• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Rosetta vs Hyperthreading

E

emjem

Golden Member
I've got a non-hyperthreading P4 2.53 (running at 2.85) crunching WUs. I also have a P4 2.6 with hyperthreading sitting idle for lack of a video card.

My question is how much does hyperthreading contribute to Rosetta crunching? That is, should I shut down the non-hyperthreading P4 2.85 and use the vid card in the hyperthreading P4 2.6, or not.

The non-hyperthreading system benchmarks 1452 (Double Precision MIPS) and 2926 (Integer MIPS) while I think the hyperthreader will run about 1200/1100, judging by some similar machines I have crunching.

I'd run some WU comparison tests but the size of the WUs is so variable that I don't think that will show much.
 
I have been thinking about that problem as well. My P4 2,8 HT has a RAC of 170 which is the same as my Athlon XP 2000+. BUT, it looks like the P4 is outproducing the Athlon when it comes to total production. The Athlon scores 1500/2500 but the P4 only scores 1200/1200, but for EACH virtual CPU, So I guess that R@H does benefit from HT, although it isn't too much, but still a bit.

Does anybody have some numbers about that? After all, I'm just guessing.

:beer:
 
I posted this over there:

"Windows task manager shows that my CPU is spending ~50% of it's time on each of the two WUs running. So on the surface it would appear that an HT cpu does twice as much work as a non-HT unit. But this could very well be false logic since not all is as it appears in the cpu world. It would be nice to have some FACTS on this issue.

As for the memory usage issue with HT I don't see a problem. Two of my P4 3.2 systems use ~69 meg for each 'cpu'. So it would seem that any system with at least 256meg of ram has room for around 80% growth in WU size."
 
Here's what's blowing my mind:

My non-HT system is a P4 2.53/533 OCed to 2.85/600 with one 512k stick of PC2100 ram (overclocked to 300), non dual channel. Boinc benchmark gives it 1452/2926.

My fastest HT system is a P4 2.6C/800 OCed to 3.3/1003 with 2 sticks of 512 PC3200 ram (not overclocked so running 5:4 400), dual channel. Boinc benchmark gives it 1358/1301.

So even if I double the HT Integer MIPS per second (2x1301=2602) the non-HT is about 12% faster. Does not compute! Maybe the boinc benchmark is flawed? Maybe the R@H app is designed for non-HT architecture? Maybe, maybe, maybe. Just like to get some facts mam, just the facts.
 
Originally posted by: emjem
Originally posted by: Freewolf
My P4 2.8 with ht does more per a day than my P4 3.0 without ht does.
Click to expand...


How much more?
Click to expand...

Looks to be at least 40 to 60 points a day worth and the wu times seem to be faster but that could just be faster wus'

 
The 3 gig started crunching on the fourth and has 1,948.28 points
the 2.8 gig started crunching on the 12th and has 1,476.05
the 2.8 also has a higher Average CPU efficiency
 
So maybe the HT is crunching ~60% points per day? And with 7% less horsepower.

I haven't run my crunchers long and consistant enough yet to make any points comparisons. Otherwise my results have got a lot of start-up problems in them.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top