Zealot
I quit My Job for Folding
WE WERE very close to Athlon 64 and Athlon FX last week but didn't get time to play with the systems. Still the world is too small a place to hide the numbers until the 23rd of this month.
AMD has sent many Athlon FXs and 64s to people around UK and continental Europe and if you are an OEM or a system integrator, you know how these babies look and perform.
The Athlon FX on Windows XP 32 bit beats Intel's latest release 3.2 GHz but that's still the Northwood core, of course.
In the Sandra memory test, Athlon FX delivers 5600 MB/s while an Intel Pentium 4 3.2 on Canterwood 875 with DDR 400 of course only delivers 5000MB/s.
In Quake 3 , which was always considered Intel's playground and patch, the Athlon FX is slightly under nine per cent faster on AMD's processor rather than Intel's "brain of a PC".
In Unreal 1024x768, it's close to 18 per cent advantage in AMD's favour.
3Dmark03 at 1024x768 shows that the FX is two per cent slower than on Intel. Pcmark03 is faster on Intel by five per cent since this is an Intel heavily optimized application while the memory score is 18 per cent faster on Athlon FX due to its integrated memory controller.
Still, it's not all roses, roses as Intel still holds the crown in all SSE 2 optimized application and the ones that use HyperThreading. I am mainly talking about rendering applications, where Intel still holds the crown but the gap that used to be huge between Intel's 3.2 and Athlon XP 3200+ is now significantly smaller.
In 3Dstudio Max, a Pentium 4 3.2 GHz with HT is about 10 per cent faster if you render just one frame but in a complete scene that you want to render took exactly the same in both systems. It took them both an hour with a small time difference that's not even worthy of mention and is like a twinkle in the eye (nimesha).
That's what you will see in reviews when they go live on 23rd in exactly three week's time µ
Fonte: http://theinquirer.net/?article=11339
Não sei porquê, mas cheira-me que o processador aqui testado será o 3400+, e daí não é de estranhar a vantagem.
Agora se for um processador abaixo do 3400+, então aí sim, isto vai rular.
AMD has sent many Athlon FXs and 64s to people around UK and continental Europe and if you are an OEM or a system integrator, you know how these babies look and perform.
The Athlon FX on Windows XP 32 bit beats Intel's latest release 3.2 GHz but that's still the Northwood core, of course.
In the Sandra memory test, Athlon FX delivers 5600 MB/s while an Intel Pentium 4 3.2 on Canterwood 875 with DDR 400 of course only delivers 5000MB/s.
In Quake 3 , which was always considered Intel's playground and patch, the Athlon FX is slightly under nine per cent faster on AMD's processor rather than Intel's "brain of a PC".
In Unreal 1024x768, it's close to 18 per cent advantage in AMD's favour.
3Dmark03 at 1024x768 shows that the FX is two per cent slower than on Intel. Pcmark03 is faster on Intel by five per cent since this is an Intel heavily optimized application while the memory score is 18 per cent faster on Athlon FX due to its integrated memory controller.
Still, it's not all roses, roses as Intel still holds the crown in all SSE 2 optimized application and the ones that use HyperThreading. I am mainly talking about rendering applications, where Intel still holds the crown but the gap that used to be huge between Intel's 3.2 and Athlon XP 3200+ is now significantly smaller.
In 3Dstudio Max, a Pentium 4 3.2 GHz with HT is about 10 per cent faster if you render just one frame but in a complete scene that you want to render took exactly the same in both systems. It took them both an hour with a small time difference that's not even worthy of mention and is like a twinkle in the eye (nimesha).
That's what you will see in reviews when they go live on 23rd in exactly three week's time µ
Fonte: http://theinquirer.net/?article=11339
Não sei porquê, mas cheira-me que o processador aqui testado será o 3400+, e daí não é de estranhar a vantagem.
Agora se for um processador abaixo do 3400+, então aí sim, isto vai rular.
