Intel predicts 10GHz chips by 2011

[Crispy]

Intel is predicting that its microprocessors will hit 10GHz by the year 2011. In addition, it is currently working on a system bus that is 10 times faster than its upcoming 400 MHz (4*100MHz) Pentium 4 system bus, working at effective speeds of around 4 GHz. What will we do with all that processor power? Intel is working hard to bring high-end applications to your desktop featuring much more video, speech interaction, and more complex and functional user interfaces.
Read more at ZDNet.

ROB'S OPINION
Hopefully, most of you are familiar with Moore's Law, so this doesn't come as a surprise to you. If not, Moore's Law states that the number of transistors in a common microprocessor will double every 18 months. As well, it can be applied to processor speed and many other computing/technology metrics. This law has more or less (no pun intended) held true since the transistor was invented. So, assuming that early 2001 is a time when 1 GHz processors are rampant let's see what we get if we apply Moore's Law:
early 2001: 1 GHz
mid 2002: 2 GHz
early 2004: 4 GHz
mid 2005: 8 GHz
early 2007: 16 GHz
mid 2008: 32 GHz
early 2010: 64 GHz
mid 2011: 128 GHz
This is very interesting indeed. Intel appears to be underestimating progress in 2011 by a full factor of 10. Even by doing that, it's still making headlines with its predictions. Some will call this crazy, but those that yell the loudest have the shortest memories.

Source: Intel predicts 10GHz chips by 2011 ? Computer Chips & Hardware Technology | Geek.com
​

 

[Dave76]

Hey, where's my 64 GHz processor?

I would settle for 10 GHz, I guess.

Looks like Intel's prediction from 10 years ago was closer that Rob's and his interpretation of Moore's Law.

And Intel was over by more than 100%.

 

[Win7User512]

Intel is predicting that its microprocessors will hit 10GHz by the year 2011. In addition, it is currently working on a system bus that is 10 times faster than its upcoming 400 MHz (4*100MHz) Pentium 4 system bus, working at effective speeds of around 4 GHz. What will we do with all that processor power? Intel is working hard to bring high-end applications to your desktop featuring much more video, speech interaction, and more complex and functional user interfaces.
Read more at ZDNet.

Source: Intel predicts 10GHz chips by 2011 ? Computer Chips & Hardware Technology | Geek.com
​

Didn't they get rid of the bus with the i3-i7's? Why go back?

And it would only be 10x faster bus than P4? Seems a bit slow, as P4 was some time ago...

 

[Crunchy Doodle]

I can twitter my facebook just fine at 2.66 GHz. What's the big hurry?

Bye.

 

[WindowsStar]

Moore's Law:
early 2001: 1 GHz
mid 2002: 2 GHz
early 2004: 4 GHz
mid 2005: 8 GHz
early 2007: 16 GHz
mid 2008: 32 GHz
early 2010: 64 GHz
mid 2011: 128 GHz
This is very interesting indeed. Intel appears to be underestimating progress in 2011 by a full factor of 10. Even by doing that, it's still making headlines with its predictions. Some will call this crazy, but those that yell the loudest have the shortest memories.

The reason Ghz (speed) has stopped getting faster and faster is because Intel cannot build a laser that can make the CPU any smaller. Currently the laser used to make CPU’s is about 3 blocks long and burns 100’s of CPU’s on a 14 inch disc of silicone. The way Intel keeps getting speed out of the processor was to go smaller or reduce the heat and make the chip run faster. Since Intel cannot go any smaller they started to making multi-processors, HT and Dual Cores etc. If Intel’s is saying 10Ghz by 2011 then they found a way to make the CPU smaller or they have a much better way to cool the processor that does not take a 14 inch by 14 inch block of aluminum heat sink. I know there were some people experimenting with glycol coolers much smaller than water coolers and the technology is proven, it was used in the early days of projection TVs.

 

[Scotteq]

Why is a story from July of 2000 news? Seriously. This is a Ten Year Old article.


We already learned that as you increase clock speeds the voltage/heat increases. Not to mention the issues of current leakage as component sizes have gotten smaller. Intel abandoned the quest for higher and higher clock speeds, rethought their approach, changed the philosophy to 'Instructions Per Clock', and brought us the Core 2 series.

 

[Anthony]

mid 2011: 128 GHz?

More like 12 cores by the end of 2011, We have hit the ceiling with the GHz!!

 

[WindowsStar]

mid 2011: 128 GHz?

More like 12 cores by the end of 2011, We have hit the ceiling with the GHz!!

But multi-cores give you diminishing returns. 4 cores seems to be about the best speed we can get, the tests on the new AMD 6 core have proven to be a small fraction better than a 4 core.

Amdahl's law - Wikipedia, the free encyclopedia

 

[Lebon14]

mid 2011: 128 GHz?

More like 12 cores by the end of 2011, We have hit the ceiling with the GHz!!

But multi-cores give you diminishing returns. 4 cores seems to be about the best speed we can get, the tests on the new AMD 6 core have proven to be a small fraction better than a 4 core.

Amdahl's law - Wikipedia, the free encyclopedia

I think you can blame the softwares too. Also, not every piece of software would have a use to multi-core CPUs.

 

[Thorsen]

Well I saw something last week that was related inversely: DailyTech - NVIDIA VP Declares Moore's Law Dead, GPUs Are Computing's Only Hope
(I know its DailyTech, but still....)


I knew the day would come when Moore's Law started to exhibit diminishing returns. Although a 10ghz processor would be nice, I thought a quad running at 3ghz was effectively 12ghz of processing equivelant power.


Proof that Moore's Law is alive and well: NEWS

Google has a new machine it is testing: Google demonstrates quantum computer image search - tech - 11 December 2009 - New Scientist

 

[solarmystic]

People people, we've all learnt that frequencies aren't the best way to gauge performance no more.. have we learnt nothing from the debacle known as Netburst and the Megahertz Myth?

AMD had the right idea with their low frequency, low staged yet wide pipelined Athlon X2s and it took the smarts of a small Intel Jewish team in Haifa, Isreal to put Intel back into the fray with the Pentium Ms which became.... the Conroe and they haven't looked back since then...

 

[w7101]

I can twitter my facebook just fine at 2.66 GHz. What's the big hurry?

Bye.

What are U talking about ..? .my intel 1.4Ghz celeron "single core"..
can Twitt fine ... with wds 7 ...Thank you very much ..

 

[Scotteq]

Well I saw something last week that was related inversely: DailyTech - NVIDIA VP Declares Moore's Law Dead, GPUs Are Computing's Only Hope
(I know its DailyTech, but still....)


I knew the day would come when Moore's Law started to exhibit diminishing returns. Although a 10ghz processor would be nice, I thought a quad running at 3ghz was effectively 12ghz of processing equivelant power.


Proof that Moore's Law is alive and well: NEWS

Google has a new machine it is testing: Google demonstrates quantum computer image search - tech - 11 December 2009 - New Scientist

Moore's Law is in regards to the Transistor Density which can be economically put into a semiconductor. Clock speeds are not a part of this rule.

 

[Thorsen]

Moore's Law is in regards to the Transistor Density which can be economically put into a semiconductor. Clock speeds are not a part of this rule.

True. I looked at this rule after posting. I always thought of it as a general computer term. things doubling was the standard for lots of things for a while.

In general RAM capacity, RAM speed, CPU speed, Bus speeds and HD capacity all followed this rule. Whenever anything was upgraded, it was usually double the previous years speed or capacity. Computing in general always seemed to follow the rule. Even going from single core to dual to quad. all are attempts to upgrade and all were in doubles.

 

[Zen00]

The fastest CPU clock ever was a little above 8ghz, on a Celeron.

 

[Crispy]

I know there has been a CPU that was clock around the 10Ghz mark but I cant remember where I saw it, If i find it ill post it.

 

[WindowsStar]

Moore's Law is in regards to the Transistor Density which can be economically put into a semiconductor. Clock speeds are not a part of this rule.

True. I looked at this rule after posting. I always thought of it as a general computer term. things doubling was the standard for lots of things for a while.

In general RAM capacity, RAM speed, CPU speed, Bus speeds and HD capacity all followed this rule. Whenever anything was upgraded, it was usually double the previous years speed or capacity. Computing in general always seemed to follow the rule. Even going from single core to dual to quad. all are attempts to upgrade and all were in doubles.

I think you meant: Amdahl's Law

Amdahl's law - Wikipedia, the free encyclopedia

 

[motc7]

To the OP I would say this.

In business, it's best to underestimate and then over-deliver, than to overestimate and under-deliver. If you do the former of the two, that falls into what I call "Mr. Scott's Law" or simply "Scotty's Law". It states that you don't really tell people how long or how much you can do, but give them an underestimate that way when you pull it off in less time and with more than they asked for, you look like a miracle worker."