Tag Info

Hot answers tagged

20

Gentle reminder: Actually, to provide a better answer for the rest of the community, please don't say something like "Do not talk about the four types of memory", as even if you know it well, there may be a thousand and one citizens of the internet arriving here hoping for a collateral answer :) EDIT: Paging is the accurate term for the following action. ...


11

Due to PAE: http://en.wikipedia.org/wiki/Physical_Address_Extension AFAIK Windows client versions, while they support PAE, don't support physical addresses past the 4 GB mark (apparently due to various buggy drivers). Windows server, and Linux, do not have this restriction.


10

That caches are internals of processor. Some are shared between cores, some are individual, depends on implementation. But all of them are located on chip. Some details: Intel Intel® Core™ i7 Processor, taken here: A 32-KB instruction and 32-KB data first-level cache (L1) for each core A 256-KB shared instruction/data second-level cache (L2) for each core ...


8

There would be little practical purpose- ram's meant to be fast, and no matter how fast your network is, you'd get speeds slower than even a local drive (which could hold swap, which the system uses when it needs to free up ram). Firewire goes up to 800 Mbits per second. Even the crappier flavour of sata go up to 3Gbits/s (granted that's drive dependant). ...


7

The Linux kernel automatically caches files in memory for efficiency. This is not a bad thing. When running free -m you will see on the right the amount of memory used in the cache. If an application needs the memory the kernel will free some of the cache. You should not try and manage it yourself.


7

Yes, used for caching and buffers. Modern OS's try to be smart...if you have 8GB of memory just sitting idle, not being used for anything, it's being wasted. So as long as no application is requesting it, the OS finds other things to use it for to try and speed stuff up. Don't worry about it -- if you application needs the memory, the OS will instantly free ...


6

Lets start with this: I think latest SMP processors uses 3 level caches so I want to understand Cache level hierarchy and their architecture . To understand caches you need to know a few things: A CPU has registers. Values in that can be directly used. Nothing is faster. However we can not add infinite registers to a chip. These things take up ...


6

You'll want to use ulimit ulimit can be used to limit memory utilization (among other things) Here is an example of setting memory usage so low that /bin/ls (which is larger than /bin/cat) no longer works, but /bin/cat still works. $ ls -lh /bin/ls /bin/cat -rwxr-xr-x 1 root root 25K May 24 2008 /bin/cat -rwxr-xr-x 1 root root 88K May 24 2008 ...


6

It's unlikely that it's really XP or Windows 7. It's probably memory leaks in the programs you're running. It's possible Windows 7 may have better features for reclaiming memory, but I doubt that. Even if you're using the exact same programs between versions of Windows, I'm sure there are little distinctions that add up that aren't really inherited from the ...


6

If I'm not wrong, the Linux kernel caches specific pages of a file, i.e. not the entire file is loaded into the page cache. One tool which you can use to figure out whether some contents of a file is in the page cache is fincore from the linux-ftools project. While it doesn't display all the cached files on disk, it gives you a rough idea of what is loaded ...


5

It's been a long time since I learned this stuff, but here goes. When an operating system launches a process, it assigns it pages from the virtual memory table. The operating system is responsible for maintaining a map from the virtual memory table to real memory or to the swap space on disk. When a process gets killed, the OS doesn't just stop giving it ...


5

The operating systems listed in the question tags (Windows and OS X) implement virtual memory, where each process is given its own address space, which is then mapped to physical memory by the OS. These mapping tables are used in cleaning up memory allocations when a process is terminated, so memory is completely freed. Physical pages may be shared among ...


5

Part of your problem may not be so much that it won't swap. If I remember correctly, the 4 GB address space is split in half on Vista 32-bit (which I assume you're using from the behaviour specified), so each application has 2 GB and the kernel and drivers have the other 2 GB. That means it's probably not possible (without workarounds) to get more than 2 GB ...


5

Have you tried a few older tricks such as defragging your hard drive? Apart from this, the only thing I can think of is you either missed something when looking at applications (My favourite tool is Microsoft / Sysinternals Autoruns) or it is a driver update / automatic update of some program somewhere that is causing problems. The registry thing is pretty ...


4

I think you should really give Soluto a try. Soluto is in beta and is free. It lists exactly how long each process and service take from your start-up, and recommends what you should do with each. As full disclosure, I work at Soluto, but I'm not ashamed to offer our first feature as a solution to your problem (I'm a long time user here and have never ...


4

The default policy of the kernel is to allow applications to keep allocating virtual memory as long as there is free physical memory. The physical memory isn't actually used until the applications touch the virtual memory they allocated, so an application can allocate much more memory than the system has, then start touching it later, causing the kernel to ...


4

RAM is also used to cache frequently accessed files. Both free and inactive RAM can be used immediately if needed. Unless Activity Monitor shows a high and steadily climbing page out value in the System Memory tab, it's no cause for concern. For example, my system has rather little free RAM, but my page out count is very low, so 4GB of RAM are definitely ...


4

SuperFetch, when enabled, will start preloading all of your most recently used programs into unused memory, up to your memory capacity. With 24GB of RAM, it will act as a ridiculously large "cache" and it'll greatly speed up your performance. However, it will NOT do anything to slow your computer down, since if a program needs space claimed by SuperFetch, ...


4

These options are intended for device and driver developers in particular. If in doubt do not touch as these values can dramatically affect your machine's performance in the negative direction. In general the default configuration will always be the most performant for your machine setup. But to answer your questions specifically, here is a break down: ...


4

Windows 7 has a built in memory tester that you can select from your boot menu. For most part, i've tended to rely on memtest x86+ livecd for checking systems, since by the time i figure i need to test the ram, everything else is already out. I'd also probably try setting my pagefile to zero, then back to automatic (to clear it out), and running sfc ...


4

Ubuntu upon detecting 4GB+ RAM automatically uses kernel with PAE extension, which supports up to 64GB RAM. Related question on Ask Ubuntu: http://askubuntu.com/questions/43422/8-gb-ram-on-64-bit-processor-using-32-bit-ubuntu


4

Issues like this is why we use OSs, it provides a layer of abstraction so you do not need to concern yourself at the application level about things like how the memory is stored on the hardware level. The layers are similar to this (for HDD drives), Each layer shows next to it who is responsible for that layer. File (File system) | ...


3

This is not a Linux issue, it's a Java issue (and completely normal behaviour). The same will happen on Windows, though I guess it may be more eager in swapping out part of those Tomcat processes. First of all, two instances of Tomcat are going to start two instances of the JVM in two separate processes. The question is whether you actually need two ...


3

There is currently no easy way of adjusting the swappiness (or so it is called) behavior of macos X. There are a few hacks available though (requires developer account & SDK): http://cestdelamerde.com/archives/22-Killing-Mac-OS-X-Swapping-How-To-Disable-dynamic_pager.html http://dropsafe.crypticide.com/article/3848 Good luck! Postscript. I guess you ...


3

The details are complex; in a single processor it is simple enough to implement some equivalent of "lock, modify, unlock" at the microcode level - or other techniques. Once you have multiple processors the subject gets complex, especially in view of cache effects. Protocols like MSI, and derivatives MESI, MOSI, MOESI, back this in modern Intel processors. ...


3

Well there is NUMA. When it is used, every core gets a piece of RAM assigned to it. This way exactly what you described happens. One processor will work with RAM assigned to it and another will work with different part of RAM. Some motherboards will not even detect RAM unless there are enough CPUs present to use it all. I don't know if this is the case with ...


3

The SHR column simply represents the amount of shared memory used by the process. It can be used for interprocess communications, but a more common scenario is that this is memory used by shared libraries that an application has linked in. From The Linux Kernel: Memory Management: Shared Virtual Memory Although virtual memory allows processes to ...


3

I just did an unscientific experiment and the answer is Close the tab then open a new one, but we are talking about marginal. My Firefox takes up 242.7MB, I opened up a new tab with Google and it went to 244.9MB, I then closed the tab and it went to 242.8MB. I then opened up Microsoft.com in a new tab and it went to 248MB, Closed it and went back to ...


3

Cache is almost always on chip for fastest access. Here is a nice diagram showing a quad core Intel CPU die with the L3 cache highlighted. When you look at pictures like this of a CPU die, large uniform areas are typically banks of on-chip memory used as a cache.



Only top voted, non community-wiki answers of a minimum length are eligible