My intro

[joelovesliving]

My name is Joe Daly. I am working at Intel to develop solutions to compute and data movement issues realted to the volumes of data related to short read sequencing.

[cariaso]

Brian Bramlett at Intel has been working on for visualizing Illumina data for quality control. I've seen some wonderful screen shots of his work, but don't have permission to share them. I think they would be of interest to many on this site. Perhaps you are connected to him or that work?

[apfejes]

Hi Joe, Your introduction sounds interesting - would you be able to tell us a little more about your work? What compute projects are you working on?

[joelovesliving]

Brian Bramlett at Intel has been working on for visualizing Illumina data for quality control. I've seen some wonderful screen shots of his work, but don't have permission to share them. I think they would be of interest to many on this site. Perhaps you are connected to him or that work?

Hi Mike,
We are still trying to figure out how to release Brian's work. Let's get a phone call for the three of us to talk.

Joe

[joelovesliving]

Hi Anthony,
We are trying to address a few issues for labs who are new to short read sequencing, or labs who don't have budgets for large support staff. The first objective is a turn key startup which integrates a baseline level of storage, compute, cluster technology, and an application toolbox so startup is seemless. Secondly, we are selectively applying Intel QuickAssist front side bus accelerators to improve the compute/[$$, cubic foot, watt] performance. As you stated in your blog, we are moving cautiously because many of the applications are still very fluid. It is also just difficult to stay ahead of Moore's law for over a long period of time, so the selection process needs to focus on those handful of things that have unlimited compute demand. We also hope to make the process of finding and installing bioinformatics apps very painless. By getting a platform that is based on open technology, and getting enough volume, we expect to stabilize execution targets for developers. I'm in Portland, OR so it woud be easy to have a phone follow up.

Thanks, Joe

[apfejes]

Hi Joe,

Thanks for the excellent summary - I appreciate you sharing that information with us.

I also very much appreciate what you're saying about helping new groups to get set up with the technology. Although the GSC is beyond that phase now, there are clearly some major hurdles to getting everything in place. I'm somewhat curious about the scalability of what you're working on, in terms of whether you expect it will be useful for large groups as well. (Although I understand getting an established group to switch would be much more difficult than getting in at the ground level.)

I hope I'm not asking too many questions. - and thanks again for taking the time to reply! (=

[bioinfosm]

Joe,
I would be curious too about the scalability of your start-up solution. We have had 1 solexa for a year, being upgraded to GA2, along with a new GA2. So we have cleared some of the startup hurdles, only to catch the newer ones and work on overcoming them