: News
: Reviews
: Editorials
: About
: Contact
: Advertising
: Privacy
: Links
: Forums



Q&A with Intel's George Alfs


An IDF Kickoff

by Josh Walrath


            This week I had the chance to fire off a series of questions to George Alfs of Intel.  As you may well know, Intel is kicking off its Fall IDF (Intel Developers Forum) on September 7.  Intel kicked off the Summer with the rapid fire release of the LGA-775 processors, PCI-E, Azalia Audio, and the 915 and 925 series of chipsets.  After that, very little was heard from Intel as compared to the releases that AMD has done.  Among these milestones that AMD has talked about are the dual core tapeout, 90 nm products shipped for revenue, and finally the showing of working dual core silicon.

            Now it is Intelís turn to show off what it has been working on this summer.  Before the IDF opened, I wanted to ask George a few questions to get us all warmed up for the occasion.  George was kind enough to get these back to me very quickly, and right in time to get this posted before IDF opens its door for this season.  Georgeís answers are in bold.


How has the company reacted as a whole to Craig Barrettís very public email rebuke to the employees of Intel on their overall performance as of late?

Alfs: It wasnít a rebuke.  Craig wanted to let employees know that he is working closely with executive staff to ensure that Intel is doing the right steps at the right time.

AMD has really pushed the performance and feature envelope with their Athlon 64 products.  What is Intel doing to keep the market focused on Intelís parts and its technology?

Alfs: The real question is, are PC users looking to open their word documents faster or evolving their PC usage as it becomes the focal point for entertainment, information sharing and business.  We have some of the highest performing products in the world, and we will continue to improve performance with faster bus speeds, frequencies, etc.  But, weíre also delivering key technologies inside Intel silicon like Hyper-Threading, which allows for faster multi-tasking and for running multi-threaded applications.  Our work with software developers on HT (remember that the OS and application infrastructure was ready at HT launch) helps set the stage for multicore in 2005.  And, we're developing La Grande (security) and Vanderpool (Virtualization) technologies and others yet to be disclosed. Combined, we are delivering super-fast platforms we are traditionally known for, but also increasingly adding technologies that address what computer users are demanding the most as the PCís role evolves.

Much has been said about the 90 nm node that the latest Pentium 4ís are produced on, but what is being done behind the scenes to ensure good yields, and how is Intel engineering around the heat and power issues of its latest products?

Alfs: Our 90nm process is mature.   We make a million Pentium 4 processors a month on this process, as well as Dothan-based mobile products.    I see a lot of industry pundits confuse our industry leading process technology with individual product characteristics.   As far as Prescott, we ensured industry support was in place at launch for running millions of transistors at incredibly fast speeds.   We will continue to ramp up our products capabilities, as well as ensure infrastructure is in place to handle those capabilities.

The Pentium-M has been a great success for Intel, as well as the design team in Israel that developed it.  What kind of advances will we see in the near future with this product?  Does it suffer from the same heat and power issues that have plagued the Pentium 4 E models?

Alfs:  Both our Banias and Dothan based Centrino products are incredibly successful.   Dothan increases Baniasís cache size to 2 megabytes, increases clock speed and other capabilities, and stays in the same thermal envelope.    I donít accept your premise on Prescott, and Dothan shows that our 90nm process is capable of producing extremely cool running processors.

With Tejas and later Pentium 4 based products cancelled, how will Intel address the desktop space in the future?  Will the Pentium-M core be the basis for later generations of desktop processors?  If so, what kind of architectural changes will need to be made to make it more competitive in this market?

Alfs: Sounds like you need to come to IDF this week!   Weíve already talked about how we are taking our entire mobile, desktop, and server products to multicore technologies in 2005.    We havenít yet talked about which cores will be used in all cases but stay tuned.  As far as our Pentium M core, it is far and away the most successful mobile core on the market, and we intend to keep it that way.


Next: Q&A Continued


If you have found this article interesting or a great help, please donate to this site.


Copyright 1999-2004 PenStar Systems, LLC.