Best storage for efficent retrieval

Question

I'm working on a project where we develop an access-application which uses share price data for approximately 10 years and over 1000 shares (this is just to indicate the size). The most critical part is the retrieval of data which need to be as fast as possible, and I wonder what methodology you recommend. As it is now, we have sorted the shares according to country and then we have a separate row for each date (and a column for each share). This takes forever to retrieve (we use all data at once), so I wonder if you know of any alternative solutions (like storing all quotes for a stock in some binary form or something).
Thanks in advance,
K J

Answer 1

"Retrieval of data" -- this isn't very precise, isit?
No, but "we use all data at once" is a bit more
precise, don't you think? To minimize the number of
times we need to query the database, we put all data
into float arrays. Right, that's why I mentioned the summarized datasets idea. Which is what your float arrays are, in a way. You could have the process that updates the detailed database table also update a summary table whose structure corresponds to your float arrays, perhaps.
Have a look at Ralph Kimball's book "The Data Warehouse Toolkit", you might get some ideas from it.