Very Extreme Edition. When?

Very Extreme Edition. When?

Post by Henri Tapa » Sat, 07 Jan 2006 03:06:02

When will we see Intel processors that have following features?

- EM64T
- two or more cores
- MMX, MMX+, SSE, SSE2, SSE3, 3DNow! and 3DNow!+ -instruction sets
- Hyper-Threading -technology
- NX-bit
- LaGrande
- Virtualization (Vaderpool/Silvervale?)
- Bumpless Build-Up Layer
- TeraHertz -technology

Henri Heinonen - filosofian ylioppilas fysiikasta

Very Extreme Edition. When?

Post by Eric » Thu, 12 Jan 2006 14:47:26

Why would you want Lagrande? or even virtualization? YUK! a waste of
silicon. EIST & Bumpless Build-Up Layer? not heard of those. What are they?


Very Extreme Edition. When?

Post by Eric » Thu, 12 Jan 2006 14:54:51

EIST is enhanced speed step technology.

AS far as i know, Intel has no plans to increase speed beyond the curent
3.8Ghz. Dumb in my view, very dumb...

Very Extreme Edition. When?

Post by Bill David » Thu, 19 Jan 2006 03:47:51

Virtualization allows running all the best software for multiple
operating systems, running multiple instances of an O/S on virtual
machines, and solving or preventing numerous other problems.

Other than allowing the best applications from multiple operating
systems to run, the average user doesn't need the capability.

-bill davidsen ( XXXX@XXXXX.COM )
"The secret to procrastination is to put things off until the
last possible moment - but no longer" -me

Very Extreme Edition. When?

Post by Gregory D. » Wed, 01 Mar 2006 14:06:57

There may be some very practical reasons for the speed limit. One may be
that the wave length of 3.8 GHz is about 3 inches, as I calculate it. What
the effective wavelength through a semiconduncting material is, I cannot
say presently, accept that it will be considerably shorter. When your
conductor starts to get near 1/4 or even 1/8 a wavelength it will likely
start behaving like a tuned circuit or antenna to some degree. This would
likely lead to all kinds of hard to predict or unmanagable interactions
going on, particularly as the effective length varies as the transistors
turn on and off. Remember also that the desired wave for is a square wave
which has harmonics of even higher frequencies, and so shorter wavelengths.


Gregory D. MELLOTT

Very Extreme Edition. When?

Post by daytrippe » Thu, 02 Mar 2006 00:49:39

On Mon, 27 Feb 2006 22:06:57 -0700, "Gregory D. MELLOTT" < XXXX@XXXXX.COM >

Right. The bizness actually collapsed two years ago under a pile of EM issues.


We managed to get wide serial interconnects running over one or two feet of
copper - and through two or more connectors, no less - at well over 5
gigahertz. And with error rates in the 1 in 10^-18 range or better. The same
technology that's now in commodity platforms, eh?

btw, we don't use no steeekeeeng "square waves" at these speeds. Definitely
*not* "desired"...

/daytripper (SI engineers ain't all dead yet, son ;-)

Very Extreme Edition. When?

Post by pete » Thu, 02 Mar 2006 20:18:06

And make another kneefall for AMD? Get real! ;)


Not anytime soon.

regards, Peter Bjn Perls