STS engine comparison: suggestions, criticism and help
Posted: Sun Jan 01, 2012 4:38 pm
First of all, happy new year to everybody.
Now, to the point: I asked Swami's permission to use his STS suite, and publish results (he kindly agreed).
My idea is to use STS to compare last\previous versions of an engine, hoping with it to help authors to see in which areas improvements are effective and where they worsen engines performance. The question is: do you think it could be useful?
I have an i7\720 quad, allowing hyperthreading.
I have run a few tests, up to now (see in attach Komodo 3\4, 10 sec/pos, no HT).
I need, however, some help on Hyperthreading: I know that for chess engines HT is bad. Unfortunately my BIOS does not allow me to switch it off, so I'm forced to use TaskManager's options, but some engines, under Arena 2.5 behave in different ways (some using 4 thr, some only 3...). In order to have a uniform behaviour, my idea is to allow HT, so that at least, all engines will work under the same conditions, Do the experts agree?
Txs for comments.
Now, to the point: I asked Swami's permission to use his STS suite, and publish results (he kindly agreed).
My idea is to use STS to compare last\previous versions of an engine, hoping with it to help authors to see in which areas improvements are effective and where they worsen engines performance. The question is: do you think it could be useful?
I have an i7\720 quad, allowing hyperthreading.
I have run a few tests, up to now (see in attach Komodo 3\4, 10 sec/pos, no HT).
I need, however, some help on Hyperthreading: I know that for chess engines HT is bad. Unfortunately my BIOS does not allow me to switch it off, so I'm forced to use TaskManager's options, but some engines, under Arena 2.5 behave in different ways (some using 4 thr, some only 3...). In order to have a uniform behaviour, my idea is to allow HT, so that at least, all engines will work under the same conditions, Do the experts agree?
Txs for comments.