drummyfish/less_retarded_wiki
5
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Update

Browse source
This commit is contained in:
Miloslav Ciz 2023-12-26 14:46:41 +01:00
parent 0d8f2a5fea
commit b35242f273
17 changed files with 58 additions and 31 deletions
Download patch file Download diff file Expand all files Collapse all files

2
cache.md
View file

@ -10,7 +10,7 @@ A cache is related and/or exploits some observations and concepts related to com
- **principle of locality**: Computers (/CPUs) tend to more often than not access data that are close to each other in memory, i.e. a CPU doesn't typically make [random](randomness.md) jumps in memory but rather e.g. reads a sequence of bytes one after another from an [array](array.md) or [struct](struct.md). For this reason when a CPU pulls something out of memory, there is a high probability of accessing an address that is nearby to this memory next time -- a cache helps us get ready for this by prefetching this nearby data and having it ready for very
fast access.
- **[memory](memory.md) hierarchy**: Mostly because of the principle of locality computer memory is divided into different levels, a chain of memories that get progressively further away from the CPU, increasing their size (decreasing price for capacity) as they get further away but also decreasing their speed. Here a cache can be seen as the closest memory to the CPU, i.e. being the smallest, most expensive but also fastest memory. By extension we can see that RAM can in many cases be seen as a "cache" for the hard drive, hard drive can be seen as "cache" for the network (after all web browsers ARE caching websites into files on the disk) etc.
- **[memory](memory.md) hierarchy**: Mostly because of the principle of locality computer memory is divided into different levels, a chain of memories that get progressively further away from the CPU, increasing their size (decreasing price for capacity) as they get further away but also decreasing their speed. Here a cache can be seen as the closest memory to the CPU (except for the [registers](register.md)), i.e. being the smallest, most expensive but also fastest memory. By extension we can see that RAM can in many cases be seen as a "cache" for the hard drive, hard drive can be seen as "cache" for the network (after all web browsers ARE caching websites into files on the disk) etc.
- **[dynamic programming](dynamic_programming.md)**: Dynamic programming is a programming technique revolving around remembering already calculated results so that we don't have to compute them again in the future -- this is basically what caches do, they remember results we obtained in relatively expensive way so that next time we can get them cheaper.
- ...