PDA

View Full Version : A64 at 2.8 ghz

f00t
10-21-03, 06:59 PM
pretty amazing results if ya ask me

linky (http://www20.tomshardware.com/cpu/20031021/index.html)
{PMS}fishy
10-21-03, 07:18 PM
Memory bandwith is killer. The P4s at stock speed are still walking all over a CPU thats clocked to the max. Run one of those P4s up to 4ghz and there is no comparision. Untill software begins to utlize the 64bit instruction set there is not much of a point, as far as I am concerned.
johnoh
10-21-03, 08:00 PM
was there a memory bandwidth benchmark in that link that I missed? Even at 2.2ghz the a64 walked all over the p4 3.2g
Feuerkampf
10-21-03, 08:22 PM
Originally posted by f00t
pretty amazing results if ya ask me

linky (http://www20.tomshardware.com/cpu/20031021/index.html)

Are you reading the same Article as me?

It all looks super positive to me, and it's from Tom who has a history of AMD bashing.
TheGr8s1
10-21-03, 08:57 PM
That was an unfair review anyways. I agree that the AMD FX-51 is a kickarse card stock or overclocked but it's compared in that review to only stock P4's. Why don't they compare a highend P4 800fsb (3.0c?) with similar cooling (Prommy) overclocked to 4ghz+. Personally I think the performance gap is very narrow in favor of either company depending on the application involved when both are overclocked. Intel still has an edge on memory bandwidth as well.
Speed_Mechanic2
10-21-03, 09:48 PM
THG's Intel-bias shows even in that seemingly in-favor of AMD review.
OC Detective
10-21-03, 10:01 PM
OK its a FX51 not an A64 and those results actually suck especially the 3D 2K1SE results - which are a meagre 20713 - everyone knows that people are getting about 26 - 27 thousand with these cpus overclocked - go check over at xtreme! I guess he has been using crappy drivers!
xtrmeocr
10-21-03, 10:40 PM
i have to agree those scores seamed awefull low.
pelikan
10-21-03, 11:35 PM
I don't know what you P4 guys are talking about. Taken a look at 3DMark01 on the ORB lately? The A64 is dominating by a longshot. Even P4's clocked at 4.3 GHz+ are behind.
OC Detective
10-22-03, 12:32 AM
Dude - who are you referring to when you state "you P4 guys"? Check my sig - and my point was that as per usual THG produces results which show AMD cpus in a bad light due to either bad set up or plain erroneous figures. My post even mentioned the very same results you are talking about.
Tyranos
10-22-03, 12:52 AM
Funny how THG seems to have the absolute worst scores for an FX51 or A64 I've come across, making it look like a slightly faster XP in performance. And also included is a price break down of the athlon system. Hmm.
OC Detective
10-22-03, 01:16 AM
If you want to have a good laugh check out the rating of the graphics card they used on the FX51 - bunch of amateurs.
Tyranos
10-22-03, 01:46 AM
128mb 5900u, all for over six hundred bucks?? wooo count me in.:p
Gerfey
10-22-03, 02:27 AM
Very funny article if you compare it to when they had the P4 at 4 ghz. I don't think there is a single benchmark that you can compare from the two, because they used different video cards so all the game benchmarks are invalid, none of the application benchmarks are of the same version so you can't compare, and they used none of the same synthetic benchmark versions. I know why they did it of course, both articles used the newest versions, but it would have been cool to see a direct comparison.
JCviggen
10-22-03, 02:38 AM
in gaming benches the FX walks over the P4 at 4G+ , in some workstation things it goes back and forth.

memory bandwidth isnt a game, and latency is often more important than bandwidth...the actual benchmarks are what matter anyway.

I think the only point in this article is performance scaling with clock speed, not that it is faster than the P4s
OC Detective
10-22-03, 03:29 AM
Yup that is about the only valid thing to draw from that review - its excellent scaleability - which means that when it moves to 0.09 or the steppings mature its gonna be real sweet for overclocking in terms of performance increase.
pelikan
10-22-03, 10:04 AM
Originally posted by OC Detective
Dude - who are you referring to when you state "you P4 guys"? Check my sig - and my point was that as per usual THG produces results which show AMD cpus in a bad light due to either bad set up or plain erroneous figures. My post even mentioned the very same results you are talking about.

I wasn't talking about your post. I was talking about PMSFishy's. Sorry, I should have been more specific.
shiyan
10-22-03, 10:18 AM
just one comment about why guys at xtreme get 27k with similar hardware... and toms only got 20k.

I think the card in the review was not overclocked all that much...
TheGr8s1
10-22-03, 10:20 PM
Originally posted by shiyan
just one comment about why guys at xtreme get 27k with similar hardware... and toms only got 20k.

I think the card in the review was not overclocked all that much...

exactly! no volt mods.....probably not even overclocked at all....
dropadrop
10-23-03, 01:08 AM
Originally posted by shiyan
just one comment about why guys at xtreme get 27k with similar hardware... and toms only got 20k.

I think the card in the review was not overclocked all that much...

I believe the videocard in toms hardware review was running at stock speeds. He was also overclocking with the multiplier, not with fsb. It would have been good to have an overclocked P4 on the side, but as Jc mentioned the important part was to see how well the k8 scales. Seems like the processor still has alot to give. (like it should ofcourse as a brand new model)