Firing Squad Bioshock Benchmark/Review

kind of a short review with no 8600gt/gts into the mix.

1950pro is really showing it's superiority over 7900gs in newer games.
 
kind of a short review with no 8600gt/gts into the mix.

1950pro is really showing it's superiority over 7900gs.

Yeah was hoping they would have the 8600 since the mentioned in early on.

And it is FiringSquad not Squid :p
 
I hate the fact they don't include 1280x1024 resolution for medium range people. Not everyone is playing 1600x1200 or higher Firingsquid.
 
I hate the fact they don't include 1280x1024 resolution for medium range people. Not everyone is playing 1600x1200 or higher Firingsquid.

Yeah, running a 7900 or 8600 or something at 16xx doesn't really tell you much since most people don't run these cards at that resolution.
 
Wow that review shows that there may be something going on for DX10 after all. All of the benches on each card showed a consistent performance equal or better in DX10 vs. DX9 which is what was promised by DX10 in the first place. Who knows, maybe DX10 will actually start to be a good thing :p.
 
Wow that review shows that there may be something going on for DX10 after all. All of the benches on each card showed a consistent performance equal or better in DX10 vs. DX9 which is what was promised by DX10 in the first place. Who knows, maybe DX10 will actually start to be a good thing :p.

Keep in mind a lot of the stuff they promised for DX10 didn't make the cut. As you can see from the article, there isn't really much of a difference visually, so the cost in performance(if there is a hit) for the added effects isn't present.

World in Conflict should be different since there is some pretty noticeable differences, and it seems they didn't tack on DX10 support.
 
the 2900xt 1gb lost to the 8800gts 320mb...

surprise!

:eek:

With fewer stream processors and less onboard memory, the GeForce 8800 GTS 640MB and 320MB trail the GTX by over 25%, but still deliver very playable frame rates in the game up to 1920x1200. At 2560x1600 the GeForce 8800 GTX and Ultra pull even further away from the GTS cards, running about 1.5 times faster in our testing. AMD’s Radeon HD 2900 XT trails NVIDIA’s GeForce 8800 GTS 320MB by over 20% at 1600x1200, but as the resolution increases the gap slowly closes until the 2900 XT is practically even with the GeForce 8800 GTS 320MB by 2560x1600. Considering that the GeForce 8800 GTS 320MB has less memory and a narrower memory interface, not to mention fewer stream processors, this has got to be disappointing to Radeon enthusiasts, but we wouldn’t be surprised if AMD has more driver tweaks in the works to improve BioShock’s performance. As it stands now though the Radeon HD 2900 XT isn’t that much faster than an X1950 Pro running under DX9 with the same graphics settings.

why must we always wait around for amd/ati to release new drivers to improve performance in new games?
 
the 2900xt 1gb lost to the 8800gts 320mb...

surprise!

:eek:



why must we always wait around for amd/ati to release new drivers to improve performance in new games?

Yeah that seems to be the only argument ati fans can come up with: just wait for better drivers. :rolleyes:

The 2900 is only ~30% faster than a x1950 pro. Propably slower or equal performance to a 1950 xtx. 8800 GTS 320 is offering equal performance compared to a 1gb 2900xt at 2560 x 1600! Now that just isn't even funny.
 
Yeah that seems to be the only argument ati fans can come up with: just wait for better drivers. :rolleyes:

The 2900 is only ~30% faster than a x1950 pro. Propably slower or equal performance to a 1950 xtx. 8800 GTS 320 is offering equal performance compared to a 1gb 2900xt at 2560 x 1600! Now that just isn't even funny.

Also Didn't Ati already release drivers to "improve performance", which they are using?
 
this game is fairly new. Give it some time.

They just put the hack in their drivers the other day.
 
give it some time?!?!

so ati users must wait 2-3 months for ati to release some decent drivers for the game.

why is it unplayable for you?

GTS vs XT
7fps @ 1600x1200
5fps diff. at 1920x1200

Seems okay to me although it needs some improvement.
 
why is it unplayable for you?

GTS vs XT
7fps @ 1600x1200
5fps diff. at 1920x1200

Seems okay to me although it needs some improvement.


i have a x1950xt. perfectly playable for me. just find it annoying that in most new games that come out we have to give ati 1-2 months to fix the drivers.
 
2900xt vs. 8800gts! It's on.

2900xt ram: 1gb vs. gts 320mb.
2900xt memory interface: 512bit vs. gts 320bit.
2900xt memory type: GDDR4 vs. gts GDDR3.
2900xt stream processors: 320 vs. gts 96.
2900xt clock speeds: 740/1000 vs. gts 500/800.

Ding Ding. And the winner is atiii...... no wait. nvidia? Sunuvabitch.
 
It was interesting to see not much performance difference between the 8800GTS 320 & 8800GTS 640 cards.
 
2900xt vs. 8800gts! It's on.

2900xt ram: 1gb vs. gts 320mb.
2900xt memory interface: 512bit vs. gts 320bit.
2900xt memory type: GDDR4 vs. gts GDDR3.
2900xt stream processors: 320 vs. gts 96.
2900xt clock speeds: 740/1000 vs. gts 500/800.

Ding Ding. And the winner is atiii...... no wait. nvidia? Sunuvabitch.

ROFL this makes post of the year hands down. But it really does show that architecture can make the whole world of difference.
 
2900xt vs. 8800gts! It's on.

2900xt ram: 1gb vs. gts 320mb.
2900xt memory interface: 512bit vs. gts 320bit.
2900xt memory type: GDDR4 vs. gts GDDR3.
2900xt stream processors: 320 vs. gts 96.
2900xt clock speeds: 740/1000 vs. gts 500/800.

Ding Ding. And the winner is atiii...... no wait. nvidia? Sunuvabitch.

LOL
 
I hate the fact they don't include 1280x1024 resolution for medium range people. Not everyone is playing 1600x1200 or higher Firingsquid.

Running cards at those lower resolutions tells you nothing, because you're bottlenecking yourself at that point. The idea is to isolate a single variable (the video card) as the performance one, not test a whole bunch of them and then end up with no idea which affected what aspects of the performance. Basic scientific method, for the win.
 
The lower resolutions would give those running older systems an idea of what to expect.

Later, other sites will bench the game I'm sure. FQ just wanted to be the first out the gate, so it's not as detailed as others will do.
 
Some cards kick ass without any AA AF, some cards continue to show great performance with AA AF while some just start sucking. Since I don't use AA/AF with my games, much less with new and demanding ones like BioShock, that article isn't much of a help.

Nor are most reviews, seen tons not running any test without AA/AF and as a result giving crap scores to the mid range new ATI cards, since they don't perform well with AA/AF.. which in turn doesn't mean jack to me since I don't use it. Argh.
 
2900xt vs. 8800gts! It's on.

2900xt ram: 1gb vs. gts 320mb.
2900xt memory interface: 512bit vs. gts 320bit.
2900xt memory type: GDDR4 vs. gts GDDR3.
2900xt stream processors: 320 vs. gts 96.
2900xt clock speeds: 740/1000 vs. gts 500/800.

Ding Ding. And the winner is atiii...... no wait. nvidia? Sunuvabitch.

loooooooooooooooooooooooooooooooooooooooooooooooooool
 
Those results are laughable. My 2900XT is actually getting much much higher frames than they reported. They definitely screwed up the 2900XT tests. My 2900XT is beating my overclocked GTS (648/1000) in Bioshock.

I call BS on that review. Actually I was having problems with the 2900XT not performing well in my rig in sig, then Gigabyte released an F5 bios last night for my motherboard and the 2900XT has had an awakening. I think the PCI-e might have been dropping to 1x, but only when using a 2900XT. They might be having a similar issue...
 
2900xt vs. 8800gts! It's on.

2900xt ram: 1gb vs. gts 320mb.
2900xt memory interface: 512bit vs. gts 320bit.
2900xt memory type: GDDR4 vs. gts GDDR3.
2900xt stream processors: 320 vs. gts 96.
2900xt clock speeds: 740/1000 vs. gts 500/800.

Ding Ding. And the winner is atiii...... no wait. nvidia? Sunuvabitch.

HAAAAAAAAAHAHAHAHA!!!!
 
I love the DX9 vs DX10 comparisons, like seriously cracking me up. Its one of those "FIND THE DIFFERENCE" types of screenshots. What a crock of shit, more of a reason to wait a while longer.
 
Running cards at those lower resolutions tells you nothing, because you're bottlenecking yourself at that point. The idea is to isolate a single variable (the video card) as the performance one, not test a whole bunch of them and then end up with no idea which affected what aspects of the performance. Basic scientific method, for the win.

it's bottlenecking a 1950pro and 7900gs?
 
another useless test based upon a cpu that hardly anyone has.

good going! ;)

I know that a small % difference occurs when you go past the 2.2ghz amd dual core mark when compared to a newer C2D, but one cant be reliant on this trend.

middle clocked cpu's with non overclocked ram should be tested as at least part of a benchmark review, if not, the only part of the benchmark review.
 
So the top of the line 1GB 2900XT gets hammered by the 320mb GTS....wow
 
Hmm dx10 has better particles, but almost everything else is near the same, but how ever I think that dx9 shadows looked more realistic then the dx10 shadows, cause in real life, shadows are not suppose to be as "Crisp" its actually quite blurred depending on the distance, and brightness of the light source.

But then I was so used to the "crisp shadow" sometimes I find real life shadows to be unrealistic :/
 
Considering this review is from "Ad" squad (they basicly eye fuck you with ads) I would take it with a large grain of salt . There reviews have always been iffy ..atleast since a few years ago. Not sure why its dropped off but I would rather see [H] put up some numbers then I will take it as truth.
 
another useless test based upon a cpu that hardly anyone has.

good going! ;)

I know that a small % difference occurs when you go past the 2.2ghz amd dual core mark when compared to a newer C2D, but one cant be reliant on this trend.

middle clocked cpu's with non overclocked ram should be tested as at least part of a benchmark review, if not, the only part of the benchmark review.

What? You make absolutely no sense.

The whole point of having a system that is a really high spec is to remove any possible CPU bottlenecks when benching performance. You WANT the bottleneck to strictly be the GPU so that you have the most accurate comparison possible. If you slow the processor down, it can become the bottleneck, and the GPU's will look closer together in speed than they actually are.

The point of these reviews is to see what the fastest GPU is, not to see what fps you should be getting with an exact system spec.
 
i have to imagine they did something wrong with those 2900XT results. i mean that's just laughable. maybe the drivers? but seriously, why can't ATi turn out drivers on par with Nvidia's?

i will say this makes my purchase of the GTS 320 feel a whole lot more justified, though :D 620/850 ftw.

i would like to see some results from another test (site) though.
 
i have to imagine they did something wrong with those 2900XT results. i mean that's just laughable. maybe the drivers? but seriously, why can't ATi turn out drivers on par with Nvidia's?

i will say this makes my purchase of the GTS 320 feel a whole lot more justified, though :D 620/850 ftw.

i would like to see some results from another test (site) though.

Well I know its not hardware, the architecture on the 2900 is not the easiest to code for, the 8800 is alot easier to code for just by looking at the architecture. Ati usually does bring out good drivers.
 
I hate the fact they don't include 1280x1024 resolution for medium range people. Not everyone is playing 1600x1200 or higher Firingsquid.

This is my biggest gripe with all the latest reviews. There are a lot of us out there with 19" 1280 x 1024 LCD monitors.
 
Running cards at those lower resolutions tells you nothing, because you're bottlenecking yourself at that point. The idea is to isolate a single variable (the video card) as the performance one, not test a whole bunch of them and then end up with no idea which affected what aspects of the performance. Basic scientific method, for the win.

It does tell you something... it tells you that there is little to gain by upgrading. What I want to know in a review is how a potential addition would change the performance of what I currently have.
 
Those results are laughable. My 2900XT is actually getting much much higher frames than they reported. They definitely screwed up the 2900XT tests. My 2900XT is beating my overclocked GTS (648/1000) in Bioshock.

I call BS on that review.

Well if your sig is correct then you have a faster CPU, 3.6Ghz vs 3Ghz used in the review. Besides, they didn't have global lighting on either (I didn't notice a difference with my rig either way on performance). I play at a slightly lower resolution than they did and I think their results are actually accurate (with vsync on). I mean if you want, you can post your own benchies and see what a faster CPU would do. Hopefully [H] will do a small review (or include it into their portfolio), that would show some people :eek:.
 
why must we always wait around for amd/ati to release new drivers to improve performance in new games?[/QUOTE]

Just a few years ago AMD was top of the food chain, I think they made a bad decision teaming up with ATI, they should have stayed independent.
 
Back
Top