'd say G80 is another r300

Did I say that? No I didn't. I may be pro nVidia but clearly there's something wrong when the enthusiast reviewing community is being lazy and just using canned benchmarks. nVidia showed how easy it was to manipulate futuremark's 3dmark03 by making special drivers to pwn the 9700 in it even though they were equal cards.

You tell me, is it easier to make a card that will have the reviewers praising since they test with predetermined sequences or is it easier to test it in dynamic environments?

What I"m asking is: IS a test easier if you know the questions or if you don't?

If you think it's easier when you know the questions look at the 3870x2, and then look at all the reviews that use these benchmarks that are canned (meaning they are in-game but do not test dynamic conditions, they're very specific). Only people that hadn't canned the benchmarks were HardOCP and I'm not trying to start a forum war (unlike some people who came on the H forum and started a thread) I'm just simply stating the facts apoppin and you should expect them.

And don't fucking insult HardOCP. Last time I checked their whole staff (Inclusive of kyle) talk to us enthusiasts... Unlike Anand who can't even help people from his own demographic when they ask him questions bout his job (I'm East Indian). That's how good of a person Anand is. Yeah he's busy, so is Kyle. You have to realize ATT and H are the biggest sites out there... Act your age Apoppin, what are you? 45-50? It's about time you stop being a child and insulting a review that uses real conditions (which costs MORE money, takes MORE time) instead of canned benchmarks.

Why would Kyle have a tough time recommending a card if he spent money/time reviewing it? Seriously think about it.... Oh wait nevermind I saw your 29 000 posts in 8 years. That's roughly 3.6ish (estimation) thousand a year, and about 10 a day. I think this is how you do it. Am I right?

You may be pro nVidia? ... that is the understatement of this year ... so far
... Ten a day ... that is the only thing you are correct about in your rather nasty reply. You also appear to be acting your "age" - juvenile.

Kyle *preaches* to enthusiasts with the nvidia Bible in his hand. He is also quick to excommunicate anyone who dares speak profanely against his god. You can have that fan-site; at least we can speak freely here as long as we refrain from personal attack. The rest of the tech world rightly makes fun of HardOCP's silly and skewed testing.

i am also quite sure it takes more time to *manipulate* settings to get their predetermined desired results. There is a reason the *rest* of the tech world uses tried-and-true benchmarks. They may be imperfect but they are less prone to manipulation and do not depend on what a reviewer "thinks" is playable.

and i do *both* kinds of testing ... at least i can give you MY benchmark and YOU can run it to see if i was accurate or not
clearly, making a demo run takes more effort than using the "official" one or one that is acknowledged as *good* and *representative* of the game by reviewers. Not so strangely, Kyle makes it *impossible* to replicate what he does in his testing. And also less strangely, he is OUT of line with impartial and *trusted* tech sites. Kyle runs a nvidia fansite [period]

 

[LOUISSSSS]

u guys should all see this guys McCartney's Nvidia Fanboy crying over at <<H. its amazing haha.

http://www.hardforum.com/showt...d.php?t=1266600&page=3

what an Nvidiot. too bad u don't get paid for it haha. i wouldn't hire u hire, u suck at being a fanboy

That would be quite enough LOUISSSSS. No need to throw gas on a spark.

Anandtech Moderator - Keysplayr2003

 

[Extelleron]

Nothing like the 9700 Pro. Were the HD 3870 X2 priced at $399~ or so, then it would be a good deal. But at $450, it's simply not worth it when you can get 2x 8800GT for a little bit more, or better yet just buy a single 8800GTS 512MB for $300~ or less and wait it out until the next generation cards.

ATI was more ahead with the X1900 generation then they are now in single card performance. It may not have been quite so clear cut then, but from today's perspective, the X1900XTX is to the 7900GTX as the 9800 series was to the GeForce 5's. Hopefully R700 will do this again, and once more ATI will be the undisputed graphics leader.

 

[apoppin]

Originally posted by: LOUISSSSS
u guys should all see this guys McCartney's Nvidia Fanboy crying over at <<<<. its amazing haha.

http://www.hardforum.com/showt...d.php?t=1266600&page=3

what an Nvidiot. too bad u don't get paid for it haha. i wouldn't hire u hire, u suck at being a fanboy

Lets see here:

1. If it wasn't for nVidia we'd be enjoying *possibly* AGP 8x RAGE3D 32mb cards
2. If it wasn't for nVidia we'd care more about a TV tuner on our video card over things we care about (fillrate, theoretical bandwidth, what kind of ram it uses, pixel pipeline-- all that good stuff)
3. If it wasn't for nVidia we'd be in the f@cking graphical stone-age.

Open your eyes people, fanboy or not, there have been fanboys since the 9700 pro came out. ATI won ONE f@cking round, ONE! And look, there are ATI fanboys who claim ATI knew what they were doing when they made the 9700-- Like hell they did, it was because of nVidia's poor drivers (Thanks Futuremark, jerks) that nVidia even lost that round.

Kyle? ... Kyle ... is that really how you feel? Now we know why the benchmarks all favor nvidia - he tolerates this crap^^^
- and nvidia is just a graphics company - a corporation out to make money - not the savior of the G-D planet



i'll be taking his posts really seriously
:roll:

 

[Red Storm]

3870 X2 under $399.

Not everyone has an nVidia motherboard, so SLI is not an option for a lot of people. Single card solutions however, are.

 

[Keysplayr]

Now THAT is aggressive pricing. Nice.

 

[dileepsv]

so..the 3870 X2 in a non-crossfire motherboard would work more efficiently and cool in comparison to 2x 3870 cards in CF?

 

[Killrose]

The only good things about the 3870x2 are that it gives the AMD/ATi camp something to throw into their box that gives them more performance and that it doesnt require a CF motherboard to do it. I'd give it at least 2 more drivers before we really get an idea of its true performance potential, but I doubt we'll see double digit percentage improvements in the real world over what we are seeing now.

Other than that it is no where the industry exciting card the 9700pro was. The 8800GT with its excellent price performance ratio has taken that crown.

 

[Tempered81]

Originally posted by: Red Storm
3870 X2 under $399.

Not everyone has an nVidia motherboard, so SLI is not an option for a lot of people. Single card solutions however, are.

HOLY!