Um, this is a little embarrassing to admit in public; I'll just say it: I'm not optimized
Aahh, you're not alone in the ranks of 'late' optimisers. But stung by your admission I've plonked akosf's S40 version onto my P4 HT 3G. Stunning - average times per Albert have plummeted from a bit more than 30K to a bit less than 8K!!
No wonder there has been floated a concern about server overload with greater WU throughput!
Ohh.. I'm dizzy.. whew ... now for the rest of my small farm..
Cheers, Mike.
Mike,
If you're going the Akos route, I'd strongly suggest you also use an optimized BOINC client, crunch3r's or truxoft, to keep your claims from dragging bottom, and there is an increasing probability of pairing 2 Akos customers in a quorum. If they both use the official app, the group can get slashed down to 4-5 credits. Be wary LOL
microcraft
"The arc of history is long, but it bends toward justice" - MLK
Mike,
If you're going the Akos route, I'd strongly suggest you also use an optimized BOINC client, crunch3r's or truxoft, to keep your claims from dragging bottom, and there is an increasing probability of pairing 2 Akos customers in a quorum. If they both use the official app, the group can get slashed down to 4-5 credits. Be wary LOL
Thank you Mike. I'd done that - the TruXoft version - and noted an almost immediate doubling of the claimed credit. The nett result is from just over 600 seconds per claimed unit of credit ( all standard ) to just under 400 seconds per claimed unit of credit ( akosf + Truxoft ). Roughly speaking it'll chew thru a bit less than four times the WU's, but claim slightly above 1.5 times the credit.... so I guess that's the basis for the concerns of 'equity' I've seen expressed elsewhere. I really can't get excited about counting in an imaginary currency anyway, except to monitor progress. However I appreciate that it may strongly motivate others to contribute though ( for whatever reason ), and thus I'm quite happy if they are accommodated in their concerns, because it will result in their co-operation to then achieve the signal analysis and the science result.
I'll think I'll leave the changes as is for a day or two, see what happens, mainly to demonstrate stability etc.. then look to the remainder of my boxes.
Cheers, Mike.
( edit ) So it's occured to me that I now can be less concerned about the speed impact of enabling the screensaver! To celebrate I've kicked back and am now sipping a nice glass of Barossa Valley red ( a merlot ) while dozily watching the LIGOs' bullseye tiptoe across the celestial globe - upon my favourite Samsung LCD flatty. Go E@H!! :-)
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Um, this is a little embarrassing to admit in public; I'll just say it: I'm not optimized
Aahh, you're not alone in the ranks of 'late' optimisers. But stung by your admission I've plonked akosf's S40 version onto my P4 HT 3G. Stunning - average times per Albert have plummeted from a bit more than 30K to a bit less than 8K!!
No wonder there has been floated a concern about server overload with greater WU throughput!
Ohh.. I'm dizzy.. whew ... now for the rest of my small farm..
Cheers, Mike.
A lady who just turned 104 years old was asked, “What's the best thing about being that old?� She said, “No peer pressure.� She's clearly not crunching; I'll try to get optimized this evening. :)
A lady who just turned 104 years old was asked, “What's the best thing about being that old?� She said, “No peer pressure.� She's clearly not crunching; I'll try to get optimized this evening. :)
Oh ..... Chipper!! ROFL!!! :=)
For all the changes in society in recent generations, truly we have not left the tribal hut! :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Just plugged in akosf's S40.03 optimized version of albert. Amazing how fast the orange cross-hairs move now, across the celestial sphere in the screen saver. The search pattern for each work unit is clearly evident; quite a dramatic improvement in the crunching! :)
The nett result is from just over 600 seconds per claimed unit of credit ( all standard ) to just under 400 seconds per claimed unit of credit ( akosf + Truxoft ).Roughly speaking it'll chew thru a bit less than four times the WU's, but claim slightly above 1.5 times the credit.
So interestingly after about week of 'akosf + Truxoft' the average is now about 220 seconds per claimed unit of credit. This is consistent with the initial underclaiming of credit by TruXoft, then a gradual up-calibration. All of the farm is now chewing through about 4 times the previous number of WU's, and now claiming about 2.7 times the credit ( compared with vanilla flavours ). Stability is good, with the exception of my laptop which required a buff 'n' clean of it's little fan and ventilation ducts, a re-seating of the graphics chip heat sink, and a general blow out of dust ( it was shutting down due to the rising heat ).
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
So it's occured to me that I now can be less concerned about the speed impact of enabling the screensaver! To celebrate I've kicked back and am now sipping a nice glass of Barossa Valley red ( a merlot ) while dozily watching the LIGOs' bullseye tiptoe across the celestial globe - upon my favourite Samsung LCD flatty. Go E@H!! :-)
I've paid little attention to posted opinion on screensaver resource consumption for Einstein at home, but I was quite shocked at the very large impact it had on my Pentium M (Banias) laptop. Removing the screen saver raised science output to akosf S-39L by over 30% on that machine. In contrast, on my Gallatin (P4 EE HT) desktop, I could not see the difference made by the screensaver--believe it is certainly under 5%, and most likely much less.
Possibly the real difference is in the graphics boards or their drivers, but just a warning--it can be substantial.
RE: RE: Um, this is a
)
Mike,
If you're going the Akos route, I'd strongly suggest you also use an optimized BOINC client, crunch3r's or truxoft, to keep your claims from dragging bottom, and there is an increasing probability of pairing 2 Akos customers in a quorum. If they both use the official app, the group can get slashed down to 4-5 credits. Be wary LOL
microcraft
"The arc of history is long, but it bends toward justice" - MLK
RE: Mike, If you're going
)
Thank you Mike. I'd done that - the TruXoft version - and noted an almost immediate doubling of the claimed credit. The nett result is from just over 600 seconds per claimed unit of credit ( all standard ) to just under 400 seconds per claimed unit of credit ( akosf + Truxoft ). Roughly speaking it'll chew thru a bit less than four times the WU's, but claim slightly above 1.5 times the credit.... so I guess that's the basis for the concerns of 'equity' I've seen expressed elsewhere. I really can't get excited about counting in an imaginary currency anyway, except to monitor progress. However I appreciate that it may strongly motivate others to contribute though ( for whatever reason ), and thus I'm quite happy if they are accommodated in their concerns, because it will result in their co-operation to then achieve the signal analysis and the science result.
I'll think I'll leave the changes as is for a day or two, see what happens, mainly to demonstrate stability etc.. then look to the remainder of my boxes.
Cheers, Mike.
( edit ) So it's occured to me that I now can be less concerned about the speed impact of enabling the screensaver! To celebrate I've kicked back and am now sipping a nice glass of Barossa Valley red ( a merlot ) while dozily watching the LIGOs' bullseye tiptoe across the celestial globe - upon my favourite Samsung LCD flatty. Go E@H!! :-)
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
RE: RE: Um, this is a
)
A lady who just turned 104 years old was asked, “What's the best thing about being that old?� She said, “No peer pressure.� She's clearly not crunching; I'll try to get optimized this evening. :)
I'm optimised and crunching
)
I'm optimised and crunching on...
Click my stat image to go to the BOINC Synergy Team site!
RE: A lady who just turned
)
Oh ..... Chipper!! ROFL!!! :=)
For all the changes in society in recent generations, truly we have not left the tribal hut! :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Just plugged in akosf's
)
Just plugged in akosf's S40.03 optimized version of albert. Amazing how fast the orange cross-hairs move now, across the celestial sphere in the screen saver. The search pattern for each work unit is clearly evident; quite a dramatic improvement in the crunching! :)
RE: The nett result is from
)
So interestingly after about week of 'akosf + Truxoft' the average is now about 220 seconds per claimed unit of credit. This is consistent with the initial underclaiming of credit by TruXoft, then a gradual up-calibration. All of the farm is now chewing through about 4 times the previous number of WU's, and now claiming about 2.7 times the credit ( compared with vanilla flavours ). Stability is good, with the exception of my laptop which required a buff 'n' clean of it's little fan and ventilation ducts, a re-seating of the graphics chip heat sink, and a general blow out of dust ( it was shutting down due to the rising heat ).
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
RE: So it's occured to me
)
I've paid little attention to posted opinion on screensaver resource consumption for Einstein at home, but I was quite shocked at the very large impact it had on my Pentium M (Banias) laptop. Removing the screen saver raised science output to akosf S-39L by over 30% on that machine. In contrast, on my Gallatin (P4 EE HT) desktop, I could not see the difference made by the screensaver--believe it is certainly under 5%, and most likely much less.
Possibly the real difference is in the graphics boards or their drivers, but just a warning--it can be substantial.