Sign in with
Sign up | Sign in
Your question

4GHz AMD64's Not Far Off

Tags:
Last response: in CPUs
Share
April 10, 2006 11:05:52 PM

Not sure if it's been posted, but it's worthy of a runaround if it has:

http://sharikou.blogspot.com/2006/04/amd-to-increase-pe...

Hmm...so much for all the Fanboys saying 3GHz is a wall...hmm....

~~Mad Mod Mike, pimpin' the world 1 rig at a time

More about : 4ghz amd64

April 10, 2006 11:45:12 PM

*drooling on floor* - that would be nice
April 11, 2006 1:06:02 AM

interesting situation perhaps... Conroe at 3Ghz beating an AMD at 4Ghz... talk about trading places :lol: 
Related resources
April 11, 2006 1:06:55 AM

Quote:
interesting situation perhaps... Conroe at 3Ghz beating an AMD at 4Ghz... talk about trading places :lol: 


Reverse that and subtract 1GHz fanboy.

~~Mad Mod Mike, pimpin' the world 1 rig at a time
April 11, 2006 1:13:54 AM

Quote:
Not sure if it's been posted, but it's worthy of a runaround if it has:

http://sharikou.blogspot.com/2006/04/amd-to-increase-pe...

Hmm...so much for all the Fanboys saying 3GHz is a wall...hmm....

~~Mad Mod Mike, pimpin' the world 1 rig at a time


Well, technically, 3Ghz is a wall. They haven't made 4Ghz yet and it's only speculation.
April 11, 2006 1:14:36 AM

Quote:
Not sure if it's been posted, but it's worthy of a runaround if it has:

http://sharikou.blogspot.com/2006/04/amd-to-increase-pe...

Hmm...so much for all the Fanboys saying 3GHz is a wall...hmm....

~~Mad Mod Mike, pimpin' the world 1 rig at a time


You miss understand what that means, clock skew is what is holding back the A64 not transistor performance.

With that in mind any processors made off that process persay will not see a 40% performance increase in transistors, due to gate structures and power hot spots.

Much of what makes say a A64 do what it does is in the design not soo much in the process.
April 11, 2006 2:02:19 AM

Wow... that would be insane... I can't wait to see if that happens... Intel's Conroe chip would be in for some trouble.
April 11, 2006 2:51:32 AM

Some fanboi blogspot site posts that AMD is announcing a process technology that might lead to 40% gains in transitor performance....

and somehow that leads to ....

AMD will ship 4GHz chips.

As cool as 4GHz AMD chips would be... that is some pathetic deductive reasoning.
April 11, 2006 2:53:22 AM

Quote:
Some fanboi blogspot site posts that AMD is announcing a process technology that might lead to 40% gains in transitor performance....

and somehow that leads to ....

AMD will ship 4GHz chips.

As cool as 4GHz AMD chips would be... that is some pathetic deductive reasoning.


http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,...

http://www-03.ibm.com/chips/news/2004/1213_amd.html

So now AMD.COM and IBM.COM are Fanboy Sites? Son, you're a little slow aren't you? A little simple deductive reasoning brings about the 4GHz. It isn't a hard concept to grasp my friend.

~~Mad Mod Mike, pimpin' the world 1 rig at a time
April 11, 2006 2:53:22 AM

Quote:
Some fanboi blogspot site posts that AMD is announcing a process technology that might lead to 40% gains in transitor performance....

and somehow that leads to ....

AMD will ship 4GHz chips.

As cool as 4GHz AMD chips would be... that is some pathetic deductive reasoning.
Agreed, oh and welcome to the Forumz!
April 11, 2006 2:56:49 AM

Quote:
Some fanboi blogspot site posts that AMD is announcing a process technology that might lead to 40% gains in transitor performance....

and somehow that leads to ....

AMD will ship 4GHz chips.

As cool as 4GHz AMD chips would be... that is some pathetic deductive reasoning.


http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,...

http://www-03.ibm.com/chips/news/2004/1213_amd.html

So now AMD.COM and IBM.COM are Fanboy Sites? Son, you're a little slow aren't you? A little simple deductive reasoning brings about the 4GHz. It isn't a hard concept to grasp my friend.

~~Mad Mod Mike, pimpin' the world 1 rig at a time I don't like it when people eat up every piece of shi7 a company feeds them. Not saying that it isn't true, but you must be cautious in a company's marketing. I'd be more apt to believe IBM's article just simply because they are a different company and AMD doesn't have near enough cap to satisfy a giant like IBM.
April 11, 2006 2:58:23 AM

Quote:
I don't like it when people eat up every piece of shi7 a company feeds them.


Likewise my Blue Friend.

~~Mad Mod Mike, pimpin' the world 1 rig at a time
April 11, 2006 3:10:04 AM

Quote:
Some fanboi blogspot site posts that AMD is announcing a process technology that might lead to 40% gains in transitor performance....

and somehow that leads to ....

AMD will ship 4GHz chips.

As cool as 4GHz AMD chips would be... that is some pathetic deductive reasoning.


http://www.amd.com/us-en/Corporate/VirtualPressRoom/0,,...

http://www-03.ibm.com/chips/news/2004/1213_amd.html

So now AMD.COM and IBM.COM are Fanboy Sites? Son, you're a little slow aren't you? A little simple deductive reasoning brings about the 4GHz. It isn't a hard concept to grasp my friend.

~~Mad Mod Mike, pimpin' the world 1 rig at a time

So you are claiming that a 40% increase in transistor performance (claimed in a press release) implies a linear scaling of shipping processor clock speeds. That is not very smart.

Since the 70s it has been known that there are semiconducting materials, such a Indium Arsenide that have 50 times the electron mobility of silicon... OMGZ AMD is gonna ship 150GHz processors next year!!!

Lame.
April 11, 2006 3:26:26 AM

Quote:
AMD is gonna ship 150GHz processors next year!!!
You heard it first on tomshardware.com
April 11, 2006 3:32:48 AM

Quote:
Not sure if it's been posted, but it's worthy of a runaround if it has:

http://sharikou.blogspot.com/2006/04/amd-to-increase-pe...

Hmm...so much for all the Fanboys saying 3GHz is a wall...hmm....

~~Mad Mod Mike, pimpin' the world 1 rig at a time



I was gonna write it off as BS, but since you posted it, I'll beleive it. A P4 at 4GHz owns, so I'd love to see what an A64 could do at that speed. It could only be beaten by like an 8GHz P4.
April 11, 2006 4:01:00 AM

i woul like to see where this guys Sources are :L would be cool but meh i dout it at least in 06

oih yea and the http://www.theinquirer.net/ is def not a creditable source
April 11, 2006 4:02:52 AM

Quote:
i woul like to see where this guys Sources are :L would be cool but meh i dout it at least in 06

oih yea and the http://www.theinquirer.net/ is def not a creditable source


Looking at your Location, Your new to the forums, and this post...it would appear there is another person who hates me....when I move to my next location this week, you gonna change yours too? I'm going pretty far man....

~~Mad Mod Mike, pimpin' the world 1 rig at a time
April 11, 2006 4:10:13 AM

hmm no i always put location Moscow Russia in all the forums i join :p 

annd as for the hating as long as you bring intelligence to the fourm which from what i read u you do so meh i dont hate you :D 

i do hate fanboys, mainly the fanboys that make up crap...and people who think you need 2-4gigs of ram to run Windows Vista...and Mac users.........and ipods >_<

-edit i just thought it was time for me to stop reading and join in the arguments :D 
April 11, 2006 4:43:09 AM

The thing that always gets to me when companies advertise percents is what they are comparing to.

Quote:
The companies announced that they have successfully combined embedded Silicon Germanium (e-SiGe) with Dual Stress Liner (DSL) and Stress Memorization technology (SMT) on Silicon-On-Insulator (SOI) wafers, resulting in a 40 percent increase in transistor performance compared to similar chips produced without stress technology, while controlling power consumption and heat dissipation.

The fact that this is relative to chips that don't have stress technology means that the relative increase is less over the current generation than it may appear.

Quote:
In December 2004, AMD and IBM announced 24% transistor performance increase with DSLSSOI, and Opteron speed went from 2.4GHZ to 2.8GHZ (3GHZ soon).

The blog itself points out that the stress technology used in current chips yields a 24% increase in transistor performance, which is very impressive, and is responsible for the clock speeds going up from 2.4 GHz to 3 GHz. That is incidentally a 25% increase in clock speeds. However, the 40% increase that they're now advertising is not on-top of this 24% from their existing stress technology but before it. That would seem to indicate that the relative increase over the current stressed generation is closer to a 16% increase in transistor performance over their current technology. Whether that means they'll be able to hit 4 GHz I don't know, but their numbers are certainly composed for best effect. Intel is no better of course.
April 11, 2006 4:46:02 AM

Quote:
The thing that always gets to me when companies advertise percents is what they are comparing to.

The companies announced that they have successfully combined embedded Silicon Germanium (e-SiGe) with Dual Stress Liner (DSL) and Stress Memorization technology (SMT) on Silicon-On-Insulator (SOI) wafers, resulting in a 40 percent increase in transistor performance compared to similar chips produced without stress technology, while controlling power consumption and heat dissipation.

The fact that this is relative to chips that don't have stress technology means that the relative increase is less over the current generation than it may appear.

Quote:
In December 2004, AMD and IBM announced 24% transistor performance increase with DSLSSOI, and Opteron speed went from 2.4GHZ to 2.8GHZ (3GHZ soon).

The blog itself points out that the stress technology used in current chips yields a 24% increase in transistor performance, which is very impressive, and is responsible for the clock speeds going up from 2.4 GHz to 3 GHz. That is incidentally a 25% increase in clock speeds. However, the 40% increase that they're now advertising is not on-top of this 24% from their existing stress technology but before it. That would seem to indicate that the relative increase over the current stressed generation is closer to a 16% increase in transistor performance over their current technology. Whether that means they'll be able to hit 4 GHz I don't know, but their numbers are certainly composed for best effect. Intel is no better of course.

Relative this, relative that...jeez man, you sound like a buddy of mine that must find relevance in everything and anything, can't just let sh!t be sh!t for the sh!t the sh!t is....no offence, but let the sh!t be sh!t!.....sh!t! 8O

~~Mad Mod Mike, pimpin' the world 1 rig at a time
April 11, 2006 9:25:25 PM

AMD just joined the 3GHz club so why is 4GHz out of the question? based on their "older tech"(as some would say) they reached it then the 4GHz isn't out, and could come late 06'

AMD 3GHz Opteron
April 12, 2006 2:07:57 AM

Quote:
AMD just joined the 3GHz club so why is 4GHz out of the question? based on their "older tech"(as some would say) they reached it then the 4GHz isn't out, and could come late 06'

AMD 3GHz Opteron


Physics...and Electromigration which in an nut shell is that the electrons in the chip move so fast that they destory its own circuitry.... happens to Northwoods P4 if u up the Vcore 1.7+ but they called it S.N.D.S or Sudden Northwood Death Syndrome, the reason behind it all is the phsical pneumonia Electromigration

because of this 4.0ghz wont be happing for a while untill new technology can make the circuitry better like a new material other then silicone. yes u can o/c to 4.0ghz eaisly however after a while they will start to die of course there are the luckey few that have run 4.0ghz stable for some time but all in dew time it will slowly become unstable and destroy itself
April 12, 2006 3:44:55 AM

finally, some good constructive posts devoid of fanboyism. :)  *breathes sigh of relief*
April 12, 2006 4:01:10 AM

Quote:
finally, some good constructive posts devoid of fanboyism. :)  *breathes sigh of relief*


Word.
April 12, 2006 4:04:23 AM

Quote:
finally, some good constructive posts devoid of fanboyism. :)  *breathes sigh of relief*


Word. .droW
Anonymous
a b à CPUs
April 12, 2006 4:10:35 AM

Yeah good thread MAdMod,learned more stuff...

Thanks too JumpingJack, FatFunkeu and ltcommander!
April 12, 2006 4:11:51 AM

Quote:
finally, some good constructive posts devoid of fanboyism. :)  *breathes sigh of relief*


Yah...doesn't happen too often does it :wink: I try to avoid waking the fanboy within me at all times, he is real nasty. :twisted:
April 12, 2006 4:13:41 AM

Quote:
Yeah good thread MAdMod,learned more stuff...

Thanks too JumpingJack, FatFunkeu and ltcommander!


Trust me, MadMad thinks we are a bunch of idots..... He is the only non-idiot in the whole of Forumz. Look at me, I'm just spewing with fanboyism and idiocy.
Anonymous
a b à CPUs
April 12, 2006 4:39:57 AM

Have'nt been posting a lot lately, ive been quite busy with school, but I did some reading on the forum...

I saw a lot of post of MMM and my opinion on him is already set. I was being sarcastic because most of the thread end up with flame war and comments based on speculation, so it was insightfull to get nice post from the people I mention.

Oh and that guy MMM made seams nice but did'nt have time to read it. 2 more exames and I'm free!
April 12, 2006 6:58:36 AM

It will be interesting to see exactly what speeds AMD will be able to reach within the next year, I do predict that it will be over 4GHz. After all, IBM has supposedly hit over 5GHz in the lab with their new power 6 chips, and AMD and IBM do share alot of CPU R&D.

I guess it's a matter of waiting to see exactly what AMD has gotten from their partnership.
a b à CPUs
April 12, 2006 7:37:22 AM

wow 4ghz, Intel P4's are at 3.8ghz for like over a year now wow what performance woa.

HAHAHAHAHAHAHAHAHAHHAHAHAHAHAHHA

Clock for clock is another thing, and Intel will have 45nm and CSI coming, not as if Intel need em to beat K8's anyhow, but whenever AMD decides to come back with somethin good Intel will have em lined up ready so eh.

You wanted to boast AMD news so ill post Intel news - thats fair aint it?

For god sakes i was a fast system not a god dam brand name moron.
April 12, 2006 8:56:26 AM

This thread about 4GHz AMD on 90/65nm SOI is just another fantasty story and wish of AMD fanboy, written on his blog. I can write about Iraq preparing to destroy the world with whatever weapon, so BE SCARED!!!!!!!!!!!!!!
Like other disscutants said the phsysics of the used materials and sizes limits K8 to achieve stable 4GHz whether 90nm or 65nm. And another thing that AMD fanboy don't want to think about is thermal dissipation which grows exponentionaly as freqfency rises. Not to mention the votlage that is needed for 4GHz operation swtiching of the transistors.
April 12, 2006 9:17:48 AM

Typical yeild is more like (apx) ~ 3.5 GHz using 65nm, the 4.0 GHz parts will be far and few between, assuming any at all get released this year even :p , ... on that die size and manufacturing process. Just like 2.8 GHz and 3.0 GHz parts on 90nm 'cost significantly more' due to yeild.

This is cold hard fact, it can be argued with but only at the folly of the poster.

Still I'd like to see the both of them compete for now.

2.5 GHz is all one really needs, after that more cores help, rather than more clock speed.
April 12, 2006 11:56:52 AM

Here where the Fanboy morons forget. It dont matter who win or loses. As long as both groups are doing good. If Amd wins. I will buy amd. If Intel wins I will buy intel. But Untel then. It just funny watching a group of babys going My brother better then yours. So People stop acting Like 6 years old and grow up.
April 12, 2006 1:40:40 PM

Quote:
and people who think you need 2-4gigs of ram to run Windows Vista...


I've run it on 256MB of RAM but that doesn't mean it should be run on that much RAM.

2GB is about what you need to have a responsive system. And even still it has to pull shit off the hard drive a lot. Of course then theres the annoying "Do you want to allow <insert every system thing you can think of here> " message you get.

You can't even pull up the damn task manager for christs sake without clicking a button to allow it. And then you have to select which processes you want to see. Granted you could probably change a setting to fix that but its retarded they implemented it that way to begin with. Especially on an admin account.
April 12, 2006 1:52:39 PM

Actually, we at futuretechnology.eu have been running a hush-hush AMD 128 bit 16 core 4 ghz chip through some benches.

The cpu is 90% liquid.

How did they get all them sparkly bubbles into it?

Wow!!

benchmark rig specs:

4 terabytes of QuadDDR PC 76000 ram
1 petabyte solid state drive w/ 950 gb/sec through put
1 google socket (we own the world) mb
1 amd 128 x16 core 4 ghz
1 Voodoo 5 "We made a come back " series vid card w/ top secret specs.
Google form factor case / gas grill and frige combo
Google holo keyboard

benchmark test results:

Figured out women in .14 hours.
April 12, 2006 2:10:49 PM

Quote:
wow 4ghz, Intel P4's are at 3.8ghz for like over a year now wow what performance woa.

HAHAHAHAHAHAHAHAHAHHAHAHAHAHAHHA

Clock for clock is another thing, and Intel will have 45nm and CSI coming, not as if Intel need em to beat K8's anyhow, but whenever AMD decides to come back with somethin good Intel will have em lined up ready so eh.

You wanted to boast AMD news so ill post Intel news - thats fair aint it?

For god sakes i was a fast system not a god dam brand name moron.


Words of a fanboy. There is no need to lower the tone. we have MMM for that.

I sincerely hope that a huge clock speed increase is not what it take to make a better chip for AMD, as this is what Intel have been doing for so long with Netburst. Undeniably it is relativley straightforward way of increasing performance - but as a tech junky I would love to see a more AMD improve their tech (as intel have) to do more work per clock.

Both AMD and intel will release 4ghz chips at some point. Maybe not with their current generations - that means P4 and k8, not conroe or K8L.
April 12, 2006 2:38:25 PM

Quote:
It will be interesting to see exactly what speeds AMD will be able to reach within the next year, I do predict that it will be over 4GHz. After all, IBM has supposedly hit over 5GHz in the lab with their new power 6 chips, and AMD and IBM do share alot of CPU R&D.

I guess it's a matter of waiting to see exactly what AMD has gotten from their partnership.


Do you follow IBM much? In terms of their technology announcements overall?

My padre works for IBM (not in the research or tech divisions) so I end up hearing about new IBM tech a bit more frequently than most people. Its interesting because they typically are a bit ahead of the PC curve. The IBM power PC chip gave a horsepower edge to macs for years and now it's IBM chips in the Xbox 360 and soon-to-be playstation 3. Point is that AMD is somewhat likely to follow being that they share alot of tech development.
April 12, 2006 3:38:56 PM

lol! anyway, it doesn't matter to me...I'll just buy behind the tech curve and wait for everyone to dump their Sockete 939s and PD's and I'm all set :D  It's all about waiting and buying at the right time ;) 
Anonymous
a b à CPUs
April 12, 2006 5:45:42 PM

End of the term =|

Thermo-Dynamics of fluids and vector/matrix math. sucky classes!

I also had Networking wich was easy and I also enjoyed my Multi-Thread class this term. I was actually able to see the benefit of multi threading and how much more complicated it can be to debug the code!
Related resources
!