Is SLI worth it?

ryane24

Member
I was thinking of getting two 8800 GT 512 graphics cards and running them SLI. That would cost about 550 dollars or so. Is it really worth it? Couldnt you get one card that could do the same for that price?
 
save the extra money and buy something else you may want

-ipod
-GPS for your car
-a gun
-new tires
-tools
-video game

whatever, SLI is a pure waste of money. It doesn't actually really increase performance all that much. Research parallel processing because that is what SLI is. Now, if SLI was distributed processing we would be having a different conversation right now.
 
2x the price = 2x the performance?

^It doesn't work that way... You might end up with 50% more performance, you will not double your performance though... It's not worth it...
 
You don't really get a performance increase either, its just more bandwidth. You may raise your FPS but nothing else, all other number crunching done by the CPU and all memory paging done by the RAM gets no benefit.

Those benchmarks are misleading when talking about real world performance.
 
You don't really get a performance increase either, its just more bandwidth. You may raise your FPS but nothing else, all other number crunching done by the CPU and all memory paging done by the RAM gets no benefit.

Those benchmarks are misleading when talking about real world performance.

More FPS... isn't that what you want?
 
More FPS... isn't that what you want?

You do. Or at least I do. Increased bandwidth is a result of sli, but it not the only distinctive benefit. The second graphics card will provide extra muscle, though it's sloppily executed. SLI problems mainly exist because of the code driving it, in other words the drivers, and the code it is being implemented upon, in other words game code. Consequently, there is future hope for the platform, or other platforms like it.
 
what about crossfire? i have seen two 3870's beat an nvidia 8800ultra, it must be worth a shot then, no?
 
I beg to differ with quite a lot people here to be honest. I used 2 7800GTX's in SLI quite a while, and I certainly dont regret that i bought those.
the only reason why i switched to an Ati card now was the power consumption (wich of the new nvidia cards is kinda retarded imo), and because i wanted to give Crossfire a shot too, see how that is ^^

anyways, if you have the money to spare, and you can get the 2nd card cheaper, if you buy them at a store at the same time for example, you can always ask 'if I buy these 2, is there soemthing you can do on the price?' . asking doesnt hurt, now does it? its how I got my xbox360 + wireless expansion kit + 4 games 60 euro's cheaper, and with 4 games instead of the 3 I went there for ^_^ )
 
I beg to differ with quite a lot people here to be honest. I used 2 7800GTX's in SLI quite a while, and I certainly dont regret that i bought those.
the only reason why i switched to an Ati card now was the power consumption (wich of the new nvidia cards is kinda retarded imo), and because i wanted to give Crossfire a shot too, see how that is ^^

anyways, if you have the money to spare, and you can get the 2nd card cheaper, if you buy them at a store at the same time for example, you can always ask 'if I buy these 2, is there soemthing you can do on the price?' . asking doesnt hurt, now does it? its how I got my xbox360 + wireless expansion kit + 4 games 60 euro's cheaper, and with 4 games instead of the 3 I went there for ^_^ )

but see now archangel you have the girl factor on your side lol. guys can't flirt with the manager for a deal... :mad:

jkjk
 
Theres no denying that games now can take advantage of dual-GPU cards, looking at real game benchmarks of two cards SLI'd makes a difference of anywhere from 50-80%.
 
if its PCI-E 2.0 its worth it, i think...especially with dual 3870's.
imagine if there were a 3870x2 PCI-E 2.0....those in crossfire would be amazing :drool:
 
if its PCI-E 2.0 its worth it, i think...especially with dual 3870's.
imagine if there were a 3870x2 PCI-E 2.0....those in crossfire would be amazing :drool:

you know what? that brings up a very strong point. up untill now, none of the pci-e 2.0 cards have fully utilized the increase in bandwidth, i wonder if this card will finally embrace the higher bandwidth?
 
OK, look I will try to explain this one more time. SLI does increase performance. It puts two GPUs in what is known as parallel processing mode, which means they are synchronized and work together at the same time. Thus increasing the bandwidth that can go through a card. However, when you play a game the video card does not carry all the weight. The motherboard, processor and RAM still have a huge play and the hard drive as well.

When you benchmark a system it runs a very specific instruction set to accomplish one simple goal. That is where you see the benchmark results go higher in certain categories. What you should be comparing is actual game play FPS and watch the system monitor (task monitor) to see how many resources are being used with game play and how your system handles it. Sure, 3Dmark may have a 50% increase in benchmarking your FPS with SLI or xFire, but would UT3 run at a full 50% increase? I doubt it.

Benchmarks do not reflect real world performance, please remember that. I have said this many times before in the past. I used to work for an OEM systems builder a few years back, and we built some of the first SLI systems. Sure they benched faster, but comparing one single high end card to two high end cards in SLI mode in actual game play, you could not really tell the difference.

Yes, it does boost some performance but is it worth it? that is a choice is up to you. The performance increase in game is not 50 to 80% more, only in benchmarks.
 
yeah i can see how you would be right, but i didnt think that benchmarks ran like a movie, instead i thought they were rendered like a real game would be, giving accurate results...
 
yeah i can see how you would be right, but i didnt think that benchmarks ran like a movie, instead i thought they were rendered like a real game would be, giving accurate results...

Nope its all pre rendered, with the exception of the meta stuff that is done in benchmarks . Plus every consumer should know benchmarks never reflect real world performance in any thing. Just because you buy a new car that was benched from 0 to 60 in 4.5 seconds doesn't mean you are going to get that performance every time. Depends on what configuration, what gas you use, the weather, the driver, the road, the tires, so on and so forth. Just like everything else benchmarks are used for marketing a product.

I hate to use car analogies for computers but that one actually kind of fits.
 
Back
Top