Hi, everyone.

At the urging of mo.v I am starting this thread to compare the speed in s/TS of HadCM3 coupled models on various types of machines. Please state the type of processor (AMD or Intel), the OS, the gigahertage (is their really such as word? If not there should be.) of the processor and the amount of RAM. Also please say whether it is a desktop or a laptop.

If you don’t these values you can find them by clicking “Computer” then “System Properties” in Vista or “My Computer” followed by “System Information” in XP.

For multi-core processors please state if you are running other WUs (and their type) or other projects on the other cores.

My ACER Intel Core2Duo 1.5 GHz laptop with 2 GB of RAM is doing 4.05s/TS at 18% finished an 80 year WU. There is an HadSM3mh running on the other core.

---

2.07 s/TS Core2 Quad Q6600 @ 2.4GHz XP Pro X64 2GB 800MHz RAM with a variety of CPDN partners.
2.02 –2.04 s/TS Core2 Quad Q6600 @ 2.4GHz XP Pro X64 2GB 1066MHz RAM with a variety of CPDN/beta partners.
2.133 – 2.286 s/TS Core2 Quad Q6600 @ 2.4GHz XP Pro X64 2GB 1066MHz RAM with x3 beta HADAM3P. The load from these models causes the CM3 to run about 11% slower.

2.87 s/TS Core2 Duo T2500 @ 2.0GHz XP Pro SP2 2GB Lenovo T60 with x1 HADAM3
2.97 s/TS Core2 Duo T2500 @ 2.0GHz XP Pro SP2 2GB Lenovo T60 with x1 beta HADAM3P

---

1. Core 2 Q6600(B3) @ 2.4, 3gb RAM (4x1gb installed) @ 667MHz RAM (PC2-5300),
XP Pro SP3 X32, Dell desktop.
2 x HadCM3 each @ 2.16 s/TS
1 x HadSM3 @ 1.44 s/TS, 1 x HadSM3MH @ 1.34 s/TS

2. Core 2 Q6600(G0) @ 3.37, 3.24gb RAM (2x2gb installed) @ 1000MHz (PC2-8500),
XP Pro SP3 X32, Custom-built desktop.
2 HadCM3 each @ 1.53 s/TS
2 HadSM3 each @ 0.99 s/TS

AMD Opteron 1210 at 1.8 GHz
2 GB RAM
OS SuSE Linux 10.3
3.40 s/TS on one core
The other core runs another project from CPDN Beta, Einstein@home, SETI@home, QMC@home, AQUA@home, LHC@home

---

The last HadCM3 model I ran was getting 1.06s/TS on average.

For reference, my desktop computer is Intel Core2Duo 3.0GHz (E8400) w/ 4GB RAM, Fedora 8 linux. The only other projects I run are SETI and Rosetta.

1st computer: laptop
Intel Celeron M @1.40GHz
XP Home Ed. SP3
504 MB RAM
Running application 6.02, 59% though an 80 yr WU
Most recent: 4.48 s/TS and dropping
*this computer has varied between 4.9 s/TS and this value in the past, but is currently continuing to slowly drop

2nd computer: desktop
Intel Core 2 Quad Q6600 @2.40GHz
Vista Home Premium (64bit) SP1
4.00 GB RAM
Running application 6.04, 39% through an 80 yr WU
2.04 s/TS
All cores rotate between this project and the following (at the moment): ABC, Aqua, Cosmology, Einstein, LHC, Milkyway, PrimeGrid, QMC, Spinhenge, and uFluids.

---

~It only takes one bottle cap moving at 23,000 mph to ruin your whole day~

Hi
Desktop (no brand name)

Windows Vista Home basic

One processor

AMD Athlon processor 3500+ 2.21Ghz

3Gb RAM

Average of 2.76 T/S

AMD Quad core Phenom 9950BE @ stock 2.6GHz, 4GB 800MHz RAM, MSI DKA790GX Motherboard - HTT @ 200MHz, PCI-E GFx card, Windows XP Pro 64bit, BOINC 6.2.19 - Homebrewed Tower system:

2.09 s/ts for 1 HadCM3ist_ model running in conjunction with 3 CPDN Beta HadAM3P runs.

The same HadCM3 run rate improved to 1.98 s/TS when the 3 Beta runs had completed and were temporarily replaced with 3 less CPU/RAM intensive, short time to completion runs from the \'World Community Grid\' Project (BOINC version)

Therefore, there can be a performance hit when running alongside other more CPU &/or RAM intensive runs.

EDIT: It is probably worth also including motherboard (or at least chipset type and GFx type, i.e. onboard or AGP/PCI-E) since these can aslo affect run rate with otherwise identical components. If not known, say on a prebuilt system, the BIOS or CPU-Z will give the information.

---

At the moment I am running this HadCM model on a desktop Athlon 64 3500+ with 2GB of memory. OS is Ubuntu Linux 8.10 64bit version. Kernel 2.6.27-13.
TS is 2.8377.

[Edit] The previous HadCM model I ran had a TS of 2.845. Some tweaking of the powersave features options in Ubuntu gained me some extra crunching power. [/edit]

Cheers

Jos

HADCM don\'t run that good on my puter, at least compared to HADSM.
My only succesfull WU took 1.6574 s/TS or 26c/h, two others failed after quite a lot of time with kind of the same speed.

---

Grüße vom Sänger

The last HADCM3 model work units I ran on either an Opteron 275 or a Opteron 285 both used to take around 1.93 to 1.97 s/TS.

Thinking of trying another one to see if they have changed in a year.

I had a problem with the displayed times on the work units speeding up about 4 times the actual process times and showing a TS of 0.21 to 0.50 seconds which was not correct, this happened twice on different work units.
So probably time to try again.

---

Thinking of trying another one to see if they have changed in a year.

With the version 6.x hadcm3, linux performance degraded significantly. Any new one you download won\'t go very fast. With my Opteron 185, it used to run 2 at 1.7-1.75 s/TS. Now it runs one at 2.3-2.4 s/TS.

Crunch Machine: VMWare 1 CPU Virtual Machine,
Q6600 @2.52 GHz 512MB DDR2 @840 MHz, WinXP32

Previous HADCM3 WU: 2.0278 sec/TS at completion.
Current HADCM3 WU: 2.0127 sec/TS at 89% complete.

Notes:
Current speed slightly faster due to optimising configuration of host and VM\'s.

This is one of four VM\'s running on Q6600 4GB DDR2, WinXP32, each VM has affinity set to one core each.

VM1 Runs 90% Seti, 10% Rosetta
VM2 runs 90% Rosetta, 10 % Seti
VM3 runs 90% CPDN (this VM), 10% Seti
VM4 set to run 90% Orbit but currently running Seti, Rosetta & CPDN (HADSM) @ 33% each