Intel & TSMC together? A Semiconductor Love-Hate Story
Day 10 of Tracking Semiconductors Industry
Today, I came across a piece of news that’s pretty major in the semiconductor world—Intel’s stock just shot up by 14% with the arrival of its new CEO. But hold on, that’s not the full story.
Over the last five years, Intel's stock price has fallen nearly 60%, even as the Nasdaq Composite Index and S&P 500 have more than doubled. Investors have been frustrated with the company's inability to compete with rivals that have capitalised on AI-driven demand.
And there’s more - TSMC, it’s rival in foundry business has pitched to rescue it through a joint venture (sources say (every source is saying - sources say, god knows which source).
A name as big as Intel has been losing market share rapidly in recent years. And if we dig deeper, it’s clear why—it seems like Intel wanted to be the jack of all trades, but in trying to do everything, it's struggling to master any one thing. Let’s break it down and get into what’s really happening inside Intel.
How Intel Operates (And Why It’s Struggling)
Intel has always been different from companies like AMD, Nvidia, and TSMC because it does everything in-house—from designing to manufacturing its own chips. This vertically integrated model gave Intel an edge for years, allowing it to control its entire process, maintain quality, and push out cutting-edge products.
But the game has changed. The rise of AI and specialized chip companies has made Intel’s approach harder to sustain. While Intel is trying to handle multiple business areas at once, its competitors are excelling by focusing on just one thing.
Intel vs. Competitors: Who’s Doing What?
AMD & Nvidia → Only design chips and outsource production to TSMC, which allows them to stay agile and focus purely on innovation.
TSMC & Samsung → Only manufacture chips for companies like Apple, AMD, and Nvidia, perfecting their production process and maintaining a lead in chip-making technology.
Intel, on the other hand, is competing in both design and manufacturing, and that’s proving to be a tough battle.
Intel’s 5 Business Segments (And Where It’s Struggling)
Intel isn’t just working in one space—it’s trying to win in five different areas at once. That sounds great, but the reality is that it’s falling behind in most of them.
1. Data Center & AI (Losing to AMD & Nvidia)
AMD has surpassed Intel in the data center market, with faster and more power-efficient chips made by TSMC.
Nvidia is absolutely dominating AI chips, holding 90% of the market. Meanwhile, Intel is still struggling to create a truly competitive AI chip.
2. PC Processors (Losing to AMD)
Intel used to dominate the personal computer market, but AMD’s Ryzen chips (again, built by TSMC) have better performance-to-price ratios.
Many consumers and laptop brands are switching to AMD, chipping away at Intel’s once-untouchable market share. (Although, I still have an Intel processor!)
4. Foundry Services (Losing to TSMC & Samsung)
Intel wants to become a foundry like TSMC—manufacturing chips for other companies.
But it’s years behind TSMC in manufacturing technology and struggling to attract major clients.
5. Automotive (Fair Win: Mobileye is Strong)
The one bright spot is Mobileye, Intel’s autonomous driving division, which is doing well in the self-driving tech space.
6. Graphics & Accelerated Computing (Losing to Nvidia & AMD)
Intel launched its Arc GPUs to compete with Nvidia’s RTX and AMD’s Radeon.
But the response has been underwhelming, and Intel’s market share in this space is still tiny.
Lack of Focus or Just Overwhelmed?
So, is Intel lacking focus, or is it simply trying to do too much at once? It’s tough to say. But one thing is clear: while Intel is juggling multiple areas, its competitors are laser-focused on one thing—and they’re winning because of it. (it’s kinda needed lesson for me rn - focus)
Intel has big plans to turn things around, including investments in new manufacturing plants and next-gen chip technology. But will it be enough to catch up to TSMC, AMD, and Nvidia? Only time will tell.
Its manufacturing is slower than TSMC, leading to late product launches.
AMD and Nvidia focus only on chip design, so they innovate faster and outsource production to TSMC, which is already ahead of Intel in chip-making.
TSMC and Samsung don’t design chips; they only manufacture them, allowing them to perfect their processes.
One of the most obvious reasons why semiconductor design companies avoid using Intel’s foundry is the conflict of interest. Since Intel designs and manufactures its own chips in-house, why would a competitor trust Intel to produce their chips in the same facility?
It’s as simple as this—why would I ask my rival, who designs and manufactures his own chips, to produce mine too? That would mean helping them stay competitive while risking my own market share. It just doesn’t make sense for companies to rely on Intel when they can go to neutral foundries like TSMC or Samsung, which specialize only in manufacturing without competing in chip design.
Intel’s Financial Performance: A breakdown of the Q3 Concall (Foundry becoming a fund sucker)
Over the past five years, Intel's financial performance has reflected the mounting challenges in the semiconductor industry. Revenue has steadily declined from $77.87 billion in 2020 to $53.10 billion in 2024, with particularly sharp drops of 20.21% in 2022 and 14% in 2023.
Breaking it down by segment:
Client Computing (CCG) has seen fluctuations, but the Data Center Group (DCG)—a key growth area—dropped from $23.41 billion in 2020 to just $12.82 billion in 2024, highlighting its struggles against AMD and Nvidia.
Intel Foundry Services (IFS), aimed at attracting external customers, grew to $17.54 billion in revenue. However, this expansion has come at a heavy cost, with significant operating losses.
These struggles have contributed to Intel’s stock falling nearly 60% over the past five years. Financial ratios like gross margins have been under pressure, largely due to rising manufacturing costs and aggressive competition.
Management, as seen in recent earnings calls, has been focused on cost-cutting measures, including workforce reductions. However, manufacturing issues and strategic missteps—such as missing sales targets for its Gaudi AI chips—have put further strain on the company. In a tough but necessary move, Intel suspended its dividend to preserve cash for critical investments.
Looking ahead to Q1 2025, Intel expects further revenue declines, citing seasonal trends, macroeconomic uncertainties, and intensifying competition. The road ahead remains challenging, and Intel will need strong strategic execution to regain its footing in an industry that is rapidly evolving.
Intel Lagged in the Node race
(Nodes Btw, In semiconductor chips, a node refers to the size of the transistors, which are the fundamental switching components. It's a measure of a key dimension in the manufacturing process, typically expressed in nanometers (nm). A smaller node number signifies smaller transistors. Smaller transistors allow more of them to be packed onto a single chip, leading to increased processing power, faster speeds, and improved energy efficiency.
Smaller nodes = More transistors can be packed in a chip = more efficiency)
This is where the lag arised . Intel used to be top leader with their 14nm tech. But only some server CPUs & mid-range graphic cards from around the mid-2010s to early 2020s utilized 14nm technology. The game's moved on, and everyone wants chips that are smaller and faster. Now, Intel's saying they've got this awesome new tech called 18A that could actually put them ahead of everyone again! That's the good news.
But here's the thing: Intel's had a bit of a rough patch lately with getting their new chip-making tech right. They've faced delays and haven't quite delivered as promised in the past. So, while this 18A news sounds great, folks are a little "seen it before" and are waiting to actually see it work and be mass-produced reliably before they fully believe the hype. Basically, everyone's hoping Intel's turned a corner, but their past hiccups mean there's still some healthy skepticism floating around.
On top of that, Intel's really struggled to get a foothold in the booming Artificial Intelligence (AI) market. Nvidia, with their super powerful graphics chips (GPUs) made by TSMC, has basically taken over because their chips are great at handling the kind of parallel processing that AI needs, and they've built up a whole ecosystem around it. So, Intel's been playing catch-up there too.
What if TSMC & Intel JV got real?
TSMC, who's been beating Intel at making chips, might actually team up with them in a joint venture! TSMC would run Intel's factories with a minority stake (less then 50%) , and even big players like Nvidia and AMD could get involved.
Why would Intel do this? Simple: TSMC is the best at making advanced chips, and Intel needs their help to catch up. It would also give Intel's struggling factories a financial boost and could unlock government support for US chipmaking. For TSMC, it means getting access to Intel's customers and expanding their US operations.
This could seriously change Intel's game. They might move away from making their own chips and focus more on just designing them, like AMD and Nvidia. This could free them up to innovate faster.
This partnership could also be a big deal for the world's chip supply. It might spread out where advanced chips are made, which is good for everyone. But it could also change the competition in the chip-making world. Plus, with the US wanting more chips made at home, this could get a lot of government attention and maybe even some policies to help Intel out.
Zoho's founder, Sridhar Vembu, thinks Intel needs TSMC's help now because for years, the smartest people in Silicon Valley went to other tech areas instead of chip manufacturing (fabs) due to better financial prospects. Investors also ignored fabs. This lack of talent in chip manufacturing is a key reason why Intel is behind now.
It's kind of a surprising "rivals into lovers" situation, with TSMC potentially helping out Intel, even though they're competitors. It's definitely a story to watch and see if it actually happens!
BTW Happy Holi!
-Abhishree