Conclusion

The fact that storage systems can't keep up with soaring CPU speeds is a major problem for computer manufacturers. There is intense pressure to find ways to minimize the number of wait states that occur in normal system operation. The principal response to date has been to use judicious amounts of cache memory—very fast but very expensive—in between the CPU and regular memory. There is good news and bad news here. The good news is that, as transistors and traces continue to shrink, it is possible to have more of the cache that sits on the same chip as the CPU. This type (L1) is by far the fastest because it can operate on the same clock as the CPU and because it can use a very wide path without experiencing skew or interference. The ...

Get Essential Guide to Computing: The Story of Information Technology, The now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.