We had a busy day on the floor of SC|06 today. Lots of visitors wanting to see demos of Sun Studio, and especially the new Thread Analyzer that will be in Sun Studio Express 3 (to be released in December).

And the response was pretty positive. We gave away a lot of DVDs with Sun Studio 11 and Express 2.

One thing we noticed that was different since last year's SC|05 in Seattle was that more people knew about Solaris and Sun Studio this time. And there were just as many people talking about having Solaris on their systems as Linux.

One researcher I talked to had an important C/C++ application he needed to deploy on a new Xeon desktop system in his lab. When I asked him whether he was using Solaris or Linux he was quick to reply that it didn't matter. He was willing to use either one. So I handed him a Solaris DVD and the Sun Studio DVD and he seemed quite happy.

We're seeing more of this sort of "agnosticism" regarding operating system. Now that Solaris is as free and open source as Linux, it doesn't really matter which OS is the target platform. Especially with the Sun Studio Express compilers running on x64/Linux.

Another crowd pleaser were the SDN Developer Forums at http://forum.sun.com/jive/index.jspa?tab=devtools . I was able to show how questions about Sun Studio get answered quickly either by the developer community at large, or by our Sun Studio software engineers around the world. It was really impressive.
But the most impressive thing I saw was at the university and national labs exhibits. In the theater part of the "e-science" booth they were showing a BBC TV video about climate change that  really  caught my attention. It showed that many of the  climate models are showing a disastrous rise in global warming that cannot be attributed to cyclic changes in sunlight or atmospheric conditions as we have seen them in the past. But how much warming? The models vary widely from 2 to 6 degrees C over the next 100 years. The problem is that trying to predict things over such a long time span results in very wide variations depending on minor differences in initial conditions. So one plan is to run thousands of climate models, with slightly differing assumptions, and see how they cluster in the aggregate.

Nice idea, but there isn't a single supercomputer center on earth with that much available time to run so many models. So one researcher hit on the idea of using the world's idle computers in a collaborative distributed network, sort of like the way the search for ET, the SETI project, is being run.

And this is now happening. And the BBC is involved. You can go to http://bbc.co.uk/climatechange and download a client program for your Windows or Linux PC (no Mac client just yet) and participate in this global experiment.

The BBC Video, tltled MELTDOWN, aired on BBC Four in February, and was quite engaging. I didn't think I would, but I actually sat thru the whole thing, glued to every word. In the video the well known British geographer Paul Rose (http://www.paulrose.org/index.html) tries to find out why the ice flows that he studies in Greenland have been shrinking. Is is global warming caused by human activity, or is it a cyclic thing that has happened before. Like Al Gore's movie An Inconvenient Truth, the BBC program gives ample scientific evidence that this is no cyclic event, and that the dynamics that caused cyclic events in the past, like the "little ice age" that hit Europe around 1816, but something far more severe than anything the earth has ever seen.

The "e-science" organization is a consortium of British universities that combine their computing resources over a grid network to do important scientific research. One of their activities is climate prediction, and you can particpate in that research. Go to www.climateprediction.net for more information. And for more information on the e-science global grid, see www.nesc.ac.uk.

I probably spent too much time in their booth talking to the climate prediction folks and looking at their displays. But ever since I spent a year at the European Weather Centre (ECMWF) in 1977, I've been fascinated by computational meteorology. And now I may get a chance to run my own climate model and participate in this new global grid. I plan to download the client program and get it running on my AMD64 Acer Ferrari laptop.

Today was my last day at Supercomputing 06. Tomorrow I fly back to the Bay Area. This was quite a show. Met lots of old friends from the good old days of scientific computers, and made some new ones. Quite a lot to assimilate. I'll probably have more to say after I read thru all my notes. Next year it will be in Reno, Nevada!