CPU ceiling?

[LSMFT]

Are we stuck around 3ghz? I haven't seen much speed improvement in years
other than more cores and 64 bit. Some 4ghz around,seems like we should
be at 7 or 10 ghz by now.

 

[Flasherly]

Are we stuck around 3ghz? I haven't seen much speed improvement in years
other than more cores and 64 bit. Some 4ghz around,seems like we should
be at 7 or 10 ghz by now.

Gigahertz myth [according to wiki] - is a software benchmark to unify
what a CPU cycle can or not do. [What is done at x4 faster speed is
identical to work engineered to be x4 more efficient at x1/4 the
speed].

http://en.wikipedia.org/wiki/File:Transistor_Count_and_Moore's_Law_-_2008.svg

A chart of transistor model density, which supposes -- will twice the
transistors at a given speed (the chart stops at 2008) be utilized
twice as efficiently at half the time?

Six years ago, a micro computer in a research lab was run at 500Ghz.
Speculation is put forth that "atomic-level" miniaturization will be
the final limit in 20 years.

Some theorists further speculate on "technological singularity" -
progress in technology will be instantaneous.

Then, presumably however sweet that would be, is to see only what you
thought you saw in multifarious advancements. A focal imperative,
possibly, for events to follow, per se as identifiably futuristic
technology, by rapidly succeeding design implementations to some
axiomatic end. Entropy, randomness, and evolution, of course, apart a
conscious event horizon of machinery that doesn't go awry -- thinking
preposterous silly thoughts, such as taking us mere mortals over.

 

[TVeblen]

Are we stuck around 3ghz? I haven't seen much speed improvement in years
other than more cores and 64 bit. Some 4ghz around,seems like we should
be at 7 or 10 ghz by now.

Gigahertz myth [according to wiki] - is a software benchmark to unify
what a CPU cycle can or not do. [What is done at x4 faster speed is
identical to work engineered to be x4 more efficient at x1/4 the
speed].

http://en.wikipedia.org/wiki/File:Transistor_Count_and_Moore's_Law_-_2008.svg

A chart of transistor model density, which supposes -- will twice the
transistors at a given speed (the chart stops at 2008) be utilized
twice as efficiently at half the time?

Six years ago, a micro computer in a research lab was run at 500Ghz.
Speculation is put forth that "atomic-level" miniaturization will be
the final limit in 20 years.

Some theorists further speculate on "technological singularity" -
progress in technology will be instantaneous.

Then, presumably however sweet that would be, is to see only what you
thought you saw in multifarious advancements. A focal imperative,
possibly, for events to follow, per se as identifiably futuristic
technology, by rapidly succeeding design implementations to some
axiomatic end. Entropy, randomness, and evolution, of course, apart a
conscious event horizon of machinery that doesn't go awry -- thinking
preposterous silly thoughts, such as taking us mere mortals over.

You got that from a Borg Cube repair manual!

 

[John Doe]

Are we stuck around 3ghz? I haven't seen much speed
improvement in years other than more cores and 64 bit. Some
4ghz around,seems like we should be at 7 or 10 ghz by now.

Gigahertz myth [according to wiki] - is a software benchmark to
unify what a CPU cycle can or not do. [What is done at x4
faster speed is identical to work engineered to be x4 more
efficient at x1/4 the speed].

http://en.wikipedia.org/wiki/File:Transistor_Count_and_Moore's
_Law_-_2008.svg

A chart of transistor model density, which supposes -- will
twice the transistors at a given speed (the chart stops at
2008) be utilized twice as efficiently at half the time?

Six years ago, a micro computer in a research lab was run at
500Ghz. Speculation is put forth that "atomic-level"
miniaturization will be the final limit in 20 years.

Some theorists further speculate on "technological singularity"
- progress in technology will be instantaneous.

Then, presumably however sweet that would be, is to see only
what you thought you saw in multifarious advancements. A focal
imperative, possibly, for events to follow, per se as
identifiably futuristic technology, by rapidly succeeding
design implementations to some axiomatic end. Entropy,
randomness, and evolution, of course, apart a conscious event
horizon of machinery that doesn't go awry -- thinking
preposterous silly thoughts, such as taking us mere mortals
over.

You got that from a Borg Cube repair manual!

He scares the hell out of Borgs...

 

[John Doe]

Are we stuck around 3ghz? I haven't seen much speed improvement
in years other than more cores and 64 bit. Some 4ghz
around, seems like we should be at 7 or 10 ghz by now.

More cores equals tremendous speed improvements.

 

[Marten Kemp]

Yep, beyond that generates too much heat. In the middle 90s, Intel talked
about bio-chips, etc. by Y2K. That never happened. Intel kept upping the
CPU speed until they ran into a heat problem.

Fortunately, they had a skunkworks in Israel that came up with the
multi-core idea.

What Intel does beyond that to improve speed, I don't know but it seems the
aliens from Roswell are no longer around to advise them.

The aliens from Roswell were undocumented and got sent back home <grin>.

One cooling technology that doesn't seem to have gone anywhere
is the Peltier-effect 'refrigerator.' Anybody know why?

 

[Conor]

Are we stuck around 3ghz? I haven't seen much speed improvement in years
other than more cores and 64 bit. Some 4ghz around,seems like we should
be at 7 or 10 ghz by now.

Speed in Mhz has been fairly irrelevent for about a decade now since AMD
came on the scene in the late 1990's.

Modern CPUs carry out more functions per clock cycle than previous ones.

 

[Flasherly]

He scares the hell out of Borgs...

Thanks -- might be a remission after extensible readings, catching up
overall on present trends the written scifi genre -- only having
picked up Ovid's Metamorphosis only last week;-- Really. . .seems
forever, now in noting popular references to The Borg, inasmuch a
sense I eclipsed by attributes inferred to clever quips concerning
Jorges Borges (a formal surrealist), having just within the past few
months gotten around to viewing a corrected association to the
StarTrek production, probably, among the best in that lot. Ah, well
-- such is modernity on the exponential treadmill [<sighs>. . . back
to Ovid and flavors of 8-A.D. Romanism].

 

[SteveH]

He scares the hell out of Borgs...

Thanks -- might be a remission after extensible readings, catching up
overall on present trends the written scifi genre -- only having
picked up Ovid's Metamorphosis only last week;-- Really. . .seems
forever, now in noting popular references to The Borg, inasmuch a
sense I eclipsed by attributes inferred to clever quips concerning
Jorges Borges (a formal surrealist), having just within the past few
months gotten around to viewing a corrected association to the
StarTrek production, probably, among the best in that lot. Ah, well
-- such is modernity on the exponential treadmill [<sighs>. . . back
to Ovid and flavors of 8-A.D. Romanism].

Taking the red pills this week then?

 

[The Seabat]

Not true. Two 3GHz cores don't give you a 6GHz CPU. It gives you more
productivity, but the speed is relative. If you have a race car
running at 200mph around a track and add another car doing 200 mph you
don't end up with one or two cars doing 400 mph! It does allow you do
double your production. Just my $.02 worth.

 

[Sjouke Burry]

Not true. Two 3GHz cores don't give you a 6GHz CPU. It gives you more
productivity, but the speed is relative. If you have a race car
running at 200mph around a track and add another car doing 200 mph you
don't end up with one or two cars doing 400 mph! It does allow you do
double your production. Just my $.02 worth.

And that theoretical speedup of more than 1 core wil only
work when there is no competition on the memory bus and/or
disks and other hardware.
Worst case there can be almost no improvement, when there
is a lot of disk i/o.

 

[John Doe]

Not true.

Says some nobody who is afraid to have his posts archived...

Two 3GHz cores don't give you a 6GHz CPU.

Computer quickness is measured by gigaflops, not single CPU speed.
Multiple CPU cores multiplies the processing power/speed.

It gives you more productivity, but the speed is relative.

That is clueless babbling.
--

 

[John Doe]

that theoretical speedup

It is a proven speed up, Slurry, you ****ing moron.

of more than 1 core wil only work when there is no competition
on the memory bus and/or disks and other hardware.

I use modern hardware. The only bottleneck is between Slurry's two
brain cells.

Worst case there can be almost no improvement

Since Slurry the ****ing moron does nothing with his PC...

When there is no speed improvement, an upgrade to multiple cores
is unnecessary, but that is a different subject.

when there is a lot of disk i/o.

I use an SSD drive, Slurry. I actually have experience with single
and multiple core CPUs, and I know for a fact that there is a huge
speed improvement moving from single to multiple core CPUs.
--

 

[Flasherly]

Taking the red pills this week then?

Good one, also tough - "the" definitive inspiration behind The
Matrix. Hire a team of scripters, I suppose, to write the dialog
according to whatever flavor whomever has been reading of late. Doubt
it's a conversant as Blade Runner status, although Machine Central -
that Platonic scene spin with Neo arguing with prior instances of a
collective-man challenging a mechanistic world ordering - appears to
have gotten respective due in discussion forums.

-
First thing every morning before you arise say out loud, "I believe,"
three times. -Ovid

 

[The Seabat]

Oh, FOAD blow-hard. How did you get out of my bit-bucket?? Have to put
your over rated self-importance back where it belongs, "Plonk!!"

 

[John Doe]

The Seabat <seabat NOSPAMboardermail.com> wrote:

<runs off to play with his imaginary kill file friend>

May as well...

I know that multiple cores CPUs are tremendously faster than single
core CPUs, partly from experience.

My upgrade path...
Athlon XP 3000+ (400 MHz FSB)
Core 2 Duo E6850, 3 GHz
Core 2 Quad Q9550, 2.83 GHz

Playing Forged Alliance. When using only one core, the one core
quickly maxes out and makes the game slow and choppy. Using two
cores makes the game playable except on big maps with many
players, both cores can be maxed out. Using four cores makes it
run smoothly in all circumstances.

If you think that clock speed is the only measure of CPU
quickness, you are horribly ignorant. If you think that four is
bigger than three, you are right, but you do not get a prize for
that.

 

[Marten Kemp]

More cores equals tremendous speed improvements.

Only for OSen and applications which can make
effective use of multiple CPUs. Once Upon A Time
the rule-of-opposable-digit was that about 20%
of the cycles of an additional CPU were consumed
as overhead.

 

[John Doe]

Only for OSen

Windows is the only real PC "OSen", you ignorant ****turd.

and applications which can make effective use of multiple CPUs.

Bullshit, ****turd. The CPU and/or user can force use of the extra
cores.

Once Upon A Time

Another ****turd who lives in a fantasy land.

the rule-of-opposable-digit was that about 20% of the cycles of
an additional CPU were consumed as overhead.

Says a Luddite ****turd.
--

 

[Marten Kemp]

Windows is the only real PC "OSen", you ignorant ****turd.


Bullshit, ****turd. The CPU and/or user can force use of the extra
cores.


Another ****turd who lives in a fantasy land.


Says a Luddite ****turd.

Who pissed in your porridge?

 

[John Doe]

Who pissed in your porridge?

Why do you pretend to know about PCs?
--