AMD vs Intel

Message boards : Number crunching : AMD vs Intel

To post messages, you must log in.

AuthorMessage
Profile Robert

Send message
Joined: 3 Jan 06
Posts: 4
Credit: 23,755
RAC: 0
Message 8462 - Posted: 6 Jan 2006, 3:37:46 UTC

I thought that I had built the best gamer that I could.An Intel P4 630 3.0 ghz,ECS RS 400-A MB,ATI Radeon X800XL video with 1 Gig of ram.But my old gamer a Sempron 2500+,ECS K600-a,ATI 9200 video with 512 meg memory crunches Rosetta faster.Did I do something wrong when I built gamer01 ?I thought that it would run rings around the old Sempron.
Any help would be appreciated.
Robert

ID: 8462 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Polian
Avatar

Send message
Joined: 21 Sep 05
Posts: 152
Credit: 10,141,266
RAC: 0
Message 8463 - Posted: 6 Jan 2006, 3:59:07 UTC
Last modified: 6 Jan 2006, 3:59:43 UTC

Your Intel should be crunching much faster... If you're basing your observations on BOINC benchmarks - It may be an illusion because you have hyperthreading enabled on the Intel. You're crunching two workunits on two 'virtual' processors. Your benchmarks per CPU are lower, but add them together... a hyperthreaded single CPU machine should turn out more work than a non-hyperthreaded CPU.

p.s. AMD vs. Intel = Ford vs. Chevy
ID: 8463 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Paul D. Buck

Send message
Joined: 17 Sep 05
Posts: 815
Credit: 1,812,737
RAC: 0
Message 8472 - Posted: 6 Jan 2006, 8:18:26 UTC

Robert,

If you look at work over time, the P4 takes longer on each work unit, but with two of them in work at a time, it gets more done per unit time.
ID: 8472 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Robert

Send message
Joined: 3 Jan 06
Posts: 4
Credit: 23,755
RAC: 0
Message 8474 - Posted: 6 Jan 2006, 8:23:54 UTC - in response to Message 8472.  

Robert,

If you look at work over time, the P4 takes longer on each work unit, but with two of them in work at a time, it gets more done per unit time.


Thanks for the answers,I'll look closer at the results.For $179 for the motherboaed and CPU I couldn't pass it up.
ID: 8474 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Paul D. Buck

Send message
Joined: 17 Sep 05
Posts: 815
Credit: 1,812,737
RAC: 0
Message 8475 - Posted: 6 Jan 2006, 8:47:27 UTC - in response to Message 8474.  

Robert,

If you look at work over time, the P4 takes longer on each work unit, but with two of them in work at a time, it gets more done per unit time.


Thanks for the answers,I'll look closer at the results.For $179 for the motherboaed and CPU I couldn't pass it up.

There are many factors that come into play, dual channel memory vs. single, cache size, memory speed, video using system memory, etc. that can change the expected performance.

Oh, and quality of the MB too ...
ID: 8475 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
FluffyChicken
Avatar

Send message
Joined: 1 Nov 05
Posts: 1260
Credit: 369,635
RAC: 0
Message 8485 - Posted: 6 Jan 2006, 16:44:17 UTC

Also note that the CPU floating point unit (I believe) is the most important for Rosetta, at this the AMD CPU's are currently superior. But for gaming your graphics card comes into play far more for modern gaming as well as you increase of RAM.

Some things prefer the AMD architecture, some the Intel.... all else being equal.

Although I do not know the relative throughput of Duron/AthlonXP/Sempron SocketA to P4's for this project.

I would completely ignore the benchmarks in boinc as a 'benchmark' of your throughput ;-)

I guess until there is a 'default' workunit we can do, relative testing will be difficult
Team mauisun.org
ID: 8485 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Robert

Send message
Joined: 3 Jan 06
Posts: 4
Credit: 23,755
RAC: 0
Message 8701 - Posted: 10 Jan 2006, 4:40:55 UTC - in response to Message 8462.  

I thought that I had built the best gamer that I could.An Intel P4 630 3.0 ghz,ECS RS 400-A MB,ATI Radeon X800XL video with 1 Gig of ram.But my old gamer a Sempron 2500+,ECS K600-a,ATI 9200 video with 512 meg memory crunches Rosetta faster.Did I do something wrong when I built gamer01 ?I thought that it would run rings around the old Sempron.
Any help would be appreciated.
Robert

I think that I've found the problem,the builder(me).The M/B wants ddr2 677mhz and not the ddr2 400mhz that I installed.Well off to Frys tomorrow to get some faster memory.
And thanks again for all your help.
Robert

ID: 8701 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
R/B

Send message
Joined: 8 Dec 05
Posts: 195
Credit: 28,095
RAC: 0
Message 8755 - Posted: 11 Jan 2006, 4:20:54 UTC - in response to Message 8485.  

Also note that the CPU floating point unit (I believe) is the most important for Rosetta, at this the AMD CPU's are currently superior. But for gaming your graphics card comes into play far more for modern gaming as well as you increase of RAM.

Some things prefer the AMD architecture, some the Intel.... all else being equal.

Although I do not know the relative throughput of Duron/AthlonXP/Sempron SocketA to P4's for this project.

I would completely ignore the benchmarks in boinc as a 'benchmark' of your throughput ;-)

I guess until there is a 'default' workunit we can do, relative testing will be difficult


Does anyone have a list of all the BOINC projects and the processors they 'prefer'? As in AMD vs Intel? I've just done trial and error type experimentation on the 3 machines I use. I run SETI,ROSETTA,EINSTEIN,AND PREDICTOR. But a full list from someone that runs alot of the projects would be interesting. Thank you.

Founder of BOINC GROUP - Objectivists - Philosophically minded rational data crunchers.


ID: 8755 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Webmaster Yoda
Avatar

Send message
Joined: 17 Sep 05
Posts: 161
Credit: 162,253
RAC: 0
Message 8756 - Posted: 11 Jan 2006, 4:45:22 UTC - in response to Message 8755.  
Last modified: 11 Jan 2006, 5:19:29 UTC

Does anyone have a list of all the BOINC projects and the processors they 'prefer'? As in AMD vs Intel?


I'd be interested in that as well. From my own recollection though:

  • Rosetta was fastest on Athlon (XP/64)
  • SETI (with optimised application) works best on Intel with SSE2 or SSE3 instruction set, and Athlon 64 with SSE3:

    • 1.7GHz Celeron (SSE2), Linux: avg 1.5 hours
    • 2.3GHz Athlon XP 3000+ (SSE), WinXP: avg 2 hours+
    • 2.4GHz Pentium4 (SSE2), WinXP: avg a little under 1 hour
    • 2.6GHz Athlon 64 3700+ (SSE3), WinXP64: avg 40 minutes
    • 3.4GHz Pentium 4, HT (SSE3), Win2003: avg 55 minutes for 2 WU


  • The old (pre Albert app) Einstein worked best on Athlons (XP/64). From memory, the Celeron took 16 hours, the P4/2.4 took 10 hours, the Athlon XP took 5.5 hours, the Athlon 3700+ took less than 5 hours and the Pentium 4/3.4 with HT took 11-12 hours for two.
  • Predictor also worked best on Athlon XP/64
  • CPDN likes fast Pentiums (even with HT on) better than Athlon


But your mileage may vary :-)


*** Join BOINC@Australia today ***
ID: 8756 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Paul D. Buck

Send message
Joined: 17 Sep 05
Posts: 815
Credit: 1,812,737
RAC: 0
Message 8769 - Posted: 11 Jan 2006, 12:12:36 UTC

Einstein@Home, both versions is best on the G5 ... they used hand coded Altivec code. At one point I was doing 51 minute work units on the G5 ... now they are 2:30 to 3:00 usually ...

My 2.0 GHz G5 with Team MacNN client does work about as fast as does Crunch3rs application on my 3.4 GHz Xeon.
ID: 8769 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
R/B

Send message
Joined: 8 Dec 05
Posts: 195
Credit: 28,095
RAC: 0
Message 8864 - Posted: 12 Jan 2006, 17:48:05 UTC

Yesterday I hooked up a fourth machine to BOINC. You'll get a laugh out of this. It's an old 133 Mhz 80 Ram cruncher. I don't think it will even run Rosetta. I've got it working on a seti unit and the estimated total time is 135 hours.
Founder of BOINC GROUP - Objectivists - Philosophically minded rational data crunchers.


ID: 8864 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Hans Schulze

Send message
Joined: 20 Dec 05
Posts: 7
Credit: 102,405
RAC: 0
Message 9169 - Posted: 17 Jan 2006, 3:10:45 UTC

Best SETI machine I ever used was a Dual-XEON HT 2800 MHz machine, crunching at about 4WU per 2 hours. SETI made good use of the large on-chip caches. Would be nice to see what some Intel EE or more recent Athlon FX chips do.
That machine shows recent credits of close to 500, where dual Opteron 242 1.6 GHz 2600+ are pulling in about 175. A new Ath64 3700+ is running around 135.
Anyone know what the "recent average credit" timeperiod is? A week?
ID: 9169 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Tern
Avatar

Send message
Joined: 25 Oct 05
Posts: 576
Credit: 4,695,450
RAC: 5
Message 9173 - Posted: 17 Jan 2006, 4:33:07 UTC - in response to Message 9169.  

Anyone know what the "recent average credit" timeperiod is? A week?


It's a "halflife" calculation - very complicated. My AMD 3700 was running about 600 w/ optimized SETI apps and clients, now it's mostly Rosetta, so running about 500. See the thread on "Recent Average Credit".

ID: 9173 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : AMD vs Intel



©2024 University of Washington
https://www.bakerlab.org