Last daily update:
2025-12-29 14:44:08 UTC
19:17:32 ago
Last incremental update:
2025-12-30 09:40:18 UTC
00:21:22 ago
Server time:
2025-12-30 10:01:40 UTC
Currently online:
4 members
804 guests
Moved from "Invites"
-
Guest
@schloss - Don't worry...you will! ;-) These things are NUTZ! Did you run the CPU-Z prog to see what instructions your proc supports? Make sure you get the right client set for your CPU instructions.
-
Guest
@schloss - s@h and LHC are really faster but there's no significant changes for CPDN and E@H... Hope they release their code soon :-)
-
Guest
@dave: how can the science app be faster with a new CC, just wondering... No, not wondering, this just makes no sense.
-
Guest
@MIC - because they are compiled with generic vars so they work on any proc. When compiling for specific instruction sets, the software is able to take advantage of that and work faster. Such as the SSE3 instructions for Intel.
-
Guest
You got it all wrong... Optimized CC will not make crunching faster. Optimized science app. will.
-
Guest
According to Speedy67 (www.marisan.nl) - "Core Client will not accelerate any calculations, but has an optimized benchmark which will partly compensate the loss of claimed credit. Optimized S@H Client does."
-
Guest
@Grenadier: not really, if more users use 'optimized' (I think it is in the gray area between legit and cheating) science apps, the average credit per WU is going down, because average credit is granted 
-
Guest
@Rytis - That's exactly what happens but because granted credit is based on the averaging of other people's claimed credit, your granted credits are usually quite a bit higher. As for the cheating, I'm in it to crunch all the WU's I can, not the points. If I can crunch more get compile more data....Get 'er done! :-)
-
Guest
@Grenadier - LOL on the backlog. I can see why Berkly doesn't though. To much effort for all the diff platforms along with Intel C++ compiler pricing. It's not free. Linux is a diff story on the pricing though since gcc is open source.
-
Guest
@Rytis - Actually, if credit were granted based on CPU time, my P90 should be claiming thousands of credits =D The closest thing you could argue a credit measures are FLOPs, but even that isn't correct as an optimized client uses less but should still claim the same as an unoptimized client (theoretically, BOINC should claim the same credit given the same WU regardless of the computer or optimizations).
-
Guest
@Puffy: My optimized Seti claimes less for each WU, as I only crunch 10% Seti, and don't use the optimized Boinc. It would inflate my Credits in 90% of my WUs, and thyt's just not fair.
-
Guest
@myName: put your user name into the box at the top of this column, click "search" and Robert will be you father's brother!
-
Guest
Why all the results from my PCs in Afghanistan are all still pending? Do they have some filtering?
Index :: BOINCstats Cafe :: Shoutbox History
