Intel’s 45 Nanometer Process: 300 Transistors on a Red Blood Cell

January 17th, 2007 |
Image for FaceBook

 
Share this post:
Facebook | Twitter | Google+ | LinkedIn | Pinterest | Reddit | Email
 
This post can be linked to directly with the following short URL:


 
The audio player code can be copied in different sizes:
144p, 240p, 360p, 480p, 540p, Other


 
The audio player code can be used without the image as follows:


 
This audio file can be linked to by copying the following URL:


 
Right/Ctrl-click to download the audio file.
 
Subscribe:
Connected Social Media - iTunes | Spotify | Google | Stitcher | TuneIn | Twitter | RSS Feed | Email
Intel - iTunes | Spotify | RSS Feed | Email
 

You might think Moore’s Law comes with an ancillary set of steps on how to adhere to it. The Law essentially says that technology develops so swiftly that chip engineers can pack twice as many transistors on a piece of silicon every two years. Performance jumps dramatically but the business proposition is even more compelling: the price for that performance drops, which is why we can afford laptops today that have more computing power than big computers did in the 1970s. In this podcast we visit Intel‘s Hillsboro, Oregon research facility and fab to talk with scientists who helped bring the newest 45 nanometer chip technology to reality.

Related Stories: IntelMooresLaw

Transcript:

Host: Jason Lopez – PodTech

Guest: Mario Paniccia – Intel

Guest: Mark Bohr – Intel

Guest: Kelin Kuhn – Intel

Jason Lopez – PodTech

I am Jason Lopez for PodTech.net. The race to make the smallest fastest chips is arguably one that has no end insight. Today’s chip companies can build transistor so small that more than 300 of them can fit on a red blood cell. The physical constraints are profound. Some of the connectors between components are made by gas and are just an atom thick. So, where to go from there? Well, here is an example, Intel uses light to test chips and it donned on engineers like Mario Paniccia, who heads the Photonic Technology Lab at Intel that lasers might be harnessed to do more than find defects.

Mario Paniccia – Intel

If we can send infrared lights through silicon to measure transistors, what if I could take silicon and send communication data through it and now do — everything we do today — modulating code to build optical components using silicon and the transmission properties of silicon. That evolution over the last couple of years has led to this program today, which we call Silicon Photonics.

Jason Lopez – PodTech

Although, laser-based chips are used away, scientists are still coming up with just-in-time innovations to build new processors that keep Moore’s Law on track, which essentially says that the number of transistors on a circuit doubles every two years as the cost to make that chip goes down.

Mark Bohr – Intel

Ten years ago many of us wondered whether we would ever get to this point.

Jason Lopez – PodTech

Mark Bohr is an Intel Senior Fellow, he spoke with me at the 45-nanometer lab at Intel’s Hillsboro, Oregon Campus.

Mark Bohr – Intel

Not only have we gotten to the point, but it probably didn’t take us quite as much time as we thought it would have.

Jason Lopez – PodTech

Now, when you say it didn’t take you quite as much time, ten years ago. How many years were you thinking?

Mark Bohr – Intel

Well, for the past ten years Intel has been developing a new generation of process technology every two years. Prior to that in the early 1990s, the pace was more of once every three years. So, we actually have picked up the pace over the past ten years. Again, we’ve gotten to this point this quickly is I think pretty impressive.

Kelin Kuhn – Intel

You don’t just walk down the street and start making transistors.

Jason Lopez – PodTech

Kelin Kuhn is the Device Manager for the Intel 45 nanometer chip in Hillsboro. She likens chips to Jumbo Jets. Just as no single person could design and construct a 400 seat plane, chip building requires massive resources and teams of people.

Kelin Kuhn – Intel

There is many years of technology innovation required; each technology builds on the previous technology. If you think about the technology today, if you compare our 45 nanometer technology to the previous technology, what you’re basically looking as a technology that takes about half the area, it’s about 20% faster and it’s about one-tenth the leakage of the previous technology. Now, keep moving that backwards, every technology generation before that built on the previous one and built on the previous one, you do that for many, many years and you can make very intrically small devices.

Jason Lopez – PodTech

The newest crop of chips from Intel is the result of some ingenuity. Mark Bohr says scientists are pushing the capabilities of traditional materials such as silicon wafers or polysilicon gate electrodes or thin oxide layers.

Mark Bohr – Intel

We’ve been scaling those dimensions, making them much smaller every couple of years, but lately we’ve been adding new materials to really enhance those transistors to get them to follow Moore’s Law. The average consumer really amazed at the amount of technology, high technology that’s in their computer at home, on their desk and their laptop. They could take a part of that chip, look inside they’d be surprised at just how much sophistication, new materials, ultra small dimensions are in that chip. The average chip may have 200 or 300 million transistors on it.

Jason Lopez – PodTech

For PodTech.net, I’m Jason Lopez at Intel’s 45-nanometer Lab in Hillsboro, Oregon.

< !—End Transcript -->

Copyright ©2006 PodTech.net. All rights reserved. Privacy policy

Tags: , , , ,
 
Posted in: Connected Social Media, Corporate, Intel, Intel Moore's Law, Technology