Re: 1teraflops cell processor possible?

> As it is, Mr. Myers (again, no relation) is being very kind about
> those fragile wings.  You can put more engineering effort into the
> foam insulation on the external fuel tank, but shuttle tiles have
> fallen off in previous flights, and they will continue to fall off if
> the current shuttle thermal management system continues to be used.

Columbia's problem, to a large degree, was precisely that the tiles
were _not_ compromised. In fact, in all tests the assessment, during
the mission, that the analysis program (massively) overestimated tile
damage was vindicated. _All_ burn-through problems - and there historically
were some before Columbia's last mission - were associated with the
wing leading edge, RCC and its support structure. The tile really _was_,
as Linda Ham said, "only a maintenance issue". But everybody had thought
about problems with the tiled areas, for various (good, historical) 
reasons; very few people had considered possible problems with RCC, and
what little analysis had happened there was deeply flawed.

And, BTW, the wings were anything but fragile: the analysis of in-flight
data showed that Columbia's left wing kept flying until all that was left
was a sagging, hollow shell. You can find traces of the structural engineers'
suprise that it held on for so long even in the quite terse CAIB report.

        Jan