Re: Multiprocessor Core
- From: emu8086@xxxxxxxxxxxxxx
- Date: 24 Oct 2005 22:59:48 -0700
Frank Kotler wrote:
> But if I may answer anyway... My thought on it is that this is going to
> ask a lot of programmers - especially compiler-writers. Making efficient
> use of such a setup would take some slick programming, I would think.
> With some exceptions, I don't see that the software we've got does an
> especially good job of managing the hardware we've got. I fear that this
> could make the situation worse. Much of the hardware is likely to be
> sitting idle much of the time, simply due to a lack of anything useful
> to do. When a "peak load" (of unknown size and duration?) comes along,
> parcelling out the labor in an optimal manner seems like a programming
> nightmare to me.
>
> That's just based on a "feeling", not on any factual knowledge of how
> such a beast would be programmed.
>>From my experience OS is the only one to blame for slowing the
applications down (it may not always effectively share the CPU
resources bettween the applications). We need to do "imaginary
programming" right from the boot sequence. In theory it should allow to
load several Operating Systems in parallel on a single multi-core CPU.
.
- References:
- Multiprocessor Core
- From: emu8086
- Re: Multiprocessor Core
- From: Frank Kotler
- Multiprocessor Core
- Prev by Date: Re: Multiprocessor Core
- Next by Date: Re: Multiprocessor Core
- Previous by thread: Re: Multiprocessor Core
- Index(es):
Relevant Pages
|
