Thread: Your views on computer specs

  1. #16
    Make Fortran great again
    Join Date
    Sep 2009
    Posts
    1,413
    What's the name of this conglomerate? So you're saying that like PNY Kingston Corsair PATRIOT and every other brand from expensive American made to cheapy Chinese made is really from the same entity?

  2. #17
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,229
    I thought there are a few memory chips manufacturers - Micron, Microchip, Qimonda, Hynix, etc.

    What's the name of this conglomerate? So you're saying that like PNY Kingston Corsair PATRIOT and every other brand from expensive American made to cheapy Chinese made is really from the same entity?
    They buy memory chips from those companies above, and slap them onto their own PCBs, and use their own branding. If you look at your memory modules (may need to take off the heatspreader if it's a high end module), you can probably see the chip manufacturer's name on the actual memory chips. Different modules from different manufacturers that use the same chips tend to behave very similarly (in terms of overclocking potential, and how well it responds to increased voltage).

  3. #18
    Registered User jeffcobb's Avatar
    Join Date
    Dec 2009
    Location
    Henderson, NV
    Posts
    875

    BFO on multithreading

    Quote Originally Posted by abachler View Post
    Heh, I've been doing multithreaded/multiprocessor code since the 486 days. .
    Yeah I used to think I was doing multithreading too prior to the advent of dual-core/dual-proc machines. Then I got my first one and had to troubleshoot a thread contention issue where two threads really WERE executing and it was a whole new ballgame.....a completely new class of bugs can manifest with multiple processors than what you see on a single-proc machine with threads....since then I don't touch anything with less than two cores (for development that is...my netbook is single-proc but I only use that for doing one-off projects or entertainment)....
    C/C++ Environment: GNU CC/Emacs
    Make system: CMake
    Debuggers: Valgrind/GDB

  4. #19
    Malum in se abachler's Avatar
    Join Date
    Apr 2007
    Posts
    3,195
    Quote Originally Posted by jeffcobb View Post
    Yeah I used to think I was doing multithreading too prior to the advent of dual-core/dual-proc machines. Then I got my first one and had to troubleshoot a thread contention issue where two threads really WERE executing and it was a whole new ballgame.....a completely new class of bugs can manifest with multiple processors than what you see on a single-proc machine with threads....since then I don't touch anything with less than two cores (for development that is...my netbook is single-proc but I only use that for doing one-off projects or entertainment)....
    lol, no I mean real multi-threading/multiprocessing. There were dual CPU 486 servers you know. Multi-CPU motherboards where all the rage before the multi-cores came out.

  5. #20
    Registered User VirtualAce's Avatar
    Join Date
    Aug 2001
    Posts
    9,607
    Among the retardations that I'm hearing are gems like "XP is low-end, Vista is faster" (which has been proven to be false) and "Vista uses 1.5GB of RAM just from booting to the desktop" (yeah right).
    You don't really need to argue this point with whomever it is. The evidence is all over the internet that is simply not true. Vista on average uses 1.5 times more RAM for any given application than XP does. Just look at the requirements for Vista and XP on any boxed app and you will see they clearly recommend more for Vista. But I doubt it takes 1.5GB just to get to the desktop.

  6. #21
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,229
    I think that's because Vista adopted the "Linux-style" disk caching. I guess people would rather have free RAM than cached disk content in unused RAM...

    Many Linux beginners ask that question, too. After running for a few hours, almost all Linux installations will report close to 0 free RAM.

  7. #22
    Malum in se abachler's Avatar
    Join Date
    Apr 2007
    Posts
    3,195
    The problem is that Vista is too aggressive in allocating the 'free' ram to disk cache. This causes excessive disk thrashing.

  8. #23
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,229
    I'm not sure how Vista does it, but for Linux, as soon as an application needs the RAM, the cached content is thrown out. It won't touch swap until nearly all RAM is occupied by applications.

  9. #24
    spurious conceit MK27's Avatar
    Join Date
    Jul 2008
    Location
    segmentation fault
    Posts
    8,300
    Well, whatever, who cares. The thing most likely to get overheated around here processing on this machine is me.

    What I really want is this:

    Zenview Command Center Six-Screen LCD Monitors: Digital Tigers Multi-Screen LCD Displays: Six Monitors
    C programming resources:
    GNU C Function and Macro Index -- glibc reference manual
    The C Book -- nice online learner guide
    Current ISO draft standard
    CCAN -- new CPAN like open source library repository
    3 (different) GNU debugger tutorials: #1 -- #2 -- #3
    cpwiki -- our wiki on sourceforge

  10. #25
    Malum in se abachler's Avatar
    Join Date
    Apr 2007
    Posts
    3,195
    That thing is puny, I'd rather have this one

  11. #26
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Location
    Inside my computer
    Posts
    24,654
    Quote Originally Posted by DavidP View Post
    Agreed!!! What's the deal with memory being so much slower than processors? We can't fully utilize the true potential of the processor with this big bottleneck on the speed of memory, even with optimizations such as caching, pipelining, etc., etc., etc.
    Because memory cannot keep up! Processors are leaky and that's why they evolve so fast.
    Memory cannot be leaky, which is one of the biggest problems.
    At least, that is what I've heard some time ago.
    We need some new revolutionary type of memory to boost speeds to processor levels in the same time. But if the clock speed stays static, then all we need is a little time. We're already up at 1.2 GHz memory.

    Quote Originally Posted by zacs7 View Post
    That's ridiculous. Socket design has an enormous amount to do with processor design. If you restricted them to the same pin pattern, you'd be restricting the design of processors. You really expect Intel and AMD to work together? Considering the benefits to either of them is close to nil.

    Consider this... A universal 1024-pin socket exists that AMD and Intel both use for their quad-core processors. However, both AMD and Intel are designing oct-core processors (both in secret, they are rivals after all). You really expect them to stay with the 1024-pin socket? Even if the design permitted, it wouldn't be optimal. You can't expect AMD to contact Intel (or vice-versa) and say "we want to develop a oct-core processor, we need a new socket".

    Even if you did that, all it would take is MIPS to strike a deal with <some motherboard manufacturer> and undercut the whole market.

    That being said... go MIPS .
    I'm going to have to agree with zacs7 on this one. To say they use different boards to lock-in is ridiculous. It's a design matter, I think.

    As for the original topic, I like the energy-efficient and cheap parts.
    I don't like overclocking, either, at least not when it comes to raising the voltage.
    Energy is more important to me than speed, and cheap parts more so than expensive ones.
    I even had to underclock the graphics card because it was consuming too much wattage for my poor PSU.
    Last edited by Elysia; 12-18-2009 at 04:27 PM.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  12. #27
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,229
    My very uneducated guess is that it's because DRAM is more analog. CPU is just a bunch of transistors switching. DRAM is actually transistors and capacitors (that actually store the information). Charging up capacitors takes more time.

    Static memory (flip-flops) used for CPU cache is a lot faster, but also a lot more expensive.

  13. #28
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Location
    Inside my computer
    Posts
    24,654
    The biggest problem with SRAM is its size. DRAM takes up very little place compared to SRAM.
    And yes, DRAM is slower because of the capacitors
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  14. #29
    spurious conceit MK27's Avatar
    Join Date
    Jul 2008
    Location
    segmentation fault
    Posts
    8,300
    Quote Originally Posted by cyberfish View Post
    My very uneducated guess is that it's because DRAM is more analog. CPU is just a bunch of transistors switching. DRAM is actually transistors and capacitors (that actually store the information). Charging up capacitors takes more time.
    So heat build up could become an issue in the memory then too if it were faster?
    C programming resources:
    GNU C Function and Macro Index -- glibc reference manual
    The C Book -- nice online learner guide
    Current ISO draft standard
    CCAN -- new CPAN like open source library repository
    3 (different) GNU debugger tutorials: #1 -- #2 -- #3
    cpwiki -- our wiki on sourceforge

  15. #30
    Registered User
    Join Date
    Dec 2006
    Location
    Canada
    Posts
    3,229
    All transistors get hot if we switch them fast, because of their capacitances. Modern CMOS gates only consume significant power (and hence put out heat) when they switch, because Vdd and ground are connected briefly when they switch (both transistor networks conducting).
    CMOS - Wikipedia, the free encyclopedia

    For DRAM, I'm not sure.

    They certainly do get warm, though. But it could just be the transistors.

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Can anyone help?
    By javalurnin in forum C Programming
    Replies: 11
    Last Post: 12-02-2009, 06:02 AM
  2. Replies: 34
    Last Post: 02-26-2006, 01:16 PM
  3. Major Computer Problem
    By Olidivera in forum Tech Board
    Replies: 10
    Last Post: 07-15-2005, 11:15 AM
  4. Tabbed Windows with MDI?
    By willc0de4food in forum Windows Programming
    Replies: 25
    Last Post: 05-19-2005, 10:58 PM
  5. Computer will not boot.
    By RealityFusion in forum Tech Board
    Replies: 25
    Last Post: 09-10-2004, 04:05 PM