Ok, 20 laps of the Sun it is ! That's 20 waxes and wanes of the signals which we demodulate from at least that general pattern giving a pure solar system barycentre perspective. And what remarkable finds they have been ! Does anyone have their favourite signal ? For me it is the black widow system PSR J1653−0158, a gamma ray (GPU accelerated) discovery.
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
Right. I had thought it was early 2025 but I can't show that. Either way it's a grand achievement!
I think you're quite right about early 2005 for the 'open to all' official project.
The 15th October 2004 date is when both Bernd and Bruce Allen (Director) had the initial accounts on the test project. This wasn't open to the general public for a time while a relatively small number of volunteers helped debug everything in preparation for a later official opening to all comers.
On 9th Feb 2005, Bruce posted this message which probably announces the real start of the project proper. In that message he said,
"I just sent out approximately 6500 account keys. This should include everyone who registered at the APS World Year of Physics Web site prior to January 20, 2005, AND who indicated willingness to be a tester."
My understanding is that the test project was deemed to be ready enough to go live when he posted so I guess somewhere between Jan 20 and Feb 9 2005 could best be considered as the actual start of the project proper.
Of course, I have a vested interest in the date of 9th Feb 2005 being declared as the 'official' project start date since that happens to be when I joined :-).
I just noticed a few days ago that I have been working on E@H for more than 20 years now. I seriously started working on a BOINC version and Windows port of our main application in July 2004, mainly with Bruce Allen (overall management and server code), Maria Alessandra Papa (application code), Reinhard Prix (BOINC API, Linux & Mac builds), all at the AEI in Potsdam in close-by offices, for a few months even joined by David Anderson, and David Hammer managing the servers at UWM in Milwaukee.
October 15 2004 might well be when we set up a new server from scratch for the first somewhat public one, but internally it must have been the second or third instance (FWIW UserID#1 is David Hammer, who is no longer working @UWM). Registration went by email, i.e. we kept a list of people that showed interest and invited them by sending them an email with the "authenticator" (username/password authentication was much later). IIRC we invited 100-1000 people per week, just what we thought the server could handle.
Porting the application to Windows was difficult, but had a huge benefit: Due to the memory model Linux is quite forgiving regarding wrong memory accesses. Windows isn't. I fixed tons of bugs in the application that went unnoticed before when it was running just under Linux, and crashed it under Windows.
Originally it was planned to launch E@H (i.e. open self registration) with the beginning of 2005, the APS "World year of physics" (of which E@H was one of the four "flagship projects") and the "Einsteinjahr" in Germany. But it was mainly the application stability that kept us hesitating until February 19. At that time Bruce - back in the US in 2005 - called me literally every day urging me to give my ok for the public launch.
Due to a lot of work that anniversary went almost unnoticed so far. We are currently discussing when and how to celebrate it.
Thank you all for sticking with us for so long, for all your help and contributions!
Porting the application to Windows was difficult, but had a huge benefit: Due to the memory model Linux is quite forgiving regarding wrong memory accesses. Windows isn't. I fixed tons of bugs in the application that went unnoticed before when it was running just under Linux, and crashed it under Windows.
Isn't that just the way of things. Run it along a Windows road and see it crash. Well done !
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
... And at the first user meeting of E@H in autumn 2008 (?) at AEI in Potsdam/Golm, at the end, Bernd showed us the server, situated in a big room in the cellar, consisting of about 100 normal PCs, they bought from a normal PC seller, in normal racks and separated from the remaining room by plastic foils for reducing admittance of dust. Bernd said, they began with 60 PCs, and when he started E@H, he was the first member. Then other members of the institute followed. And he said, they are planing a bigger computer, which ended up in the ATLAS, some years later, and in Hannover.
Thanks for the reminder!
To be precise:
The computer cluster you saw must have been "Merlin". This cluster was built in 2002, and it was the third (or so) data analysis computing cluster that Bruce built. To work with it, two people were hired: an admin (Steffen Grunewald) and a scientific software developer (me). This cluster was build mainly to analyze data from GEO600, the AEI's own GW detector (located near Hanover). The relation to Einstein@Home is mainly personnel - Steffen wrote the first validator for our application, and he set up the network of download mirrors (at AEI, Glasgow, UWM and CalTech). And me - well, you know.
If you saw Merlin in the basement, this means that the second wing of the AEI Golm building was just finished, and Merlin was moved there. Previously it had been in a larger office, located on the second floor of the older wing. The windows were blinded with copper tinfoil to prevent the sun from further heating the room. Terrible to work in there.
Also too in that room in the basement must have been the AEI download mirror (a single 19" rack machine that may go as "Einstein@Home server"), and possibly a few (desktop) machines to compile and test the applications for Linux, Mac and Windows. The core E@H servers (web-, database- and download-server) were located at UWM and managed by David Hammer.
The planned successor to Merlin was "Morgane" (Merlin's sister), the first cluster of 1U rack servers that we built and used for data analysis from ground-based GW detectors (LIGO and GEO600). I think Morgane is no longer in operation either.
Judging from the above this user meeting must have been well before 2008, possibly 2006. In 2007 Bruce became director of the AEI in Hanover, and took me with him.
Edit: I just found this page about the event. According to it, the event took place on Dec 18 2006. For people interested in the history of E@H, I recommend Bruce' talk, which is still available from the linked archive page.
What a great recording of Bruce's talk, makes the powerpoint easier to follow. I especially note that a reviewer of the early ideas had thought them "too well conceived and organised" to receive a grant !? Such is the minefield of funding applications I suppose ... :-0
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
Thanks Martin for reminding
)
Thanks Martin for reminding us of the anniversary.
Ok, 20 laps of the Sun it is
)
Ok, 20 laps of the Sun it is ! That's 20 waxes and wanes of the signals which we demodulate from at least that general pattern giving a pure solar system barycentre perspective. And what remarkable finds they have been ! Does anyone have their favourite signal ? For me it is the black widow system PSR J1653−0158, a gamma ray (GPU accelerated) discovery.
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
Mike Hewson wrote: Right. I
)
I think you're quite right about early 2005 for the 'open to all' official project.
The 15th October 2004 date is when both Bernd and Bruce Allen (Director) had the initial accounts on the test project. This wasn't open to the general public for a time while a relatively small number of volunteers helped debug everything in preparation for a later official opening to all comers.
On 9th Feb 2005, Bruce posted this message which probably announces the real start of the project proper. In that message he said,
"I just sent out approximately 6500 account keys. This should include everyone who registered at the APS World Year of Physics Web site prior to January 20, 2005, AND who indicated willingness to be a tester."
My understanding is that the test project was deemed to be ready enough to go live when he posted so I guess somewhere between Jan 20 and Feb 9 2005 could best be considered as the actual start of the project proper.
Of course, I have a vested interest in the date of 9th Feb 2005 being declared as the 'official' project start date since that happens to be when I joined :-).
Cheers,
Gary.
I just noticed a few days ago
)
I just noticed a few days ago that I have been working on E@H for more than 20 years now. I seriously started working on a BOINC version and Windows port of our main application in July 2004, mainly with Bruce Allen (overall management and server code), Maria Alessandra Papa (application code), Reinhard Prix (BOINC API, Linux & Mac builds), all at the AEI in Potsdam in close-by offices, for a few months even joined by David Anderson, and David Hammer managing the servers at UWM in Milwaukee.
October 15 2004 might well be when we set up a new server from scratch for the first somewhat public one, but internally it must have been the second or third instance (FWIW UserID#1 is David Hammer, who is no longer working @UWM). Registration went by email, i.e. we kept a list of people that showed interest and invited them by sending them an email with the "authenticator" (username/password authentication was much later). IIRC we invited 100-1000 people per week, just what we thought the server could handle.
Porting the application to Windows was difficult, but had a huge benefit: Due to the memory model Linux is quite forgiving regarding wrong memory accesses. Windows isn't. I fixed tons of bugs in the application that went unnoticed before when it was running just under Linux, and crashed it under Windows.
Originally it was planned to launch E@H (i.e. open self registration) with the beginning of 2005, the APS "World year of physics" (of which E@H was one of the four "flagship projects") and the "Einsteinjahr" in Germany. But it was mainly the application stability that kept us hesitating until February 19. At that time Bruce - back in the US in 2005 - called me literally every day urging me to give my ok for the public launch.
Due to a lot of work that anniversary went almost unnoticed so far. We are currently discussing when and how to celebrate it.
Thank you all for sticking with us for so long, for all your help and contributions!
BM
Bernd Machenschalk wrote: I
)
And we thank YOU for all of your hard work and keeping up with Einstein@Home for 20 years!
HooRah!! Congratulations!
Proud member of the Old Farts Association
Bernd Machenschalk
)
Isn't that just the way of things. Run it along a Windows road and see it crash. Well done !
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
Dang, I missed my 15th
)
Dang, I missed my 15th anniversary last year ;) So yeah, let's pop some corks soon...
Einstein@Home Project
Oliver Behnke wrote: Dang, I
)
I'm ready anytime! ;*)
Proud member of the Old Farts Association
astro-marwil wrote:... And
)
Thanks for the reminder!
To be precise:
The computer cluster you saw must have been "Merlin". This cluster was built in 2002, and it was the third (or so) data analysis computing cluster that Bruce built. To work with it, two people were hired: an admin (Steffen Grunewald) and a scientific software developer (me). This cluster was build mainly to analyze data from GEO600, the AEI's own GW detector (located near Hanover). The relation to Einstein@Home is mainly personnel - Steffen wrote the first validator for our application, and he set up the network of download mirrors (at AEI, Glasgow, UWM and CalTech). And me - well, you know.
If you saw Merlin in the basement, this means that the second wing of the AEI Golm building was just finished, and Merlin was moved there. Previously it had been in a larger office, located on the second floor of the older wing. The windows were blinded with copper tinfoil to prevent the sun from further heating the room. Terrible to work in there.
Also too in that room in the basement must have been the AEI download mirror (a single 19" rack machine that may go as "Einstein@Home server"), and possibly a few (desktop) machines to compile and test the applications for Linux, Mac and Windows. The core E@H servers (web-, database- and download-server) were located at UWM and managed by David Hammer.
The planned successor to Merlin was "Morgane" (Merlin's sister), the first cluster of 1U rack servers that we built and used for data analysis from ground-based GW detectors (LIGO and GEO600). I think Morgane is no longer in operation either.
Judging from the above this user meeting must have been well before 2008, possibly 2006. In 2007 Bruce became director of the AEI in Hanover, and took me with him.
Edit: I just found this page about the event. According to it, the event took place on Dec 18 2006. For people interested in the history of E@H, I recommend Bruce' talk, which is still available from the linked archive page.
BM
What a great recording of
)
What a great recording of Bruce's talk, makes the powerpoint easier to follow. I especially note that a reviewer of the early ideas had thought them "too well conceived and organised" to receive a grant !? Such is the minefield of funding applications I suppose ... :-0
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