Archive for December, 2010

Outlook 2007 upplevs segt och andra suspekta fel.

No Comments »

Well

Som admin kan livet vara hårt ibland när man brottas med problem hos kunder.
Att datorer “aldrig kan gå för snabbt” vet vi ju alla om och när problemet slår till så är det lätt att leta efter en nål i en höstack.

Vanligast är ju att man tänker att virusprogrammet har rullat ut en dålig uppdatering/patch från servern eller att nätet är segt (många kör vpn) och i alla dess otaliga former.

En tråkig uppdatering som uppdagades relativt fort som har undgått mig är KB2412171.
Så kort och gott, har man problem med Outlook 2007 så skall denna på anmodan av MS avinstalleras.

Lite länkar som pekar ut mer information om detta.

http://www.howto-outlook.com/news/kb2412171.htm
http://social.answers.microsoft.com/Forums/en-US/officeinstall/thread/e8273e72-b9f6-4f44-b621-c04d0553427f
http://support.microsoft.com/kb/2412171


Cuda efficiency at seti@home

No Comments »

Hi everyone.

For newcomers at seti@home i’m called Vyper and i’m the creator and member of a team called GPU Users Group. I’m also a beta tester over at Lunatics website.

The main reason i did this was to compare different Cuda versions from 6.09 and onwards during the extended down time of the seti@home servers. I did this because i wanted to show numbers of what we die hard crunchers already know but hasn’t got the efficiency of it presented.

So therefore i did some thorough testing and made a spreadsheet of all the details so we all can benefit of it.
Please be aware of there are old unreleased test software that hasn’t been released to the public but the outcome of all the tests has resulted in a highly optimised client that has evolved significantly in terms of throughput and stability.
Final versions is downloadable over at the download section at Lunatics downloads section.

If you start to read the document you’ll notice some huge variations in speed and that depends on client crashes which results in shorter times but has invalid data. I didn’t erase that, instead i’m mentioning this here.

At page 22 you see that one executable is faster than others to a certain amount. Rest assure that is due to crashing clients which couldn’t be run at the same time properly. So the fastest executable at page 23 is not fast in real life but broken 🙂

Well click on the document to see the full data.

Cuda efficiency

//Vyper