HyperThreading

somerandomguy

Distinguished
Jun 1, 2001
577
0
18,980
I was just reading about HyperThreading <A HREF="http://www.extremetech.com/article/0,3396,s=1031&a=16756&app=1&ap=2,00.asp" target="_new">here</A> at Extreme Tech, and it says that HyperThreading is already built into the Pentium 4 chip, but is disabled.
Is this true, and if so, what's the delay?
When will we be able to buy chips with HyperThreading enabled?

"Ignorance is bliss, but I tend to get screwed over."
 

charliec2uk

Distinguished
Jul 26, 2001
249
0
18,680
I think that hyperthreading (which does look P.D.G.) is to be available at first in Xeon chips at some point next year. Although there don't appear to be plans for this technology to filter in to desktop processors for wome time yet.

But I reckon that Raystonn should know better than me.

Charlie

Democracy Bernad, it must be stopped!
 
G

Guest

Guest
Hyperthreading looks good indeed. Nice way of keeping your cpu running when memory latency and bandwith constraints would otherwise stall it. Im just curious how OS's are going to handle this, since it should be entirely transparant to the OS. That mean Windows 2000 Professional for example, would see a single HT (Jackson) Xeon as a dual CPU. Would a dual HT-Xeon then require windows 2000 advanced server ?

---- Quote of the day: "I hope admin can delete massages.."
 
G

Guest

Guest
i believe win2k will see a single HT xeon as a single CPU.

no proof but think i read it on the forum at aces.

i had a drink the other day... opinions were like kittens i was givin' away
 

somerandomguy

Distinguished
Jun 1, 2001
577
0
18,980
That mean Windows 2000 Professional for example, would see a single HT (Jackson) Xeon as a dual CPU.

According to the article I linked, some database software (at least) will indeed show the CPU with HyperThreading as being a dual CPU.

"Ignorance is bliss, but I tend to get screwed over."
 
G

Guest

Guest
Where is the link ?

---- "Creative soundcards are excellent, <snip> It's worth the low price." (c) intel_inside
 

somerandomguy

Distinguished
Jun 1, 2001
577
0
18,980
In my first post. <A HREF="http://www.extremetech.com/article/0,3396,s=1031&a=16756&app=1&ap=2,00.asp" target="_new">Here it is again</A>, in case you still can't find it. :smile:

"Ignorance is bliss, but I tend to get screwed over."
 
G

Guest

Guest
Good link, thanks a lot.

Whoooohooo.. I'm an "addict".. Free drinks for everyone !
All I need now is a good signature ;-) Common Juin, gimme some inspiration !

= The views stated herein are my personal views, and not necessarily the views of my wife. =
 
G

Guest

Guest
The answer to my question is right in there:

"Obviously, OS support was needed, and much validation would be required with the OS. Given that the chip became available during Windows XP development timeframe, it made sense for XP to be the OS with initial support."

I had no idea the OS had to support it. Apparently its still up to the OS to detect a thread (well, virtual CPU) is iddle, and issue a HTL instruction to that virtual processor so that other concurrent threads get all the processing power. I thought the cpu would manage that by itself.. but then again, what do I know ?

= The views stated herein are my personal views, and not necessarily the views of my wife. =