Dual Core vs HyperThreading vs Dual CPU

[Danial Thom]

--- "Andrew P." <infofarmer at gmail.com> wrote:

> On 1/10/06, Marc G. Fournier <scrappy at hub.org>
> wrote:
> >
> > I'm going to assume that Dual Core is better
> (can't believe that they took
> > a step back) ... but, is how does it rate?  I
> know that HyperThreading is
> > definitely != Dual CPU ... but how close does
> Dual Core get?
> 
> There is extensive evidence (google for that,
> please), that
> HT is even slower than a single core in quite a
> few applications.
> Moreover, the whole HT implementation has been
> shown to
> be a security risk. In the near future intel is
> going to spend $1.9bn
> on its new marketing campaign. If you wanna be
> part of it,
> buy their CPU, half of the money will be in
> your nearest billboard.
> 
> Dual-core is a new, and a very smart concept,
> which is exactly
> equal to a dual-cpu configuration in terms of
> performance per
> core - plus it provides a huge cut down on
> power consumption,
> and a theoretically hugely faster
> interconnection between the
> cores (they are physically many times closer).

Thats not entirely true, as its not *exactly* the
same. It looks the same to an O/S, but things are
wire differently, so there are likely to be some
differences.

> 
> By 2010 we'll see 4-core, 8-core and maybe even
> 16/32 solutions.

Intel has implied that they will have
multi-multicore processors (more then 2) a lot
sooner than you think. But for now the multicore
thing is just marketing hype because most O/Ses
don't utilize DP efficiently enough to make the
gains worthwhile. You'll likely see more urgency
to produce them when OS'es can actually benefit
from them. Most people today who use DP system
just "assume" they are faster or better without
having any real clue if they are or not.

DT

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com