One thing leads to another at Intel: it's make-or-break time for the 14A process.
On the heels of its president Lip-Bu Tan's announcement to bring back hyper-threading technology on its next processors, Intel today confirms the risks the company is taking with its next engraving nodes. Intel is no longer completely up to speed with Samsung and, above all, TSMC. Its next etching node, the Intel 18A, would only enable it to come up just short of TSMC, which is why the company has already decided to reserve this technology for its processors alone: external customers are now targeted for the Intel 14A, which is still a long way from being ready.
TSMC's roadmap seems to put Intel and Samsung far behind © Tom's Hardware
In fact, Intel's technological ambitions could even be described as a last-chance node. As Lip-Bu Tan reminded us, Intel has reallocated funds from promoting the 18A to its partners to the 14A. More modern, this etching node is better able to compete with TSMC and, as a result, attract new customers. These new customers are essential to ensure the profitability of such a fine and... costly etching process. Intel makes no secret of the fact: if no major customer is found for 14A, it will be the end of Intel's ambitions for the most advanced processes.
Processes whose costs have become exorbitant, while etching techniques are increasingly difficult to master. It's no coincidence that, a few months ago, it was Samsung that announced its intention to stop competing directly with TSMC on 2 nm. Samsung prefers to focus on improving the profitability of the nodes currently used by the South Korean. The problem is that, while Samsung has already accepted TSMC's dominance, and Intel looks set to do the same barring a major turnaround on 14A, TSMC could well find itself alone on the most advanced etching nodes: a monopoly that should please no one, since, without a competitor, TSMC could take advantage of the situation to significantly raise its prices.
