Reuters reports that Chen, who grabbed the reins in March, is considering a shift in strategy that would pivot the firm’s foundry focus toward its next-generation 14A process.

Sources say this could give Troubled Chipzilla a fighting chance at overtaking TSMC in key technology areas, which is vital if it wants Jobs’s Mob or Nvidia to even consider picking up the phone.

That means the 18A process, which former CEO Pat {Kicking] Gilsinger was pitching to external clients as if it were the second coming of silicon, has quietly stopped taking new business. Chen has lost faith in 18A’s appeal to outsiders and may be ready to pull the plug.

Industry watchers reckon that such a move would mean writing off hundreds of millions, if not billions, of dollars already invested in 18A and its spinoff, 18A-P. The money may be gone, but the embarrassment might be harder to wash off.

Chipzilla’s PR flacks refused to entertain the idea directly but were quick to point out that Intel itself is the “main customer” for the 18A process. Mass production of the Panther Lake laptop chip, supposedly “the most advanced processor ever designed and produced in the United States,” is still pencilled in for late this year.

“Chen and the senior management team are committed to strengthening the product roadmap, rebuilding customer trust, and improving financial health. We have set our priorities and will take the necessary actions to turn the tide,” Intel insisted.

Meanwhile, the cocaine nose jobs of Wall Street are watching closely, since bringing in external customers is make-or-break for Chipzilla’s foundry dream. The trouble is, the 18A process has slipped behind schedule, while TSMC’s N2 is progressing smoothly and its N3 process is already underway.

Sources indicate that the entire matter could be resolved as early as the July board meeting, although given the dollar signs attached, some board members want to delay any final decision until the autumn.

Intel is now trying to tailor the new 14A process to suit potential mega-clients, but it's unclear whether it can get it ready in time or at all. Meanwhile, it still owes Amazon and Microsoft a small batch of 18A chips under contract, meaning a sudden switch to 14A would be more wishful thinking than technical reality.

Amazon and Microsoft have so far declined to comment, probably because they're too busy double-checking Intel’s delivery dates. Chipzilla, for what it's worth, insists it will honour all existing commitments.

Once touted as Chipzilla’s return ticket to the top table, the 18A process now looks like another costly detour in its long, winding road to irrelevance.