Account | Einstein@Home

Gary Roberts

Member since: 09 February 2005

Country: Australia

BOINC ID: 12521

Background

After many years of working in a University environment, I spent the last 8 years of my working life doing the same work, but as the principal of my own business. The business prospered and progressively acquired a number of computers, so I was able to get involved in Seti Classic just after it first started. By the time classic finished and the BOINC era had started, I had become hooked on the concept of DC but was looking for something with a bit more of a mainstream fundamental scientific research theme in the area of physics and astronomy. In early 2005 I joined Einstein@Home just before it went public. In mid 2005 I joined LHC@Home. Finally, in early 2009 I joined Milkyway@Home.

Initially, I used my business computers but I was able to supplement their output by using computers belonging to business associates. In 2006 when I was starting to wind back my business interests, I decided to expand my own fleet and not rely on the generosity of others. I was quite a fan of the Tualatin Pentium III architecture at that time and was able to purchase bulk lots of ex-business hosts very, very cheaply. The biggest expense was the ongoing running costs. At the height of this expansion, I had around 200 hosts running, all fully funded out of my own pocket.

The Tualatin crunching period lasted well into 2008 when I started shutting down large numbers and upgrading the rest. I kept the case, optical drive and HDD and put in new motherboards, CPU, RAM and PSUs. I started with Q6600 quads but ended up with larger numbers of Q8400 quads when they got cheap enough. I also used quite a few pentium (E6300) dual cores which perform quite well and overclock very easily. Over a period, I had scaled back from over 200 to around 60 machines that cost less to run but had a much higher output. In 2009, I joined Milkyway. When their GPU app was launched, I added a number of ATI 4850 GPUs to existing hosts to do the crunching and continued using the CPUs for Einstein. I discontinued the Milkyway CPU app on all hosts.

In 2011, I added some more hosts which were owned by my daughter (as well as a few more of my own). She runs a commercial real estate business and is a great fan of Apple iMacs. All those were set up crunching on Einstein and the staff who use them each day don't seem to notice the extra load. Most of my own hosts were running Linux so I felt quite at home with OS X and managing BOINC on it.

In 2012/2013, when the Einstein GPU app had matured, I purchased a number of mainly nVidia GTX650 GPUs which seemed to give about the best 'bang for buck' at that time. I've since purchased some AMD HD7770s and HD7850s which, with the latest drivers, seem to be even better. I built a number of new Sandy Bridge or later generation hosts to take the new cards, after finding that some older motherboards didn't allow the full performance to be achieved.

Opinion

Many years ago I did an Engineering Degree at the University of Queensland and Physics was part of the curriculum in years 1 and 2. I did extremely badly at Physics. Most Engineers at the time considered Physics to be somewhat impractical and just wanted to get on with the real stuff of Engineering. Theoretical concepts such as "ripples in the fabric of space-time" or "the dual wave/particle nature of light" were seen as relatively unimportant when judged against the more practical concepts of Engineering.

These days I reckon I might have mellowed a bit and I've become interested in what's happening with fundamental research. To me, it is quite important to support those who have a chance of making the discoveries that just might really allow us to understand this enigmatic universe we find ourselves in. It is too depressing a thought to imagine we are marooned on this speck of dust we call home without a practical means of ever realistically being able to escape from our solar system. We should be able to colonise other planets or moons but to even just visit a nearby star system seems impossible, let alone the rest of the galaxy. So I've chosen to support projects which might increase our knowledge and understanding of the universe. The hope is that some of these endeavours may help to discover new physics and with that, new possibilities for eventually exploring the cosmos.

Einstein is my favourite project for a number of reasons. The project has obvious scientific merit. It always has work. The staff are very responsive and make every effort to look after the needs of the volunteers. Worthwhile discoveries are being made - the volunteers can see progress. When problems or malfunctions occur, the project has the funding and resources to deal with them appropriately and professionally. You know the project is here for the long term, with a decided sense of purpose about what it is doing.

When the project was first launched, Prof Allen worked tirelessly to get problems fixed, to keep people informed, and to help people who were having difficulty getting started. Such 'leading by example' always encourages volunteers to continue their participation.