Just buy them an 8088 with an 8087 math co-processor
There is an engineering term called "scope creep" -- when your idea gets bigger and fancier until it is impossible to complete.
We are seeing this with climate models -- there is enough chaotic action in our atmosphere that a given model will yield wildly different results each time it is run. What a lot of the 'warmists' do is cherry pick their data and then cherry pick their results. They get more grant money that way.
The British Royal Meteorological Society's director has a begging bowl in his hand --
from the Beeb:
Lack of computing power 'hindering weather forecasting'
Prof Paul Hardaker, head of the Royal Meteorological Society, was speaking to the Commons science committee.
He said more supercomputers were needed to carry out complex calculations, but finding the money was "an issue".
However, the potential economic benefits of more accurate forecasts were "enormous", the MPs heard.
The science and technology committee is carrying out an inquiry into the use of science by the Met Office and the effectiveness of the body as a whole.
Emphasis mine -- given the stunningly poor predictions they have been giving in the last ten years, it's amazing that they have any funding given them at all. If they were working in the real world, they would have gone bankrupt years ago -- meanwhile, private for-profit businesses like
WeatherBell are thriving.
More:
But Prof Hill, Prof Hardaker and a third witness - Prof Alan Thorpe, the director general of the European Centre for Medium-Range Weather Forecasts - all agreed a lack of supercomputers was a major problem.
"I recognise there's an affordability issue and we have to make priorities, but it is a significant limitation in our capability at the moment," Prof Hardaker said.
As Steve Martin used to say:
Well, excuuuuuuuuse Me!
Looking at the Top 500 list of worldwide supercomputers, the
United Kingdom Meteorological Office has two supercomputers rated at #'s 279 and 280 in the list of the worlds most powerful. That is not shabby. The Top 100 systems are either classified, computer manufacturer research systems, nuclear, geophysics (oil) or university systems -- the idea that a simple Met office would need its own system of that magnitude is fucking hubris on their part.
Brings to mind the proverb:
a bad workman always blames his tools
Posted by DaveH at October 30, 2011 4:03 PM