There is a specific type of dispute that only occurs in the tech industry; it is quiet, technical, and fought through X posts, but it has significant implications for the entire sector. Elon Musk launched that argument last week, specifically targeting IBM, and it addressed a topic that the semiconductor industry has been subtly uneasy about for years: what do chip process names really mean these days?
In a dramatic announcement, IBM revealed what it called a 0.7-nanometer manufacturing technology, which it claimed to be the world’s most sophisticated chipmaking technique. In a blog post, the company was honest about the fact that “7 angstroms doesn’t correspond to the width of the contacted metal wires in the chip, as it traditionally did many generations ago.”” Alright. They said it, at least. However, Musk reacted to a post on X that described the naming convention as “highly misleading.” Musk is never one to overlook something when he thinks it merits examination. He concurred. pointedly.
His recommendation was simple: rename process nodes based on how many atoms make up the smallest feature size. “That would be most accurate imo,” he wrote. It’s a brief comment that’s simple to ignore. However, it starts a genuinely complex discussion about how the semiconductor industry interacts with the outside world, or more often, fails to do so.
This issue is not brand-new. In 2021, Intel completely reorganized its naming scheme, renaming its 10-nanometer process “Intel 7” and its 7-nanometer process “Intel 4.” Not as a result of the transistors getting smaller. Because the branding had become so detached from physical reality, Intel would have appeared to lag behind rivals if it had continued to use traditional labels. The nanometer race had subtly shifted from measurable transistor dimensions to marketing positioning, and TSMC and AMD had been playing a similar game.
When interpreting Musk’s criticism, this context is important. It’s possible that people have long felt that something didn’t add up, which is why his objection is so compelling. The label begins to feel more like a brand narrative than a technical specification when a company announces a “0.7-nanometer” chip and then explains in the same breath that the number doesn’t describe any physical dimension of the chip.
Here, Musk is also playing his own game. His Terafab project, which aims to produce up to a terawatt of computing capacity within a year, puts him in direct competition with IBM for talent, attention, and funding in the same market. It is genuinely difficult to determine whether his criticism is purely principled or partially strategic. Most likely both, as these things typically are.
Even so, when the billionaire posturing is removed, his fundamental argument remains valid. Outsiders are unable to comprehend the naming culture that the semiconductor industry has developed. This opacity causes serious confusion for investors, policymakers, and the general public who are attempting to track the chip race, particularly as it turns into a geopolitical issue. It feels like more than a naming dispute to watch this unfold on social media, with Musk as the challenger and one of the oldest tech firms in America as the target.
Clearer standards are thought to be long overdue in the industry. It’s still unclear if this reform is the result of regulatory pressure, public pressure, or manufacturers’ collective decision to end the confusion. For the time being, IBM’s 0.7-nanometer designation is still in effect, Musk’s atoms-based substitute is still only a proposal, and the discrepancy between the names of chips and their actual physical characteristics remains as great as it has been for many years.
