Hey, wait a minute.
When you’re in a hole, stop digging.

Get a supercomputer.

That’s how things looked last Monday at High Performance Computing on Wall Street.

 This show, a combined effort of Russ Flagg and Pete Harris, now at A-Team Group, has been a popular one despite a schedule that coincides with heads of state visiting the UN to give speeches. This year, the Pakistani ambassador must not have been staying at the Roosevelt, because there were no scanners at the door or big guys with earpieces. Maybe Pakistan was staying at the Marriott.

Anyway, the show maxes out at 800 and had more than that registered. What’s hot?

Duh.  Risk.

Although one could plausibly argue that compute power was not responsible for getting the American banks into this mess – it was sheer stupidity, oh all right, combined with greed. So long as the profits rolled in, very few followed the lead of Toronto Dominion’s CEO who decided if he couldn’t understand it he didn’t want to own it, or Goldman who apparently jumped out and maybe took the short side. More common were those who wanted to let the money roll in as long as it could, immortalized in the words of Citi’s Chuck Prince to the effect that as long as the music is playing we’ll keep dancing.

Who needed HPC to avoid that?

“Supercomputers” are just so 80s. Now the term is HPC and it is more than just a change in spelling The new approache relies strongly on clusters or grids, uses industry standard processors from AMD or Intel and runs on commonly understood operating systems such as Linx and Windows.  They can be implemented relatively quickly and sometime run on the floor — as the New and Improved and Rejuvenated Crays can do…see below.

Filed under: Technology