Dual Core Pentiums?

[Mxsmanic]

But aren't more and more home users using their PCs for
media servers?
No.

Installing TV tuner cards in them and
storing live Tv feeds on the hard drive for streaming
to their home networks?

No.

 

[me]

I.E. "what do you want

to do?" Well, dad gummit, I'd like the thing to boot up in the same state
it was in when it left. In my case, the TV. My TV, the thing it's
supposedly replacing, does.

Excellent point!!

 

[Matt]

No need to wait. There's a ton of applications written for those 'Dual
Intel CPU boards" that "have been out for decade."



Non sequitur. The "decades worth of hardware" you talk about was not for
the "home user."

For all the (non) sense your argument makes you might as well pine that
there aren't more 'home user' space suits on the market since 'space
craft have been out for decades'.

I expect the difference is that dual core is going to be much cheaper
than dual CPU. That will lead to an economy of scale for the software.

 

[David Maynard]

I expect the difference is that dual core is going to be much cheaper
than dual CPU.

I would imagine that's the theory, assuming die yields are good enough.

That will lead to an economy of scale for the software.

Yeah. People tend to be a bit more reluctant writing software for things
that don't exist vs things that do.

 

[rhys]

Dual-core CPUs or dual CPUs give a significant performance advantage to
those who multi-task and to those who use SMP-aware apps like Photoshop.
While this may not describe a majority of the home market, there are a
significant number of "power users" who can benefit.

Exactly. I built a dual proc Pentium II in 1997, a faster model of
same in '99 for Windows NT, and a dual proc PIII in 2001 for Windows
2000 running several dual-proc aware apps, mostly in the graphic
design realm.

My question is: shall my next PC (which is about due) be dual proc or
dual core? I've seen the benefits for years now, and in my situation a
single processor gets stalled, given my work habits.

Thoughts appreciated.

R.

 

[Mxsmanic]

My question is: shall my next PC (which is about due) be dual proc or
dual core?

What's wrong with two or more dual-core processors?

 

[rhys]

What's wrong with two or more dual-core processors?

Nothing, I suppose, but it's overkill for what I do: multitask under
Win2000 with slightly slower processors, a whack of RAM for Photoshop
and InDesign and a very good video card. I can get less bleeding edge
processors and I don't require super-fast drives.

I'm not doing 3-D renders. Two dual-cores under a 64-bit OS would
indeed rock.

R.

 

[Fuckknuckle]

Jesus. How old are you guys? Old PCs couldn't multitask at all. They ran
friggin DOS ferchrissakes. Windows couldn't multitask until version 3.0, and
even then it was not multitasking as people know it today. The BSOD was a
daily occurance for any brave souls who attempted it.

Seems to me that part of the appeal of UNIX's arrival was the
multitasking capability.
Multitasking has been around for long, long time. Maybe we should be
using the
term 'multiprocessing'.

 

[Fuckknuckle]

And who said "PCs?" The statement was "Multitasking has been around for
half a century" and, contrary to your apparent assumption, 'PC' and
'computer' are not synonyms, much less 'IBM PC running Microsoft software'.



Not true.



Poorly written code BSODs whether you're multitasking or not.

Amen.

 

[Fuckknuckle]

Let's get one thing straight: The number of processors, or cores,
involved has absolutely nothing to do with "multitasking." Speed,
perhaps, but not "multitasking."



Is a fact.



So no single processor system can do 'true multitasking' (whatever the
heck you think 'true multitasking' means) and your 'dual core' processor
will be only able to 'truly' handle 2?

That's just nonsense.



And yet you found a way.

which is pretty much unavoidable in the absence of logic

 

[Fuckknuckle]

THere is no difference doing it on a single core P4/XP CPU than a dual
core.

Other than the speed, which is important to many people.

Ah..another ****tard who buys into the benchmarks. Tell me, does it
mean you can type faster when browsing web whilst ripping a DVD at the
same time or does it in fact make **** all difference?

Ok, you must have Downe's Syndrome.

 

[Fuckknuckle]

If you can't do all of those at the same time on a single core then
your PC is ****ed. I do all of that on a XP2500+ with 1GB RAM.

In relation to speed, I meant.

 

[Mxsmanic]

Seems to me that part of the appeal of UNIX's arrival was the
multitasking capability.

The main appeal of UNIX was that it was (nearly) free, and source was
readily available. From a technical standpoint it never had that much
to recommend it; in many ways, it was a solution looking for a problem
(although it eventually found its niche).

Multitasking has been around for long, long time. Maybe we should be
using the term 'multiprocessing'.

Yes. For what it's worth, multiprocessing has been around for a long,
long time, too.

 

[BP]

Seems to me that part of the appeal of UNIX's arrival was the multitasking
capability.
Multitasking has been around for long, long time. Maybe we should be using
the
term 'multiprocessing'.

Yeah, that's what I was talking about. And within the context of the IBM
compatible PC. Not mainframes, workstations, or even Macs (although a Mac
would fit into this group, wouldn't it?).
The poor-prick OP just wanted to know whether to build a new machine with
the new processor.

 

[David Maynard]

****knuckle writes:




The main appeal of UNIX was that it was (nearly) free, and source was
readily available.

That is an terrible oversimplification. Unix was, at least in theory, 'open
source' (read about the AT&T/BSD UNIX license wars) but, at the time, that
was of little use to anyone but manufacturers who then sold their
particular Unix 'flavor' with their hardware. e.g. Ultrix on VAX, AT&T Unix
on HP, etc.

BSD itself wasn't originally free software either. It was sold, with the
University of California's blessing, at a nominal cost only to people or
institutions that had already purchased licenses permitting them access to
the source code of AT&T Unix.

Licensing problems was one of the spurs to developing things like Minix and
Linux.

From a technical standpoint it never had that much
to recommend it; in many ways, it was a solution looking for a problem
(although it eventually found its niche).

That, no doubt, is why DARPA chose it to be the preferred "universal
computing environment" linking together Arpanet research nodes, which
evolved into the Internet.

Things that people take for granted today as 'obvious' were anything but
back in the early days and one of the radical new concepts in UNIX was
hardware transportability with a consistent software environment. An
incredibly powerful concept, that meant one could transport applications
across platforms and developers would be presented with the same basic
environment regardless of whose computer you were using. Prior to UNIX you
could identify the computer by the O.S. and that's 'what kind of
programmer' you were. If the O.S. was VMS then the computer was a VAX and
you were likely a 'VAX/VMS Programmer," and that's where your code would
end up, on a DEC VAX, but UNIX told you nothing of the hardware because it
could be running on any machine.

One can immediately see the benefits to ARPANET. Applications could be
developed ONCE and commonly applied to all machines regardless of the
particular hardware employed at any one site.

 

[Mxsmanic]

That, no doubt, is why DARPA chose it to be the preferred "universal
computing environment" linking together Arpanet research nodes, which
evolved into the Internet.

I'm sure they chose it for reasons other than technical superiority; if
they wanted the latter, they could have used Multics. Licensing and
availability issues were very important, then as now. The fact that
UNIX was a stripped Multics that could run on readily available hardware
was probably a factor, also, though. And remember that there wasn't
really much (any?) competition.

One can immediately see the benefits to ARPANET. Applications could be
developed ONCE and commonly applied to all machines regardless of the
particular hardware employed at any one site.

Multics was written in a high-level language long before UNIX.

 

[David Maynard]

David Maynard writes:




I'm sure they chose it for reasons other than technical superiority;

And what makes you so "sure?"

if
they wanted the latter, they could have used Multics.

Multics was long on promise but routinely short on delivery and was
perpetually behind schedule, by years seemingly approaching decades. By
1985 Honeywell (acquired it in 1970, just one year after the first system
emerged) had tried to cancel the thing 5 times, it doesn't instill gobs of
confidence when the maker is trying to assassinate it, till they succeeded
in 85 on the 6'th try. Numerous ports were contemplated but none of any
major significance made it.

There were some revolutionary ideas in Multics but it takes more than just
the ideas.

Licensing and
availability issues were very important, then as now. The fact that
UNIX was a stripped Multics that could run on readily available hardware
was probably a factor, also, though. And remember that there wasn't
really much (any?) competition.




Multics was written in a high-level language long before UNIX.

There's more to it that just programming in a high level language.

 

[Conor]

Other than the speed, which is important to many people.

Which you only get it the application is SMP. Oh and don;t forget 2 x
1GHz processors are nowhere near as fast as 1x 2Ghz.

Ok, you must have Downe's Syndrome.

Another marketing believer.

 

[Conor]

In relation to speed, I meant.

Does it make the data from the net transfer any faster? Does it make
the CD/DVD writer any quicker? Can you read any faster? Can you type
any faster?

Next...

 

[Mxsmanic]

And what makes you so "sure?"

Because UNIX isn't technically superior.

Multics was long on promise but routinely short on delivery and was
perpetually behind schedule, by years seemingly approaching decades.

Its only real problem was that it was ahead of the hardware available at
the time.