Intel is looking to unlock another source of funds from an initial public offering of its programmable-chips division. Early estimates suggest it could be a multibillion-dollar boost for the chip maker as it invests to take on Nvidia and other rivals.
Intel (ticker: INTC) is preparing an IPO of its Programmable Solutions Group in the next two to three years. The company said it would improve the performance of the unit, open it up to outside investment, and give Intel a boost to its coffers.
Intel shares were flat, at $35.70, in recent trading.
The PSG unit is likely worth around $16 billion, based on normalized annual revenue of $2 billion, according to Oppenheimer analyst Rick Schafer. He kept a Perform rating on Intel stock with no target price.
That’s a disappointing price tag on the face of it, considering that Intel acquired the business with its $16.7 billion acquisition of Altera in 2015. Still, managing PSG as a stand-alone business from the start of next year ahead of its planned IPO could give Intel some time to boost that valuation.
Raymond James analyst Srini Pajjuri wrote that the unit could be valued in a range of $19 billion to $23 billion at the time of its IPO, assuming its annual revenue climbs to $2.5 billion. He has an Outperform rating on Intel stock and a $42 target price.
The PSG unit specializes in field-programmable gate array (FGPA) chips, which can be easily reprogrammed after manufacture, making them useful for specific tasks and popular choices in a range of industries such as defense and telecommunications.
“With a business like FPGA, Intel has insight into customer growth for the next couple of years to make [the] IPO valuable to them. Also, if that IPO money is coming in two to three years, you have to wonder what Intel is planning to fund at that time,” said Ian Cutress, a semiconductor industry analyst and CEO of More Than Moore.
Intel has said it intends to keep a majority stake in the PSG unit, but that it could explore private investment in the business ahead of the IPO to accelerate its growth.
It could use the money to invest in multiple fronts. CEO Pat Gelsinger is pursuing an expensive effort to expand Intel’s contract chip-making operation. The company also needs funds to build out its artificial-intelligence chip pipeline to compete with
Nvidia
(NVDA) and
Advanced Micro Devices
(AMD), which are threatening its market share in data centers.
Truist Securities analyst William Stein said the strategic rationale of the spinoff seems to be allowing Intel to refocus its AI efforts. He has a Hold rating and a $37 target price on the stock.
“It appears the company is retaining the parallel compute architecture that customers have chosen for AI training workloads (Habana Gaudi GPU), and divesting (at least a portion of) the other parallel compute architecture (FPGA) that once appeared to be a strong contender for AI acceleration but has not precisely lived up to that prospect,” Stein wrote.
There’s also an intriguing executive change as part of the spinoff. Sandra Rivera, currently general manager of Intel’s key Data Center and AI Group, will move over to head the PSG unit. That means Intel will be looking for her replacement just as Nvidia is making greater inroads in the data-center market due to the AI prowess of its chips. Investors might like to see an external executive with strong experience in AI chips as Rivera’s replacement to help stave off the threat from Nvidia.
Write to Adam Clark at [email protected]
Read the full article here
