You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Say what you will, it is my personal opinion that Lip-But Tan is currently failing one of the key aspects of leadership. If you don't think Intel employees are stressed out and uncertain of the future then you don't know anyone that works at Intel. He has failed to give employees a vision that they can believe in. Does he have a plan and a vision? I'm sure he does. Based on his past success I'm hopeful that he will be able to pull it off. But he has done nothing to instill this vision to the rank and file. Perhaps he has communicated his vision clearly to upper management, but they aren't the ones that will make or break the company. That will come from having employees that understand the vision and are committed to it.
That isn't to say he has to spend all his time talking to employees or send out weekly cheer leading videos, but he needs to do more than give his employees the silent treatment. You are free to feel that employees should spend their free time cruising the internet looking for reasons to believe in their company, but I respectfully disagree.
IMO, communicating too much about "vision" could be harmful to Intel as a company. Rank and file employees need to execute well.
For example, Jensen Huang is a very visionary leader. He successfully bought Mellanox and tried to buy ARM, both of which were visionary moves. But did Huang communicate his vision and intentions clearly and widely about these acquisitions when he made them?
Success results from vision and execution, not from communication of the vision. Apple did not communicate its iphone plan in 2007. OpenAI did not communicate its progress on ChatGPT in 2022. On and on.
In Intel's case, they over-communicated about what 18A was going to do and its timeline, which likely caused TSMC to accelerate its own timeline a bit.
Say less, do more. Surprise your customers (positively) and competitors (negatively).
I don't think I said Intel didn't need to reduce headcount. However Lip-Bu Tan needs to give the employees that remain a reason to believe. I've seen what happens when a company doesn't do that. The employees will stick around in a bad job market, but they become less committed and are actively looking for an opportunity to be elsewhere.
Intel needs to change its culture. There is no way around it. AMD has 28000 employees. TSM has around 70k employees. It does not make sense for Intel to have 108k employees.
I think the reason layoffs at Intel are being discussed is that Pat Gelsinger didn't manage the company properly. He constantly mentioned "the Grovian culture" was back, but looking back, I don't know exactly what he meant.
IMO, communicating too much about "vision" could be harmful to Intel as a company. Rank and file employees need to execute well.
For example, Jensen Huang is a very visionary leader. He successfully bought Mellanox and tried to buy ARM, both of which were visionary moves. But did Huang communicate his vision and intentions clearly and widely about these acquisitions when he made them?
Success results from vision and execution, not from communication of the vision. Apple did not communicate its iphone plan in 2007. OpenAI did not communicate its progress on ChatGPT in 2022. On and on.
In Intel's case, they over-communicated about what 18A was going to do and its timeline, which likely caused TSMC to accelerate its own timeline a bit.
Say less, do more. Surprise your customers (positively) and competitors (negatively).
Gelsinger spent his time telling the world all the great and amazing things he said he was going to do and then lost credibility when he failed to deliver. That isn't what I'm talking about. I'm talking about giving employees within your company something to believe in.
You are free to disagree, but I believe the employee base executes best when they believe in what they are working towards. The people I typically talk to don't feel like they are getting that from Lip-Bu Tan right now. I've heard 20 and 30 year veterans say they have never seem morale lower and this isn't like it is the first time they've seen layoffs. It is largely because they feel completely ignored by the CEO and have no idea what the "new Intel" even looks like.
Gelsinger spent his time telling the world all the great and amazing things he said he was going to do and then lost credibility when he failed to deliver. That isn't what I'm talking about. I'm talking about giving employees within your company something to believe in.
You are free to disagree, but I believe the employee base executes best when they believe in what they are working towards. The people I typically talk to don't feel like they are getting that from Lip-Bu Tan right now. I've heard 20 and 30 year veterans say they have never seem morale lower and this isn't like it is the first time they've seen layoffs. It is largely because they feel completely ignored by the CEO and have no idea what the "new Intel" even looks like.
The low morale is likely a direct result of Intel's persistently low stock price. As an investor in $INTC, even I have low morale because I don't see near-term catalysts to make the stock price great again.
As an investor, I don't need vision talks from Intel's CEO. Instead, I want to see:
* A smooth ramp of the 18A process.
* Truly great products emerging from 18A, specifically PTL and CWF.
* Competitive data center (DC) products launching in 2026.
* A shift in Intel's culture to become fast and nimble.
All of these points primarily boil down to execution.
To boost morale, I suggest that LBT issues some one-time stock grants to remaining employees after layoffs, especially while the stock price is this low.
The low morale is likely a direct result of Intel's persistently low stock price. As an investor in $INTC, even I have low morale because I don't see near-term catalysts to make the stock price great again.
As an investor, I don't need vision talks from Intel's CEO. Instead, I want to see:
* A smooth ramp of the 18A process.
* Truly great products emerging from 18A, specifically PTL and CWF.
* Competitive data center (DC) products launching in 2026.
* A shift in Intel's culture to become fast and nimble.
All of these points primarily boil down to execution.
To boost morale, I suggest that LBT issues some one-time stock grants to remaining employees after layoffs, especially while the stock price is this low.
Recently, there was an article from SemiAnalysis about how AI models are evolving and their implications. The title is "Scaling Reinforcement Learning: Environments, Reward Hacking, Agents, Scaling Data."
2023 and 2024's language models were "next token" predictors; they primarily used tensor computations, which NVIDIA AI chips and ASIC accelerators handled very well. However, now the so-called "thinking models" or "reasoning models" are becoming more and more popular because they have far fewer hallucinations and are much more capable. A game changer, so to speak.
In my understanding, and as detailed in SemiAnalysis's article, CPUs may be more frequently needed in these reasoning models and future agentic models. This is because models may need to visit websites to gather information ("environment compute") and may need to render images or videos, tasks that require CPUs and traditional graphics processors to perform well.
That article also discussed implications regarding memory size and other factors. I somehow got the impression from that article that AI compute speed is no longer the bottleneck in the entire system.
To become a meaningful player in the AI/DC market, Intel really needs to be a forward thinker. They should engage with the best labs to understand their workloads' characteristics and determine how to make 2025/2026's AI models run faster and more efficiently. They also need to leverage IFS's packaging capability to accelerate time to market, among other things.
The low morale is likely a direct result of Intel's persistently low stock price. As an investor in $INTC, even I have low morale because I don't see near-term catalysts to make the stock price great again.
As an investor, I don't need vision talks from Intel's CEO. Instead, I want to see:
* A smooth ramp of the 18A process.
* Truly great products emerging from 18A, specifically PTL and CWF.
* Competitive data center (DC) products launching in 2026.
* A shift in Intel's culture to become fast and nimble.
All of these points primarily boil down to execution.
To boost morale, I suggest that LBT issues some one-time stock grants to remaining employees after layoffs, especially while the stock price is this low.
Does Intel have the money to be able to afford stock grants? I would not be surprised if they repeat the salary cuts that took place under Gelsinger to preserve capital. The good news is that other semiconductor companies with fabs are not hiring much either so Intel can afford to treat fab employees like dirt temporarily.
The low morale is likely a direct result of Intel's persistently low stock price. As an investor in $INTC, even I have low morale because I don't see near-term catalysts to make the stock price great again.
As an investor, I don't need vision talks from Intel's CEO. Instead, I want to see:
* A smooth ramp of the 18A process.
* Truly great products emerging from 18A, specifically PTL and CWF.
* Competitive data center (DC) products launching in 2026.
* A shift in Intel's culture to become fast and nimble.
All of these points primarily boil down to execution.
To boost morale, I suggest that LBT issues some one-time stock grants to remaining employees after layoffs, especially while the stock price is this low.
