siliconbruh999
Well-known member
Let me guess Politics ?
The Fall of Intel
- Thread starter Fred Chen
- Start date
MKWVentures
Moderator
Elimination of monthly roadmap changes and having less than 5 Program managers per Program. Powerpoint, Agile evangelist, problem solving mentors, timeline graph makers are going away (LBT is taking care of this).
siliconbruh999
Well-known member
Very Nice no need to have too much PMs/PPT guys what are Problem solving mentors lol first time I have heard about this.
Monthly roadmap change is a disaster.
Last edited:
Is this like a life coach but for bloated corporations?
Maybe it's just me but the other thread about new hiring trends at Intel caused me a small epiphany.
Things look bleak there right now with the layoffs and restructuring, but it now seems to me like what I've heard new military recruits go through at boot camp. (At least how it was in the 60s 70s and 80s.) The leadership at boot camp works to break down much of the recruits' psyches so they can build warriors in their place.
Perhaps Lip Bu considers the tasks at Intel like boot camp. Thoughts?
Things look bleak there right now with the layoffs and restructuring, but it now seems to me like what I've heard new military recruits go through at boot camp. (At least how it was in the 60s 70s and 80s.) The leadership at boot camp works to break down much of the recruits' psyches so they can build warriors in their place.
Perhaps Lip Bu considers the tasks at Intel like boot camp. Thoughts?
No. Program managers and scrum masters are alternative information and control paths for senior management who aren't capable of detailed oversight and technical judgement, or just choose not to. They think their development managers might cover up issues or fudge schedules, because they're personally invested. In reality, the program managers are less technical than the development managers, and just as likely to cover up problems, because they get dinged if the schedules slip too or objectives are not met.
I was just kidding, but you’ve described program managers very accurately.
But, very good program managers can make or break a project. Only if they’re (allowed to be) actually ruthlessly independent and accountable separate from product and engineering.
I've never met a great program manager. If I were LBT, I'd eliminate the entire function. If the engineering management chain can't figure out how to manage the project overall, I'd recommend firing them too, and find people who can.
I think program management makes sense when you have a project that crosses a lot of functional boundaries, where the engineering managers are more functional. Program managers are less technical than the functional mangers, because the functional managers need to be technically deep in their function.
That said, I think program managers often are not effective at their jobs. Someone should be in a technical/managerial position for a decade or two before they become a program manager in order to be effective... in reality they are often freshly minted MBAs who can make nice presentations.
In principle, I agree. But in practice, in four different companies, I never ran into a program manager who lived up to the cross-functional expectations, or really understood the development processes. Second and third level engineering managers should form a team and be leading this, not someone who follows a process recipe.
Agreed. But the people fascinated by engineering management are not typically people who enjoy process execution for the sake of good process.
@Daniel Nenni It happened I just found an insightful article you wrote 13 years ago on Semiwiki:
Intel Foundry; Intel says fabless model collapsing… really? (Daniel Nenni on 04-28-2012 at 7:00 pm) where there is an interesting discussion in response to the EETimes article: Intel exec says fabless model ‘collapsing’.
You highlighted with vision, "This is the entire fabless semiconductor ecosystem (Synopsys, Cadence, Mentor, ARM, TSMC, UMC, GlobalFoundries, QCOM, BRCM, NVDA, AMD, and hundreds of other companies) against Intel. Hundreds of billions of dollars in total R&D versus Intel’s billions."
It seems like an unconscious linking of the same thought of yours with another article I posted on LinkedIn: (Exploring Historical Paradigm Shifts in the Si Foundry Industry.). That article was my private observation and deliberation shared with TSMC’s former Chairman, Dr. Morris Chang, in May 2012:
[My Observation]
How TSMC can in-depth cooperation with eco-customers and eco-partners to jointly define, create and develop performance-differentiated 20nm low-power (e.g. static/standby or dynamic/active power) process flavors are the extremely important strategic challenge for coming years. The future generations of 20nm (i.e. FinFET) shall support to serve optimized low-power consumption (e.g. mobile APU) and high-performance (e.g. NVDA GPGPU) segment technologies ...
Looking back to 2012, surprisedly, TSMC indeed had two major tech nodes at the 20nm Si gate-length process:
• 20nm planar gate, mainly captive run for Apple (a bunt strategy)
• 16nm FinFET (with 20nm gate-length), crowd source serving both low-power consumption (e.g., mobile APU) and high-performance (e.g., NVDA GPGPU) segments with a few different flavors.
Attachments
It just doesn't make sense...