I know you'll think I'm an nVidia fan but...

[linkgoron]

I think sites should've used A64 X2 4800+ in benches because I heard that the 80+ nvidia drivers use DC chips and the Nvidia cards get better performance with them. Most high-end users have DC chips, so benches with an X2 would've reflected the performance of the 7800 in high-end PCs better...
maybe I'm a fanboy, but that's how I feel.

 

[Kyanzes]

Problem is there is no NVIDIA driver available with DC support other than beta. AFAIK even the 8x.xx driver set released today is still beta.

Edit: it also wouldn't be fair until we get an ATI driver that is also DC aware, it's my own point of view ofc

 

[linkgoron]

Originally posted by: Kyanzes
Problem is there is no NVIDIA driver available with DC support other than beta. AFAIK even the 8x.xx driver set released today is still beta.

Edit: it also wouldn't be fair until we get an ATI driver that is also DC aware, it's my own point of view ofc

But from what I've seen sites were using the 81 drivers...

 

[imported_Rampage]

There is no fairness in business.

If ATI has no dual core drivers.. too bad so sad.. but if they get pummelled by Nvidia because they dont have them.. tough. Welcome to the real world where you either put up a good fight, get beat up and go home.

 

[Wreckage]

Originally posted by: Kyanzes
Problem is there is no NVIDIA driver available with DC support other than beta. AFAIK even the 8x.xx driver set released today is still beta.

Edit: it also wouldn't be fair until we get an ATI driver that is also DC aware, it's my own point of view ofc

Beta ATI drivers were used in most if not all the reviews. I think if you are going to review a high end card it should be on a high end (dual core) chip.

And it is not a matter of being fair, it's a matter of what's available.

 

[Rage187]

Whats stopping Ati from releasing DC drivers?

thats right, nothin.

 

[lifeguard1999]

Generally speaking, you want the fastest CPU available to remove as much CPU limitation as possible. The single core FX-57 is 2.8 GHzl; the single core FX-55 is 2.6 GHz; single core 4000+ is 2.4 GHz; the dual core 4800+ is 2.4 GHz. Personally though, I agree with you. I would like to have seen dual-core benchmarks. Maybe that is being saved for a follow-on review.

 

[Suspicious-Teach8788]

Originally posted by: Rage187
Whats stopping Ati from releasing DC drivers?

thats right, nothin.

Uhhh their mass delays stop them from releasing DC drivers =P

 

[Ackmed]

Originally posted by: lifeguard1999
Generally speaking, you want the fastest CPU available to remove as much CPU limitation as possible. The single core FX-57 is 2.8 GHzl; the single core FX-55 is 2.6 GHz; single core 4000+ is 2.4 GHz; the dual core 4800+ is 2.4 GHz. Personally though, I agree with you. I would like to have seen dual-core benchmarks. Maybe that is being saved for a follow-on review.

What he said, the FX-57 is the fastest CPU out, and should be used. Im sure there will be an article from AT, FS, or someone else on dual core CPU's.

Reviews should however use 2gigs of ram. 1gig is no longer good enough, for some newer games, and high settings. BF2 and Fear are two good examples of this. The game stutters with 1gig, and is much smoother with 2. Stuttering causes low frames, which show up in final numbers.

 

[blckgrffn]

Originally posted by: Ackmed

Originally posted by: lifeguard1999
Generally speaking, you want the fastest CPU available to remove as much CPU limitation as possible. The single core FX-57 is 2.8 GHzl; the single core FX-55 is 2.6 GHz; single core 4000+ is 2.4 GHz; the dual core 4800+ is 2.4 GHz. Personally though, I agree with you. I would like to have seen dual-core benchmarks. Maybe that is being saved for a follow-on review.

What he said, the FX-57 is the fastest CPU out, and should be used. Im sure there will be an article from AT, FS, or someone else on dual core CPU's.

Reviews should however use 2gigs of ram. 1gig is no longer good enough, for some newer games, and high settings. BF2 and Fear are two good examples of this. The game stutters with 1gig, and is much smoother with 2. Stuttering causes low frames, which show up in final numbers.

:thumbsup:

 

[linkgoron]

Originally posted by: blckgrffn

Originally posted by: Ackmed

Originally posted by: lifeguard1999
Generally speaking, you want the fastest CPU available to remove as much CPU limitation as possible. The single core FX-57 is 2.8 GHzl; the single core FX-55 is 2.6 GHz; single core 4000+ is 2.4 GHz; the dual core 4800+ is 2.4 GHz. Personally though, I agree with you. I would like to have seen dual-core benchmarks. Maybe that is being saved for a follow-on review.

What he said, the FX-57 is the fastest CPU out, and should be used. Im sure there will be an article from AT, FS, or someone else on dual core CPU's.

Reviews should however use 2gigs of ram. 1gig is no longer good enough, for some newer games, and high settings. BF2 and Fear are two good examples of this. The game stutters with 1gig, and is much smoother with 2. Stuttering causes low frames, which show up in final numbers.

:thumbsup:

Can't you OC a 4800+ to FX55 speeds (or 57?)

 

[fishbits]

I think sites should've used A64 X2 4800+ in benches because I heard that the 80+ nvidia drivers use DC chips and the Nvidia cards get better performance with them.

I disagree. The vast majority of gamers are running single-core rigs, so a single core comparison between the two cards would be the most informative. I sure as heck wouldn't be happy to read a review saying the 7800 got one set of stats, then not have an idea how it would actually perform on my single core system.

Honestly, I can't think of a reason for doing what you propose other than to set it up to make ATI look bad. You could have just as easily proposed a comparison between Nvidia cards alone with different drivers if it was simply the DC drivers you were interested in (and it is very interesting). When we allow "stunting" like that, each GPU maker can say "OK, let's do a new review with my new special drivers that I made just for this specific test case (specific game, specific CPU, etc)." It's just a bit too much in my book and begs for abuse. If ATI's current drivers don't get all the power they can out of dual core yet, hopefully they will in the near future. They've been taking it on the chin from me for their recent flubs, but I'm not looking to pile on while they're down.

 

[linkgoron]

Originally posted by: fishbits

I think sites should've used A64 X2 4800+ in benches because I heard that the 80+ nvidia drivers use DC chips and the Nvidia cards get better performance with them.

I disagree. The vast majority of gamers are running single-core rigs, so a single core comparison between the two cards would be the most informative. I sure as heck wouldn't be happy to read a review saying the 7800 got one set of stats, then not have an idea how it would actually perform on my single core system.

Honestly, I can't think of a reason for doing what you propose other than to set it up to make ATI look bad. You could have just as easily proposed a comparison between Nvidia cards alone with different drivers if it was simply the DC drivers you were interested in (and it is very interesting). When we allow "stunting" like that, each GPU maker can say "OK, let's do a new review with my new special drivers that I made just for this specific test case (specific game, specific CPU, etc)." It's just a bit too much in my book and begs for abuse. If ATI's current drivers don't get all the power they can out of dual core yet, hopefully they will in the near future. They've been taking it on the chin from me for their recent flubs, but I'm not looking to pile on while they're down.

Well the vast majority of "gamers" use 6600GTs or lower cards. so basically we shouldn't even review 7800GTX's should we?

 

[fishbits]

What on earth does that have to do with setting up graphics cards tests with newly-released drivers that depend upon having a type of CPU all but unheard of in the desktop world until recently? I never said we shouldn't test X2s or 7800s. Anyone can come up with "stunt benchmarks" to make a certain product look artificially better than the competition. I was giving you the benefit of the doubt with the last post, but since you want to be smarmy about it, I'll simply say that you can take your greasy fanboyism and cram it.

 

[Elcs]

Dual Cores arent that widely used in games thus far and driver support isnt exactly perfect yet.

So I personally think the fastest single core processor should be used until that situation changes. The difference between a DC 4800+ and FX-57 may be minor but every reasonable effort should be made to reduce CPU bottlenecking.