The Chip Insider®–Vision is not engineering. Lessons from Gary Dickerson, Ajit Manocha, and Pat Gelsinger.

Summary:

Maxims: Vision should be colorfully expansive with limited details Vision is not engineering. It's a tactical error to lend too much detail to a vision. While vision is important, it should only be a rough sketch of a future that's easy for followers to comprehend and execute to, while lacking details… details which often become lightning rods for tactical distractions or critics hunting your failures. General David Petraeus noted getting vision right as the first essential job of a strategic leader… Two semiconductor leaders that got this right are Gary Dickerson with his 'AI is the future' vision and Ajit Manocha with his $1T semiconductor industry forecast. One who got it wrong was Pat Gelsinger with his IDM 2.0 strategy for Intel.

Gary's AI vision was first enunciated almost 10 years ago in 2017, when he laid out a new "era of artificial intelligence…" The next stage of evolution for Gary’s vision came in July 2018, with a tactical linking of AI to what Applied Materials does, with a DARPA partnership… Next… Gary elaborated with his "New Playbook" strategy… Gary’s master class in grand strategy continued with his sustainability warnings … with a vision that linked AI as the "biggest inflection of our lifetimes" to a vision of power-hungry data centers consuming the world’s electricity, and the central role Applied Materials would play. …

Ajit Manocha's forecast that semiconductor sales would reach $1 trillion by 2030 came as the 2010s were coming to a close… Sales had hardly passed the $400B bar in 2017. A trillion dollars was a big number that forced CXOs to look up from their quarterly earnings microscopes a see a… bigger… longer-term picture. Strategically, this forced them to acknowledge needs they could not fulfill… all questions SEMI was perfectly suited to address, as the CXOs could only address them collectively. So, Ajit’s strategic vision of $1T formed into the multi-threaded vectors of a grand strategy…

Pat Gelsinger’s IDM 2.0 strategy … was a showman’s vision that was mostly tactical and light of vision … the linkage of IDM 2.0 to the tactical details and to customer value was obtuse… 5N4Y added to the difficulty as it shifted the problem internally... Sadly 18A should have been a victory lap for Intel, as it would reach production quarters ahead on TSMC. I8A was also a full generation ahead of TSMC with BSP… Yet Intel couldn’t follow through because of the tactical errors. It was a strategic communications exercise in wordsmithing with tactical traps that left shattered glass instead of hardened steel.

“Strategy requires a sense of the whole that reveals the significance of respective parts.”
— John Lewis Gaddis