REVERSE Hyperthreading for AM2! - Page 11 - Techist - Tech Forum

Go Back   Techist - Tech Forum > Computer Hardware > New Systems | Building and Buying
Click Here to Login
Closed Thread
 
Thread Tools Display Modes
 
Old 06-23-2006, 11:33 PM   #101 (permalink)
Lord Techie
 
Nitestick's Avatar
 
Join Date: May 2005
Location: смерти для спаме
Posts: 8,473
Default

i'm pretty sure that it wasn't really a chip IBM was testing rather than just a bunch of transistors. i'm not sure it was actually processing anything. a chip would be considerably slower than what IBM has shown these transistors are capable of but still i would guess impressive. also keep in mind this experiment was said to have been conducted at 3 degrees above absolute zero and the home overclocker won't be able to pull that off
__________________

Nitestick is offline  
Old 06-23-2006, 11:41 PM   #102 (permalink)
Techie Beyond Description
 
Apokalipse's Avatar
 
Join Date: Jun 2003
Location: Melbourne, Australia
Posts: 14,559
Default

true, though it did get 350GHZ at room temperature.

anyway, let's get back to reverse hyperthreading
__________________

__________________
Apokalipse is offline  
Old 06-23-2006, 11:47 PM   #103 (permalink)
Banned
 
Join Date: Jun 2006
Posts: 232
Default

Not much to get back to, IMO. Some people are saying that its good enough even if its not official (I'm in those), others are saying that they want it to be official or its a no-go. Either way, we still get to wait for Conroe to see if the option in the BIOS works. And the day after Conroe is released, AMD is supposed to make a statement on it. So its still Conroe, lol.

Either way, since both companies have the technology, it doesn't really give anyone the upper hand. Its great for us though, since our performance increases dramatically in most of our applications.

I guess some of us could talk about how this was a slap in the fact, Gaara and myself included. He might still stick to having it be official, but I was speechless to see this avaliable in AM2, lol. AMD kept it under some pretty d*mn good wraps to make us think that it wouldn't be here for a long long time.

<3 surprises.
Infomatic is offline  
Old 06-23-2006, 11:57 PM   #104 (permalink)
Lord Techie
 
Nitestick's Avatar
 
Join Date: May 2005
Location: смерти для спаме
Posts: 8,473
Default

i still think it is very doubtful. from virtually any standpoint it doesn't seem to be the best course of action to hide the technology on the chips nor does it seem as if AMD could have figured it out this quickly. the closest to official reports from AMD suggested something around 2009 in the way of RHT and AMD's R&D department aren't normally optimistic or pessisimistic withe their estimates. generally when they say something will be ready it will be ready. well if AMD have got it down "more power to them". i don't think it's impossible the rumour is true i'm just not going to hold my breath on it.....i might die
Nitestick is offline  
Old 06-25-2006, 02:16 AM   #105 (permalink)
Techie Beyond Description
 
Apokalipse's Avatar
 
Join Date: Jun 2003
Location: Melbourne, Australia
Posts: 14,559
Default

AMD works together with IBM on a lot of things. SiGe transistors is one of them
__________________
Apokalipse is offline  
Old 06-25-2006, 02:43 AM   #106 (permalink)
Banned
 
Join Date: Jun 2006
Posts: 232
Default

Not that again...*groan*
Infomatic is offline  
Old 06-25-2006, 05:49 PM   #107 (permalink)
Master Techie
 
Join Date: Feb 2005
Posts: 2,078
Default

So do you guys think if RHT and Multiplexing works out maybe Amd and Intel will develop more advanced versions and start makeing cpus with a lot of cores running at slower speeds? Like say 10 cores at 500mhz but if you needed 2ghz for a program it would combine 4 and have the others run other smaller programs.

Thats what I always thought they should do. I figured that would work out well because I assume they could make 10-500mhz cores run cooler than 2-2.5ghz cores, especially since then you could have the processors that are generating the most heat be spaced out more.
dario03 is offline  
Old 06-25-2006, 06:55 PM   #108 (permalink)
Banned
 
Join Date: Jun 2006
Posts: 232
Default

The improvement in performance won't be twice what it is of one of them. So 10 x .5Ghz cores won't make a 5Ghz Single Core. 2 x .5Ghz might get you a .8Ghz. Also, for now, you can only make 2 cores act like 1, or 4 cores act like 2, or 8 cores act like 4. You can't make 8 cores act like 1, atleast not yet.

Also, it would be impractical. What WILL happen is that you can get Quad Cores that will become Supercharged Dual Cores. And so on.
Infomatic is offline  
Old 06-25-2006, 08:50 PM   #109 (permalink)
Master Techie
 
Join Date: Feb 2006
Posts: 2,992
Default

Quote:
Originally posted by Infomatic
Also, it would be impractical. What WILL happen is that you can get Quad Cores that will become Supercharged Dual Cores. And so on.

would that introduce Reverse reverse HT?


Take a quad core, make it into a Supercharged Dual core, but then into a Mega Supercharged Single core??


wow.....

I wish they would do this for GPU's or somthing...
__________________



May the wind always be at your back and the sun upon your face, and may the winds of destiny carry you aloft to dance with the stars
Lord AnthraX is offline  
Old 06-25-2006, 09:04 PM   #110 (permalink)
Banned
 
Join Date: Jun 2006
Posts: 232
Default

No. Reverse HT is what would make a Quad Core into a supercharged Dual Core. Anything more than that is not possible. There is no way, currently, to make a Quad Core into a really fast Single Core.
__________________

Infomatic is offline  
Closed Thread

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off




Copyright 2002- Social Knowledge, LLC All Rights Reserved.

All times are GMT -5. The time now is 07:32 PM.


Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2017, vBulletin Solutions, Inc.