Re: How much longer before the PC HDD is replaced by flash memory OT

After replacing The Legend with a small shell script on Wednesday 07 Jun
2006 19:32, the following appeared on stdout:

If they did overcome these problems, would there be any need to make
CPU's faster and faster?

It depends on the application, Tox. Some large database servers benefit
from fast processors in SMP configurations. Games machines benefit from
fast CPUs, although there is talk of a PPU (physics processing unit)
co-processor (which I suspect AMD will have the edge over Intel
integrating these into their chipsets since HyperTransport is an open
specification) which will shift an awful lot of gaming's processing
drain off the CPU. Image manipulation and scientific applications need
pure CPU grunt, too. And then there's compiling from source, which I
have to admit I do a lot of. It's very satisfying to watch compiler
messages fly past rather than limp, especially if you're debugging.

Not forgetting, of course, that your CPU and interconnecting buses have
to be able to keep up with the other devices that will be throwing data
and instructions at it. At present, this isn't an issue as the maximum
internal sustained disk to cache data rate is somewhere in the region
of 60-80MBps. SATA 2's burst rate of 300MBps is just that: Burst from
controller to cache, assuming the read-ahead or write-behind cache is
large enough for the entire data transfer. This is why disks with a
larger, faster cache and a faster controller score better in synthetic
benchmarks. The actual disk internals are much slower in a single disk
situation.

Those disks with perpendicular (vertical) domain recording like the
newer Seagates are a little faster internally, but they still won't max
out a SATA 1 connection, let alone a SATA 2. The speed gain is due to a
higher data density on the platter, meaning more data passes the head
at a given RPM than it would for a horizontal recording strategy.
--
Radio glossary #2
Aerial: A piece of modern art which is anathema to women, transparent to
birds but not to wind, the slightest breath of which brings it crashing
down.

.

Relevant Pages

  • Re: What can I check to fix system performance?
    ... it seems you have plenty of memory available: ... copies of files you have read of written lately, in a cache, in case ... processes per CPU, or 40 in all. ... Consider the disk structure. ...
    (comp.os.linux.setup)
  • Re: SP2 Incompatible with my computer (ASUS P4C800, 9600SE, SATA)
    ... Try dissabling the CPU L1\L2 Cache in Bios ... I am not sure about the SATA thing. ... -An Asus Radeon 9200 AGP card ...
    (microsoft.public.windowsxp.setup_deployment)
  • Re: SATA Performance with Intel ICH6
    ... computers cpu time. ... Can I do anything to optimize writing from memory to disk? ... From the SATA bus side there should be no problem. ... sys 0m7.407s ...
    (Linux-Kernel)
  • Re: What core speed were/are north bridges and/or MCHs clocked at?
    ... The CPU writes into some controller register the address range it might ... The data disk writes into memory must get in sync with CPU caches, ... no cache flushing instruction and stuff is done when disk finishes ...
    (comp.sys.ibm.pc.hardware.chips)
  • Re: write cache hit ratio??
    ... does anyone know what exactly is the meaning of "write cache hit ratio" of a filesystem? ... access to disk. ... later when the CPU is busy elsewhere, or when the cache fills up. ...
    (comp.unix.solaris)