Intel denies 10nm node cancellation rumour

Intel has hit back at claims that it is giving up on its quest to hit volume production on a 10nm semiconductor process node, issuing a rare response to the rumours claiming to be 'making good progress on 10nm.'

Claims that Intel's troubled 10nm process, which is now several years overdue, was being abandoned altogether began circulating late yesterday following the publication of an article by SemiAccurate citing anonymous sources that 'the process is indeed dead.' With former AMD production division GlobalFoundries have put its own 7nm process node on indefinite hold after hitting many of the same problems that have plagued Intel, the story seemed plausible - but Intel, in a rare response to rumour and three days ahead of its scheduled quarterly earnings call, has claimed different.

'Media reports published today that Intel is ending work on the 10nm process are untrue,' the company, which traditionally responds to such stories with a generic statement as to how it refuses to comment on rumour and speculation, claimed on its Intel News Twitter account late yesterday. 'We are making good progress on 10nm. Yields are improving consistent with the timeline we shared during our last earnings report.'

The 10nm process node was originally scheduled for 2015 as part of Intel's tick-tock development cycle which saw it alternate biennial process node shrinks and architectural improvements. Following delays to its implementation, the company admitted defeat on tick-tock and switched to incremental architectural improvements each year followed by a process node shrink as-and-when possible - and 'when possible' has been stretched increasingly further out, with Intel admitting that mass-produced 10nm parts won't appear until 2019 while its 10nm server processors are delayed until 2020.

That revised schedule is one with which Intel is sticking, however: Mass production of 10nm by 2019, with the company likely to stick at that process node for some considerable years after while it works on the next sticky shrinkage down to 7nm or below.