FX-55 vs. X2 4800+

[formerprof]

I am running with an FX-55 CPU on the ASUS A8n-SLI Deluxe board. My CPU is
always at full load because SETI is running 24/7. Could I expect any
noticeable performance gains from the 4800+ X2? Presumably the SETI
calculations would be primarily confined to one of the two cores.

My memory is running at DDR 400, 2T. I've read somewhere that the newer X2
Athlon 64s have improved memory bus performance. Is there a chance that my
chips will run at 1T with the newer CPU? I have 2x1gb 3,3,3,8 DIMMs.

All help will be most gratefully received. Good wishes to all.

formerprof

 

[Phil DeBecker]

I am running with an FX-55 CPU on the ASUS A8n-SLI Deluxe board. My CPU is
always at full load because SETI is running 24/7. Could I expect any
noticeable performance gains from the 4800+ X2? Presumably the SETI
calculations would be primarily confined to one of the two cores.

My memory is running at DDR 400, 2T. I've read somewhere that the newer X2
Athlon 64s have improved memory bus performance. Is there a chance that my
chips will run at 1T with the newer CPU? I have 2x1gb 3,3,3,8 DIMMs.

All help will be most gratefully received. Good wishes to all.

formerprof

An X2 CPU would undoubtedly give you a huge increase in performance,
since as you say one core would run SETI while the other would be free
to run whatever else you are doing.

As far as the 1T/2T, I would not expect the CPU to change that, since
the command rate is determined by the motherboard, not the CPU. In
any case you would see no measurable difference in overall performance
even if you did change the command rate to 1T.

The 4800+ X2 is a good upgrade; it will run single-threaded
applications about as fast as your FX-55, and since it is fabricated
with a 90nm process instead of the 130mm process of the FX-55 and runs
at 2400mhz instead of 2600 mhz, it will run much cooler.

Phil

 

[Mark A]

An X2 CPU would undoubtedly give you a huge increase in performance,
since as you say one core would run SETI while the other would be free
to run whatever else you are doing.

As far as the 1T/2T, I would not expect the CPU to change that, since
the command rate is determined by the motherboard, not the CPU. In
any case you would see no measurable difference in overall performance
even if you did change the command rate to 1T.

The 4800+ X2 is a good upgrade; it will run single-threaded
applications about as fast as your FX-55, and since it is fabricated
with a 90nm process instead of the 130mm process of the FX-55 and runs
at 2400mhz instead of 2600 mhz, it will run much cooler.

Phil

I thought SETI ran in the background only when other tasks are not running.
If so, then you would likely see a decrease in performance when running
other stuff if you switch to the 4800 X2.

 

[formerprof]

I thought SETI ran in the background only when other tasks are not
running. If so, then you would likely see a decrease in performance when
running other stuff if you switch to the 4800 X2.

1. Phil -- Thanks for your note. I was under the impression that the Athlon
64 chips had the memory controller built into them rather than as part of
the chipset, and I also vaguely remember that the 1T/2T issue was supposed
to have something to do with insufficient "drive" (or whatever) for large
chips. Am I mistaken?

2. Mark -- do I understand you to mean that each individual X2 core is
slower than the FX-55? D you know what the comparative "number" of the FX-55
would be if it were numbered in the series with the rest of the Athlon 64
CPUs?

Thanks so much to both of you for your help.


formerprof

 

[Rob Stow]

1. Phil -- Thanks for your note. I was under the impression that the Athlon
64 chips had the memory controller built into them rather than as part of
the chipset,

The AMD64 chips all have the integrated memory controllers, but
those controllers look to the BIOS settings to see what memory
timings/latencies to use.

and I also vaguely remember that the 1T/2T issue was supposed
to have something to do with insufficient "drive" (or whatever) for large
chips. Am I mistaken?

This recent article has a good, basic, description of what the
various latency/timing numbers is all about:
http://techreport.com/etc/2005q4/mem-latency/index.x?pg=1