INTEL CPU TO REACH 4-5GHz!!!

[AmdMELTDOWN]

just for those who want to know, I saw <A HREF="http://www.theinquirer.net/25050116.htm" target="_new">this</A> article at the Inquire

please no monkey buisness needed here, thanks! ;-}

"AMD/VIA...you <i>still</i> are the weakest link, good bye!"

 

[kavbear]

I don't get it. That's just some speculation from somebody that has nothing to do with Intel. Nowhere did he say that Intel's roadmap specified 4 or 5 ghz. He said "blah blah blah 2 ghz blah blah would lead us to believe blah blah 4ghz."

No man stands so tall, as when he stoops to tweak his rig.

 

[juin]

intel official say up to 10ghz is the limit
If i remember well all.

All this website is highly addictive is the new drug is take all me time.

<P ID="edit"><FONT SIZE=-1><EM>Edited by juin on 05/26/01 03:16 AM.</EM></FONT></P>

 

[mjdunn]

Yeah well how fast did they say the PIII would get? Didn't hit their mark...more marketing BS.

96.3 % of Statistics are made up.

 

[juin]

"All I know is by the end of 2002 both AMD and Intel are going to hit 15Ghz+"

I dont think any cpu will get to this point

"Yeah well how fast did they say the PIII would get? Didn't hit their mark...more marketing BS."

Maybe i just say is the official¸limit


<P ID="edit"><FONT SIZE=-1><EM>Edited by juin on 05/26/01 03:18 AM.</EM></FONT></P>

 

[bhc]

As I stated in one of my previous posts, I believe Intel or AMD (or anybody) will have difficulty to go 5 GHz (let alone 10 GHz) with their current core designs. It is NOT because the circuit won't work since I can certainly believe that the P4 architecture can go 10 GHz, but the real problem is heat management. Even though you can lower the core voltage with the 0.13 technology migration, the extra heat from high clock speed will still be a serious problem. I predict that to go beyond 5 GHz, you will probably need circuit solution to combat the heat problem, which may involve extensive redesign and re-optimization. They will probably still call it "P4" and the 21-stage pipeline will probably still be there, but it's really not the same as the current P4.

**Spin all you want, but we the paying consumers will have the final word**

 

[74merc]

I tend to agree with you there, eventually physics are going to get in the way.
conductivity can be improved to drop the temps some, reoptimize for smaller die, .10 micron or such, etc...
eventually it will reach a point where the transistors can't operate faster.
There will physically be a dead end point, 10 billion cycles per second is pretty damn fast, when will the speed of electrons moving over the conductive substance be the limit?
2ghz is fast, 10 is freakin awesome, but there will be further hurdles to pass if it comes to be.

and, to play devil's advocate, it was said in the days of Pentium 100s that we'll have to go to fiber optic because of the limit of electricity.
who the hell knows.

----------------------
Independant thought is good.
It won't hurt for long.

 

[Guest]

"There will physically be a dead end point"

We've been saying this for many many years.. 100 Mhz used to be some magical barrier, then 1 Ghz.. now its 10 Ghz.
I dont see no barrier yet.. just a lot of technological challenges.

---- Owner of the only Dell computer with a AMD chip

 

[noko]

So 32 bit chips are running out of steam, on with the 64 bit chips then.

 

[74merc]

yea, you can only spin an engine so fast, eventually the only way to make more power is to make it bigger.
:|
same thing with the processors.

----------------------
Independant thought is good.
It won't hurt for long.

 

[Guest]

Another illustration is you can either raise the speed limit to get traffic to move faster (CPU Speed in mhz)or widen the freeway. (number of bits)Another way to move stuff faster is to push the "cars" (data)through more efficiantly, which is what SSE does.

It's interesting to note that Nvidia claims the Geforces pump data faster than a Pentium 3 CPU, and the core only runs at 300 megahertz on the latest cards. The reason though, is the 256 bit wide freeway, compared with the Pentium and Athlon's 32 bits.

Aklein

Life is hard...Live with it.

 

[khha4113]

<b>compared with the Pentium and Athlon's 32 bits.</b>

I thought for <b>Pentium</b> it is 128 bit and <b>Athlon</b> is 64 bit.

Good or Bad have no meaning at all, depends on what your point of view is.

 

[Guest]

If you'll check the Intel CPU roadmap, you'll see that they're planning on releasing the "Itanium" 64 bit processor sometime this year. So, obviously, the Pentium 4 would be 32 bits, as is the Athlon.



Aklein

It's raining outside, and my lawn has grown a foot overnight!

 

[Guest]

"I thought for Pentium it is 128 bit and Athlon is 64 bit."

I believe you're thinking of the width of the data path to the L2 cache. PIII, P4, Athlon, Duron, etc. are all 32-bit CPUs.

"Arte es vida"

 

[peteb]

I think the author is taking their assumptions a little too far?

I'll agree that the roadmap for taking Northwood up looks good, but I think they draw the line too steep. I am not saying that Intel cannot release chips this quickly, but why would they?

As far as I can make out the 1Ghz PIII shipped 1H 2000, cannot get an accurate date, be glad to get one if anyone has it. The article predicts 2.4Ghz by Q2 2002. So we have ~ 2 years (maybe as litlle as 21 months or as much as 27 months) to add 2.5X processor speed. Not this article 'predicts' that Intel will ramp up another doubling of speed in 9 months? That doesn't make sense - if they did that, they'd just lock everyone out from buying now. It's good for companies to release strategically - not flood at the top end (assuming their plants let them). I cannot see why Intel or AMD would flood the market with high end processors when they can play the market out, leave headroom for a little competition between each other and maximise market revenue.

I do not doubt that technically it is possible for either company to release very fast processors but I doubt it makes economic sense.

-* This Space For Rent *-
email for application details