Cash???
- Thread starter Dantrag
- Start date
- Status
vassil3427
Posts: 633 +0
You can think of it as a short term memory on the processor. The more the better...
Cache is a very fast memory used to store very frequent accessed information and routines to speed up a process..
It's normally emmbeded inside the same chip, or can also be connected through a high speed bus... (embedde = the memory transistors are in the same chip as the processor core logic)
See here a chapo example:
Let's say you are the processor and you are currently working in a the actual application that's running, let's put counterstrike.
The L1 cache is limited to the instruction set that the processor carries, and is used to access with very little delay to functions that are necessary to run the application. (load from memory, unload from memory, acces pci devices throught the system bus..etc.) Normally this cache is quite small (16-128kb max). Although 128 kb L1 cache is only seen in very high end cpu's. Normally works at the same speed of the cpu itself or at half of that. This memory is very expensive.
The L2 cache is a bit slower, cheaper (but still very expensive) than the L1 cache. It is also bigger in size. 128kb,256kb,512kb, 1mb, 2mb...
P4 have 512, celerons have 128/256 (depends on the model). High end Pentium 4 Xeons (commonly used for servers) have 1mb-2mb. You normally end paying double the price for a processor like that.
In the L2 cache the processor stores the information that has to access at high speed and without delay. This time this are information of the applications running.
Let's say that instead of storing that info in the main memory, the cpu has very fast access to that information, so it can process faster common tasks. It evades having to do the trip to the main system ram every time for a process that occurs very often in a thread.
I don't know if I explained myself to good... i'll look for a article about cache and put the link here
It's normally emmbeded inside the same chip, or can also be connected through a high speed bus... (embedde = the memory transistors are in the same chip as the processor core logic)
See here a chapo example:
Let's say you are the processor and you are currently working in a the actual application that's running, let's put counterstrike.
The L1 cache is limited to the instruction set that the processor carries, and is used to access with very little delay to functions that are necessary to run the application. (load from memory, unload from memory, acces pci devices throught the system bus..etc.) Normally this cache is quite small (16-128kb max). Although 128 kb L1 cache is only seen in very high end cpu's. Normally works at the same speed of the cpu itself or at half of that. This memory is very expensive.
The L2 cache is a bit slower, cheaper (but still very expensive) than the L1 cache. It is also bigger in size. 128kb,256kb,512kb, 1mb, 2mb...
P4 have 512, celerons have 128/256 (depends on the model). High end Pentium 4 Xeons (commonly used for servers) have 1mb-2mb. You normally end paying double the price for a processor like that.
In the L2 cache the processor stores the information that has to access at high speed and without delay. This time this are information of the applications running.
Let's say that instead of storing that info in the main memory, the cpu has very fast access to that information, so it can process faster common tasks. It evades having to do the trip to the main system ram every time for a process that occurs very often in a thread.
I don't know if I explained myself to good... i'll look for a article about cache and put the link here
- Status
