Re: Break HD with sw

Tech-Archive recommends: Speed Up your PC by fixing your registry

Hello Fernan,

On Mon, 26 Sep 2005 15:04:05 UTC, "Fernan" <Fernando@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:

> Hello,
>
> I have a c++ application to check/sinchronize files in different memory
> devices. I have:
> - HD office computer
> - HD home computer
> - external 160GB backup HD
> - 512MB memory stick
>
> This application searches up to 20,000 files (10GB max)
> and compares size and date, does it in 60 seconds.
> My old version (VBA Excel) did this in 15 minutes!,
>
> My quiestion: can a HD be so stressed that can break?
> I sinchronized the files in the memory stick (450MB) daily with the
> excel version for a whole year with no problems(check all the files, and
> replace with new files), after 2 weeks of using the c++ version, th memory
> stick breaks.
> Could this be from 1 year of use, or the application broke it?
>
> Thanks

Some devices have a limited number of write or read/write cycles.
Standard memory should never be a problem. Most disk drives will last
five or more years. They are normally hurt by power cycles more than
actual read/write positioning.

Some of the newer memory devices may have certain limitations. You
can read up on what your memory stick is made with and what kind of
memory it uses. That technology, or perhaps just a part of the stick,
has some limit that you have found.

One of the most common memory types that has limited read/write
cycles is CMOS. Anything labeled EEPROM may also have limits.
CMOS Memory is used to hold your PCs hardware configuration and
is relatively small. Most are limited to about 10,000 writes.
Some also have a larger, but relatively finite read limit as well.

There was a time when PC's lost/gained time and some people
wrote programs to update their CMOS Clock Values far too often.
They soon discovered these programs destroyed the CMOS memory
and thus the utility of the PC. Some took less than a couple
days to break and the next boot failed.

Memory sticks and similar devices may also have some time limit
that they can be used. When there is an internal component that
needs charged exceeding that limit could be a problem. This isn't
your problem and you aren't likely to run into it. Such memory
is used for very limited purposes.

As you are finding out, there are times when we need to know
the details of the devices we are working with and take care to
implement a design that does not exceed some design decisions
the manufacturer has made.

David
.

Relevant Pages

  • Re: Desperately Need Help - System Reboot/Crashes
    ... but I suspect that may not be good advice. ... >>I know Windows 9x loads into the upper portion of memory and suspect ... >>Docmem or a similar DOS based memory test program ... CMOS memory bit got mangled by a power spike. ...
    (microsoft.public.windowsxp.general)
  • Re: Tech query
    ... Either way try reseating the memory, if there is more than one memory stick then try them individually in each slot. ... Also try resetting the CMOS memory (usually a jumper to swap and swap back) and make sure that the processor cooler is correctly fitted with sufficient thermal paste or pad. ...
    (uk.people.silversurfers)
  • Re: AMD64 assembly code optimization
    ... that's why I wouldn't have any prefetch in short loops. ... replacing the wide-spread memory references to access a small array ... My processor has 64K L1 data cache and that in takes 3 cycles to fetch ...
    (comp.lang.asm.x86)
  • Re: [patch 10/10] *Tests* Scheduler profiling - Use immediate values
    ... The 48 cycles sounds to me like a memory load in an unloaded system, ... If we put 2 markers for irq entry/exit, ... cycles_t time1, time2; ...
    (Linux-Kernel)
  • Re: 6502 trashing memory cycles...
    ... this point the opcode has not been decoded and the cpu _has no way_ of ... If you 'stole' that memory cycle ... I'm saying however that whatever it is, during these cycles, accessing ... corresponded to an RTS opcode? ...
    (comp.sys.apple2)