Athlon64 @ 1.4GHz: more and better benchmarks

[kuk]

I know that the 1.4GHz's are older hammer engeneering samples, but this is the best review that I've come across. It's in french, but graphs are a universal language. Notice that unlike the other german review, this one compares all processors while running on DDR333 RAM.

Article starts here ...
Notice the CPU temperature comparison at the end of the page
Benchmarks start here

Kuk

 

[cliftonite]

Wow, thanks

 

[RSMemphis]

Hmm, I like those benchmarks.
It seems that the Hammer offers at least 10-15% more power than an Athlon or P4 at the same frequency.

And that memory latency is yummy.

 

[zephyrprime]

Hey, look at the synthetic memory benchmark. With ddr333, the athlon64 on score 2379 but it should be able to approach the max theoretically possible speed of 2600 like the P4 does.

 

[Sid03]

Originally posted by: RSMemphis
Hmm, I like those benchmarks.
It seems that the Hammer offers at least 10-15% more power than an Athlon or P4 at the same frequency.

i thought they were predicting 25-30% performance over the same speed athlon? and whats up with the gaming scores... theyre horrible!

 

[kuk]

Originally posted by: Sid03

Originally posted by: RSMemphis
Hmm, I like those benchmarks.
It seems that the Hammer offers at least 10-15% more power than an Athlon or P4 at the same frequency.

i thought they were predicting 25-30% performance over the same speed athlon? and whats up with the gaming scores... theyre horrible!

From the little french I understand (curtesy of knowing a latin-based language), the low gaming scores are a result of some AGP/BIOS conflict with the video card, thus a motherboard related problem.

 

[Sid03]

Originally posted by: kuk

Originally posted by: Sid03

Originally posted by: RSMemphis
Hmm, I like those benchmarks.
It seems that the Hammer offers at least 10-15% more power than an Athlon or P4 at the same frequency.

i thought they were predicting 25-30% performance over the same speed athlon? and whats up with the gaming scores... theyre horrible!

From the little french I understand (curtesy of knowing a latin-based language), the low gaming scores are a result of some AGP/BIOS conflict with the video card, thus a motherboard related problem.

i see. why even bother printing bogus results then?

 

[RSMemphis]

KUK, you read correctly, there are some problems with the AGP implementations.

Sid03, I guess just to prove that they tested it. Also, I still think there may be a 20% increase, but why would I trust that website anyway?

Plus, I am a little worried that a lot of people see this is a desktop savior... This thing is gonna kick ass for workstation number crunching, but I am not sure that it will be great for gaming rigs.
I think it is meant for AMD to conquer the business world. And the Barton will stay for the home user... (Mmmmm, Barton).

 

[meson2000]

Well, you have to remember that the Athlon 64 is going to ship with 1meg of L2 cache to start with.
That will improve the results further......
I am waiting to upgrade my current rig with an Athlon 64. August will be the 2 year anniversary
of my current rig. I should be able to get one by then. 3400+ drool......

 

[ls32]

Originally posted by: Sid03

Originally posted by: RSMemphis
Hmm, I like those benchmarks.
It seems that the Hammer offers at least 10-15% more power than an Athlon or P4 at the same frequency.

i thought they were predicting 25-30% performance over the same speed athlon? and whats up with the gaming scores... theyre horrible!

As AMD has pointed out, they need to have quality drivers that are optimised for 64 bit platforms. This could be why the 3dmark 2001 scores are so low compared to the 32 bit platforms while the cpu intensive tests have the opteron outperforming their equivalent mhz predecessors. Once again it seems drivers are holding back the true performance of the cpu.

 

[LordOfAll]

I wouldn't read to much into this. First off this is a 256Kb cache version, and will arrive last in a string of hammer releases. Since you are looking at this chip being released at least 6 months from now the clock will likely be much higher. what i think you are looking at is the "Duron" of the future. In that light this should pummel the celeron.

 

[kuk]

Bump for the evening crowd.

 

[jeffrey]

"From the little french I understand (curtesy of knowing a latin-based language), the low gaming scores are a result of some AGP/BIOS conflict with the video card, thus a motherboard related problem."


I have heard of the Athlon 64 AGP issues before. This isn't the first time that I have read a preview that blamed poor Athlon 64 performance on AGP issues. This is the type of issue that would get a lot of attention by this forum. I would like to see some serious testing of this issue after the chip is released. If I remember correctly Nvidia was trying to work around this issue on their motherboard implementation and avoid the AGP tunnel that AMD is trying to fix.

 

[EdipisReks]

i REALLY hope that they fix the problems. otherwise, yuck.