The structuring of these test are a little questionable and when you get down to comparable models the scores were pretty close (the Pentium D 830 held up pretty well against the X2 3800, etc...). I think it's pretty funny to give price to performance to AMD, but Cnet loves headlines.
I'd be all over AMD dual cores if I felt nVidia made better chipsets, having been burnt by the stutter bug on the nForce 3 and the issues with large file transfers on the nForce4 they are off my radar. My current Athlon 64 3200+ is doing okay on a Via KT890-based board but is not feature complete (no dual core support, sloppy southbridge performance) that will hopefully be resolved with the KT9xx series chipsets. That leaves Sis (no thanks) and ATi (poor southbridge implementation as well). All AMD really needs is a top of the line chipset and I would switch to them as a primary platform tomorrow, until then I'll use Intel for work and play with a little AMD for playtime as well.
I'll keep waiting for something to impress me HTT, I'm probably way to picky in this area. I've been managing large PC environments for years at work and have some pretty detailed criteria whent it comes to chipsets and stability.
Regarding the CNET article, this got me thinking...
I think there needs to be a new way to measure "performance" in the dual core era, the current "what goes fastest is best" doesn't make much sense any longer. I think at least for now the onboard memory controller on the K8 architecture will keep AMD in front on any benchmark, and if that is your interest then go for it. But what about actual usability? Is there really that much of a difference?
I play all my games with Vsync enabled, so I don't really care if I can get more than 60fps ever. I'm much more interested in the idea of getting a steady 60fps at a decent widescreen resolution, anything over that and I never see it. What can't I run at this kind of rate, CoD 2 and F.E.A.R., both of which are GPU and not CPU bound. So why do I care? Yes, with Vsync off I get some pretty high numbers, not the highest, but very respectable. But what does it get me when actually playing?
Editing is another interesting area to look at, since I was just doing some video editing and compiling this weekend I came to some conclusions. I was taking assorted holiday videos from other family members, ripping, cleaning up and putting into a "best of" montage. I spent a ton of time in Premiere Elements, PhotoShop and had WMP10 running in the background so I could pick music. After several hours I was pretty happy and decided it was time to compile and burn; did I kick it off and marvel at how fast my dual core was working? Nope, I kicked it off and went to work on some disk labels and cover art while it worked at 100% on the other core. Sure, it could have compiled faster with both cores pumping, but I didn't buy a dual core to win benchmarks, but to be productive.
How nice is it to be able to work in FrontPage 2003, PhotoShop and have other assorted apps open and never bog down, or have the sound stutter when something needs a lot of CPU, etc... I know some of this is anecdotal and very user specific, and I have no doubt that is someone devised a way to test all of this AMD would still be faster on paper, but I ask you; who cares? I have no doubt that Windows Vista will come along and make me want a faster CPU, and by then it could easily be an AMD based system, but until they come up with a better way to measure performance on dual cores I can only take articles like this so seriously.
CNET took the entire product line and treated it as relevant when really they should have taken a good look at the Pentium D 820, 830 and 840 with the X2 3800+, 4200+ and 4400+ ; the only honest contenders for 90% of the market. The 840EE and the 4600+/4800+ are fairly irrelevant.
frontpage 2003 doesn't hog the cpu as word when you get over 40 pages with diagrams lol so the word docs around 6mb, also excel when trying to import around 20000 lines of data lol thats a test
I would have to say that you are fairly heavily opinionated (IMO at least, lol). I have to admit that when someone such as yourself (an obvious candidate for "most likely never to be labeled an enthusiast user") writes such an intricate and detailed post based on a benchmarking article for enthusiasts... I tend to think "what a knob".
Someone who enables vsync (lmfao) and just wants one core rendering multimedia is not a power user. Thats the same as a racing driver that never goes over 60kph/mph and never rev's over 3000 (rofl).
The other thing I have to say (I hate reading forums sometimes, purely because there are too many people to flame) is that Intel is being hammered for their dodgy chipsets, if you cannot find an Nforce4 that works (since it is now working great for Intel) go buy a nice DULL Celeron based laptop and play minesweeper (marvel at your 60fps).
klbf
2 Intern
•
2.7K Posts
0
November 28th, 2005 22:00
also the intel 820 D was the average cpu which the other 3 were based on
Message Edited by klbf on 11-29-2005 12:50 PM
ravik521
2 Intern
•
1.6K Posts
0
November 28th, 2005 22:00
bob_c_b
556 Posts
0
November 28th, 2005 22:00
The structuring of these test are a little questionable and when you get down to comparable models the scores were pretty close (the Pentium D 830 held up pretty well against the X2 3800, etc...). I think it's pretty funny to give price to performance to AMD, but Cnet loves headlines.
I'd be all over AMD dual cores if I felt nVidia made better chipsets, having been burnt by the stutter bug on the nForce 3 and the issues with large file transfers on the nForce4 they are off my radar. My current Athlon 64 3200+ is doing okay on a Via KT890-based board but is not feature complete (no dual core support, sloppy southbridge performance) that will hopefully be resolved with the KT9xx series chipsets. That leaves Sis (no thanks) and ATi (poor southbridge implementation as well). All AMD really needs is a top of the line chipset and I would switch to them as a primary platform tomorrow, until then I'll use Intel for work and play with a little AMD for playtime as well.
HTT Good HT Bad
2 Intern
•
1.3K Posts
0
November 28th, 2005 23:00
Bob what about the ALi/Uli chipsets, they seem pretty decent
Message Edited by HTT Good HT Bad on 11-28-2005 07:22 PM
klbf
2 Intern
•
2.7K Posts
0
November 28th, 2005 23:00
klbf
2 Intern
•
2.7K Posts
0
November 28th, 2005 23:00
herb366
1.5K Posts
0
November 28th, 2005 23:00
bob_c_b
556 Posts
0
November 29th, 2005 00:00
I'll keep waiting for something to impress me HTT, I'm probably way to picky in this area. I've been managing large PC environments for years at work and have some pretty detailed criteria whent it comes to chipsets and stability.
Regarding the CNET article, this got me thinking...
I think there needs to be a new way to measure "performance" in the dual core era, the current "what goes fastest is best" doesn't make much sense any longer. I think at least for now the onboard memory controller on the K8 architecture will keep AMD in front on any benchmark, and if that is your interest then go for it. But what about actual usability? Is there really that much of a difference?
I play all my games with Vsync enabled, so I don't really care if I can get more than 60fps ever. I'm much more interested in the idea of getting a steady 60fps at a decent widescreen resolution, anything over that and I never see it. What can't I run at this kind of rate, CoD 2 and F.E.A.R., both of which are GPU and not CPU bound. So why do I care? Yes, with Vsync off I get some pretty high numbers, not the highest, but very respectable. But what does it get me when actually playing?
Editing is another interesting area to look at, since I was just doing some video editing and compiling this weekend I came to some conclusions. I was taking assorted holiday videos from other family members, ripping, cleaning up and putting into a "best of" montage. I spent a ton of time in Premiere Elements, PhotoShop and had WMP10 running in the background so I could pick music. After several hours I was pretty happy and decided it was time to compile and burn; did I kick it off and marvel at how fast my dual core was working? Nope, I kicked it off and went to work on some disk labels and cover art while it worked at 100% on the other core. Sure, it could have compiled faster with both cores pumping, but I didn't buy a dual core to win benchmarks, but to be productive.
How nice is it to be able to work in FrontPage 2003, PhotoShop and have other assorted apps open and never bog down, or have the sound stutter when something needs a lot of CPU, etc... I know some of this is anecdotal and very user specific, and I have no doubt that is someone devised a way to test all of this AMD would still be faster on paper, but I ask you; who cares? I have no doubt that Windows Vista will come along and make me want a faster CPU, and by then it could easily be an AMD based system, but until they come up with a better way to measure performance on dual cores I can only take articles like this so seriously.
CNET took the entire product line and treated it as relevant when really they should have taken a good look at the Pentium D 820, 830 and 840 with the X2 3800+, 4200+ and 4400+ ; the only honest contenders for 90% of the market. The 840EE and the 4600+/4800+ are fairly irrelevant.
klbf
2 Intern
•
2.7K Posts
0
November 29th, 2005 00:00
wtcnbrwndo4u
2 Intern
•
1.5K Posts
0
November 29th, 2005 01:00
The X2 4800+ runs around $800 with the P4EE going a good $1000.
klbf
2 Intern
•
2.7K Posts
0
November 29th, 2005 05:00
WSA
1 Message
0
December 10th, 2005 11:00