Moving the 800 lbs. Gorilla

About > Themes > Moving the 800 lbs. Gorilla

In 1998 a global team of leading software developers assembled to help Advanced Micro Devices (AMD) design the 64-bit microprocessor to once again challenge Intel’s dominance in the personal computer market.  The team spanned five continents and had never physically met, but they all knew one another’s work and shared a passion for engineering perfection.  Their eighteen month effort produced the technical specifications for the most advanced mass-market processor to date.  They worked side by side with one another and with AMD’s most senior hardware architects and veterans of the legendary Digital Equipment Alpha processor design team.  The team completed their work and went on to other interests but took pride in helping AMD give the world the Athlon 64 processor, the last remaining Intel alternative in the personal computer market.

Senior engineers had crafted the basic structure of what was to be AMD’s assault on Intel’s Itanium, the crown-jewel of commercial PC computing.  Unlike Intel, AMD was very much a hardware company and was outnumbered by Intel fifty to one.  There was no vast software development talent pool on staff, no consultants available to hire, no large sums of marketing money to offer.  Leading the enterprise Independent Software Vendor Relations group at AMD, I was challenged with identifying technical leaders in the global software industry, to recruit these experts to help with completing the high-level design of this processor, and to motivate them to do this for free.

Top notch talent does not sit idle; dynamic individuals like these in the software industry are already heavily engaged in ventures of their own.  They do not look for jobs, jobs look for them.  They do not advertise, their work advertises for them.  They are found at the core of major software innovators like Oracle, IBM, Novell, Computer Associates and others.  My strategy was to draw on software development experience, find them, understand their work, and only when prepared enough to understand each of their respective fields and motivations, reach out to them and convince them to help us.  Moreover, all this had to be done quickly and without promise of elaborate partnerships or even financial reward.

I started with the Athlon 64 architects, and together created a profile of the specific talents we needed:  operating systems and data-crunching engines.  I reduced the scope to the software world of Oracle, Sybase, IBM, Novell, Computer Associates, and the Linux community.  Microsoft had a dedicated and separated AMD team.  Over a couple of months, I worked from one contact to the next until like a global jigsaw puzzle our dream team of software experts had materialized.

Each team member accepted the challenge: “what features would you like to see in a processor five years from now?”  How could they refuse?  Intel had sent them specifications and prototypes of their new processor but did so very late in the design phase, much after the fact to facilitate any changes they may have wanted.  All felt slighted by Intel’s approach, particularly those who were directly competing with Microsoft, and so readily enlisted with us.

The team never met but knew of one another’s work.  We scheduled weekly international conference calls to review sections of 200-plus page specification documentation for the Athlon 64.  Our team consisted of three US-based executive vice presidents of engineering, one chief architect in UK, another in the Czech Republic, and one chief technology officer in South Africa.  Each call started promptly, everyone was on time, focused and ready to work.  We could hear the Czech Republic team member’s family and baby boy in the background week after week as he took calls at night.  After a few weeks the global team found out that he lived in a two-room apartment and four people were trying to sleep in the same room during our call.  Busy executives in Silicon Valley, chief technologists in the UK and South Africa, world-renown processor architects in Austin, all changed their calendars to accommodate a better time for the call.

The team was cemented in purpose.  Week after week we tore through reams of data, hundreds of lines of code, dozens of impromptu experiments, occasional temper tantrums, and my regular nagging the team to reach consensus.  The team never quite revealed all the personal sacrifices they made to spend the time outside of the call to contribute to the specification of this processor, but they never really needed to – the finished work spoke for them.

AMD received hundreds of hours of expert consulting for the cost of a few long distance conference calls.  Each expert team member was professionally gratified to have passed on their expertise to be implemented in hardware for their own software to run better at some point in the future.  I learned that physical, cultural, economic and social separation does not stand in the way of a united motivated team focused on overcoming a challenge to which they are properly matched.


%d bloggers like this: