Chip giant Intel should strike a deal with Apple to produce 14- and 10-nanometre microprocessors, enabling it to fully utilise its moribund Fab 42 facility in Chandler, Arizona.
That is the opinion of Bill McClean, president of analyst group IC Insights.
With sales growth stalling and competition from the ARM eco-system shifting from mobile devices to the data centre, Maclean believes that Intel needs to be radical to re-ignite the company's growth.
"Faced with mounting pressure to turn things around, IC Insights believes that Intel must strike back and embark on extraordinary strategic moves in order to regain lost momentum," wrote McClean in a research note.
In a bid to win customers among tablet computer OEM and ODMs, Intel announced in November that it would pay the price difference between its newest Atom-based "Bay Trail" x86 processor and an equivalent ARM processor if the manufacturer used Intel's system-on-chip microprocessors in its product.
According to McClean, this programme could cost Intel as much as $500m - demonstrating both its keenness to expand into mobile computing, as well as the generally higher cost of Atom microprocessors.
"Does it make sense for Intel to let Fab 42 sit idle as an empty shell or will Intel consider another radical move and put the fab to use manufacturing ICs for another, possibly competing, company? It is no secret that Apple is looking to move some, if not eventually all, of its application processor production out of Samsung, its primary competitor in the smartphone and tablet PC markets," writes McClean.
The value of Apple's current business with Samsung stood at $3.4bn in 2013, according to IC Insights' estimates, which includes iPhone, iPad and iPod product lines.
While TSMC and GlobalFoundries are expected to win a slice of Apple's business at the 14-nanometre process node, IC Insights claims that Intel is in an ideal position to secure a chunk of Apple's application processor business at the 10-nanometer node using Fab 42.
"In fact, since TSMC made it known that it was not interested in dedicating an entire fab to an individual company, Intel could score a big foundry production win by dedicating a large portion of Fab 42 entirely to Apple," claims McClean.
However, Intel would either have to persuade Apple to adopt Atom - which would also make its own semiconductor design team, based on its PA Semi acquisition, surplus to requirements - or to use its ARM licence to produce microprocessors based on its biggest rivals' products.
By eliminating high entry costs for big data analysis, you can convert more raw data into valuable business insight.
A discussion of the "risk perception gap", its implications and how it can be closed