TheGreatGrapeApe
Champion
- Feb 18, 2003
- 20,043
- 0
- 50,960
There a few things that need to be cleaned up in the description if it's to remain a sticky.
You need to change the wording of the description of Xfire, you make it sound like ATi copied nV when if anything it's nV who copied ATi and Metabyte's methods of doing multi-card graphics. nV's SLi bears no resemblance to 3Dfx's SLI in anything other than name. And SLi didn't come first in the rebirth, that was Alienware's failed ALX and XGI's volari V8 Duo. And ATi was doing their SIMFusion thing the whole time with Evans & Sutherland.
Also you need to split the solutions up, since your first example of the same cards applies to SLi but not Xfire. So it should be clearly defined which one you're talking about, you examples lets someone guess, but it needs to be explicit.
And in the memory example following it, you talk about memory and then talk about speeds. The 7800GTX is a bad example for memroy since there is athat large difference. But when mentioning memory and talking about size you should mention the size being affected and that they will disable the added memory. You're sort of mixing your ideas here. Then right after that you say you can mix/mathc brands but make sure they have the same memroy and clocks, yet right before that you imply/state they can have different memory/clocks.
For an Xfire section you should add that while the cards need to be the same underlying architecture, they can have different numbers of shaders or pipelines and still work together, like an X1900GT + X1900CF combination.
Also when talking about multi GPU benefits, rememebr to mention that one of the areas of greatest impact is also in the minimum fps, where it will usually increase min fps and smooth out dips into the lower numbers. It's an important factor that's usually missed in the discussion and is the area that would be most noticeable to people not looking simply to push large resolutions or high settings.
And when talking about does it always improve performance, you should also mention that sometimes it actually hurts performance when using the additional card.
When talking about the memory doubling myth you may wish to expand on how the memory is used and why it doesn't 'double the memory' when in Crossfire/SLi mode.
When talking about 2 midrange beating 1 high end, the thing to rememebr is that there are exception to that, like 2 GF7800GTs beating 1 GF7800GTX-512 for less money, or 2 GF8800GTS beating a single GTX or Ultra for the same money, but it depends on the situation, because there's no clear cut rule for that.
I'd get rid of the 'which is better' comment, you could discuss that with intelligent people night and day and never get and answer, and posting to any one benchmark doesn't solve that question either. So I'd pretty much avoid it unless you take the time to explore it further than a link to a single group of tests.
Also I'd avoid posting a list of cards, unless you can keep it up to date and include all the sub cards, better to just link to AMD & nV's list thus the onus is on them and not you.
BTW, what the Crossfire/SLi RAMs sections? Sounds like you're talking about the 'certified' parts. But the way its written sounds like VRAM.
If you're mentioning 'Crossfire/SLi editions' etc. you might want to mention that for AMD the boxes stamped 'Crossfire Edition' come with the ribbon connectors required to run cards in Xfire.
There's also a few spelling mistakes, but I can clean those up directly if you want.
You need to change the wording of the description of Xfire, you make it sound like ATi copied nV when if anything it's nV who copied ATi and Metabyte's methods of doing multi-card graphics. nV's SLi bears no resemblance to 3Dfx's SLI in anything other than name. And SLi didn't come first in the rebirth, that was Alienware's failed ALX and XGI's volari V8 Duo. And ATi was doing their SIMFusion thing the whole time with Evans & Sutherland.
Also you need to split the solutions up, since your first example of the same cards applies to SLi but not Xfire. So it should be clearly defined which one you're talking about, you examples lets someone guess, but it needs to be explicit.
And in the memory example following it, you talk about memory and then talk about speeds. The 7800GTX is a bad example for memroy since there is athat large difference. But when mentioning memory and talking about size you should mention the size being affected and that they will disable the added memory. You're sort of mixing your ideas here. Then right after that you say you can mix/mathc brands but make sure they have the same memroy and clocks, yet right before that you imply/state they can have different memory/clocks.
For an Xfire section you should add that while the cards need to be the same underlying architecture, they can have different numbers of shaders or pipelines and still work together, like an X1900GT + X1900CF combination.
Also when talking about multi GPU benefits, rememebr to mention that one of the areas of greatest impact is also in the minimum fps, where it will usually increase min fps and smooth out dips into the lower numbers. It's an important factor that's usually missed in the discussion and is the area that would be most noticeable to people not looking simply to push large resolutions or high settings.
And when talking about does it always improve performance, you should also mention that sometimes it actually hurts performance when using the additional card.
When talking about the memory doubling myth you may wish to expand on how the memory is used and why it doesn't 'double the memory' when in Crossfire/SLi mode.
When talking about 2 midrange beating 1 high end, the thing to rememebr is that there are exception to that, like 2 GF7800GTs beating 1 GF7800GTX-512 for less money, or 2 GF8800GTS beating a single GTX or Ultra for the same money, but it depends on the situation, because there's no clear cut rule for that.
I'd get rid of the 'which is better' comment, you could discuss that with intelligent people night and day and never get and answer, and posting to any one benchmark doesn't solve that question either. So I'd pretty much avoid it unless you take the time to explore it further than a link to a single group of tests.
Also I'd avoid posting a list of cards, unless you can keep it up to date and include all the sub cards, better to just link to AMD & nV's list thus the onus is on them and not you.
BTW, what the Crossfire/SLi RAMs sections? Sounds like you're talking about the 'certified' parts. But the way its written sounds like VRAM.
If you're mentioning 'Crossfire/SLi editions' etc. you might want to mention that for AMD the boxes stamped 'Crossfire Edition' come with the ribbon connectors required to run cards in Xfire.
There's also a few spelling mistakes, but I can clean those up directly if you want.
( u always have helped me 
Facebook
Twitter
Instagram