Yes I understanded that UD only support 1 core. I'm using 100% of that 1 core out of the 2 cores available and in theory it should be 2 times faster than the pentium 4 HT which is only running at 50%. (using 1 thread)

If i use BOINC at 10% cpu, i should expect it to be 4 times faster than pentium 4 HT running 50%

thanks

Aynfong, glad you switched to BOINC. The HT wont benefit too much, but the Pentium D will.

Since i've been very successful in deploying and running BOINC+UD side by side on an 1 HT machine, i can tell from own experience that that gives a very very considerable improvement....there's more 'gap filling' then the none HT users can appreciate...Threadmaster contains the lot and has been set to allow only 95% of the combination....got to do something with the thing

BOINC on its own on an HT machine ** still treats it as 1, at least i've not figured out how to get it to run 2 BOINC WU's same time, hence why i have UD on the side for the gap filling, which is 7 hours CPU time out of 45 hours Clock time as of this writing for the last FAAH on UD. I'm thus very curious what you're going to show in Taskmanager once the BOINC is running !!!!

** I've wondered more than a few times if i really have an HT...all the independent CPU identification software says it is, the BIOS has it activated, but never seen 50%....unless Shuttle has a special implementation?

PS Fly your Flag and put your country in the Forum Profile and the My Grid Profile...points are added to the country of choice...right now Europeans are much more proud than the USAens added up. We're so far ahead. Merkens tell me, Merkens don't fly their flag (so much).....G Dubbaya will take offense

ciao

Yep thats why i set my profile to 1 CPU on advise here on the forum quite some time ago when i kicked off with BOINC

Something else that occurrs to me in the Single Processor, HT Processor, Dual Processor debate....

THe HT as we know has only one core and definitely only one FPU. The descriptions of the Pentium D say it is 2 separate cores. Does this mean it has 2 FPUs (the most usefull bit from our point of view) or just 2 main processors still sharing one FPU - either fully or partly? I cant find any processor block diag for the P4D that shows this. I am still not sure that a P4D is not in some way INFERIOR to 2 real separate processors, by sharing something such as FPU or memory controller/address decoder.

Benchmark programs would have to excercise BOTH cores at the same time in order to be sure that it has totally separate FPU and other resources. Benchmark progs I have seen dont say how they do it.


I have 3GHz Pentium D, now with 2GB RAM and BOINC Benchmarks are :-

14/07/2006 08:17:06||Benchmark results:
14/07/2006 08:17:06|| Number of CPUs: 2
14/07/2006 08:17:06|| 1501 floating point MIPS (Whetstone) per CPU
14/07/2006 08:17:06|| 2417 integer MIPS (Dhrystone) per CPU
14/07/2006 08:17:06||Finished CPU benchmarks


My Laptop (Dell Lattitude Pentium M 1.8GHz 512 GB) reports:-

14/07/2006 08:17:01||Benchmark results:
14/07/2006 08:17:01|| Number of CPUs: 1
14/07/2006 08:17:01|| 1599 floating point MIPS (Whetstone) per CPU
14/07/2006 08:17:01|| 3387 integer MIPS (Dhrystone) per CPU
14/07/2006 08:17:01||Finished CPU benchmarks

I know the BOINC benchmarks are not supposed to be very reliable, but a 3 GHz P4D appears slower that an 1.8 GHz P4!!