Intel’s stock leads Dow decliners for second straight day, as cost-savings goals could stretch out further

Intel’s stock
INTC,
-4.54%
closed down 4.5% at $34.69 in Wednesday’s session, as concern over profit growth cast a shadow over the chip maker’s AI announcements.

Intel shares sustained a 4.3% drop Tuesday to close at $36.34, after Intel unveiled its next generation of AI hardware on the opening day of its developer conference. Intel’s stock last logged two consecutive sessions as the weakest performer on the Dow Jones Industrial Average
DJIA
back in June, when the company disclosed news about its foundry services.

Chief Executive Pat Gelsinger’s “keynote tone was positive” as he showcased Intel’s new AI goods at the company’s Innovation conference on Tuesday, according to Bernstein analyst Stacy Rasgon. But the tone from Chief Financial Officer David Zinsner later was “a bit more sanguine,” particularly when it came to gross margins.

Intel’s margins have been a troublesome point for the former Micron Technology Inc.
MU,
-1.46%
CFO since he joined Intel.

On his first earnings call as Intel CFO in late January 2022, Zinsner said he felt “comfortable with a 51% to 53% range in gross margin for the year,” after having just reported fourth-quarter gross margins of 55.4%.

While Intel went on to report margins of 53.1% for the first quarter of 2022, that was the last time margins came close to the forecasted range. Since then, the closest margins have come to the target were the 44.8% margins that Intel posted in the second quarter of 2022, while the lowest they have dropped was to 38.4% in the most recent first quarter. Then, Intel posted its largest quarterly loss on record.

While Rasgon said many of the margin drivers Zinsner mentioned Tuesday are nothing new, “his expectations for margin expansion next year sounded muted” compared with current Wall Street expectations.

“I think year over year, we’ll see gross margin expansion,” Zinsner told analysts during Tuesday’s webcast. “It may not be hundreds and hundreds of basis points next year. It may be more modest for the reasons I talked about from a headwind perspective. But we do expect margin expansion next year.”

Read: Intel’s stock slides to become Dow’s worst as inventory ‘air pocket’ overshadows AI chip rollouts

The Bernstein analyst said his Street consensus showed expectations of a 550-basis-point improvement in gross margins for calendar 2024, suggesting that the company’s forecast for $4 billion to $5 billion in annual savings on the cost of goods sold “might be a 2025 story at best.”

Rasgon has a market-perform rating on Intel with a $34 price target.

Rasgon titled his note to clients “Some bark, some bite …” not only in reference to the essence of Intel’s presentation, but also as a way of apologizing for his dog “Chip,” whose barking briefly interrupted the webcast.

Over at Oppenheimer Research, analyst Rick Schafer, who also had a perform rating on Intel shares, said he doesn’t see Intel as a serious threat to Nvidia
NVDA,
-2.94%
in the accelerator market.

“We see x86 headwinds in PC/server as ARM penetration and accelerator adoption grow,” Schafer wrote in a note dated late Tuesday. Arm Holdings
ARM,
-4.10%,
which returned to the public markets last week, uses a different architecture than Intel’s x86, one that favors low power consumption over the high performance that generates a lot of heat and uses a lot of electricity.

The Oppenheimer analyst said that Zinsner “tempered” margin expansion expectations, and estimated Intel’s margins could fall in the low- to mid-40% range, compared with the 60% long-term target.

Schafer remarked that Intel’s work on a 256-node graphics processing unit (GPU) cluster, with each node housing eight of Intel’s Gaudi accelerator chips, was “most likely, a competitive response to Nvidia’s DGX GH200.”

Nvidia unveiled its DGX GH200 Grace Hopper superchip in early August with an expected release date in the second quarter of 2024. Schafer noted that Intel execs see 80% of the DGX target market as using eight or fewer GPUs, and that the company is targeting that market with its Gaudi accelerators for AI inference and training large-language models. Despite Intel’s efforts, however, Schafer said that “Nvidia’s dominance in training LLMs is unlikely to be disrupted.”

Citi Research analyst Christopher Danely held onto his $34 price target and neutral rating on Intel’s stock, with some thoughts on Intel’s habit of talking up Intel Foundry Services, or IFS, its nascent third-party foundry business. The company has yet to name a large customer.

Foundries, or “fabs,” as they are known in industry parlance, are complex, super-clean facilities where transistors are etched onto wafers of silicon. Intel, like Texas Instruments Inc.
TXN,
-0.64%
and Micron, operates its own fabs, while companies like Nvidia, AMD, and Apple Inc.
AAPL,
-2.00%
are “fabless,” meaning they use a third-party fab like Taiwan Semiconductor Manufacturing Co.
TSM,
-1.00%
to produce their chips.

IFS is meant to compete with TSMC and other third-party fabs like those run by Samsung Electronics Co.
005930,
-0.29%
and GlobalFoundries Inc.
GFS,
-0.88%,
and Citi’s Danely said he is not a fan. He wrote that he believes Nvidia is IFS’s “pre-payment customer” for third-party foundry services, but “mostly for leverage against TSMC.”

“We believe Nvidia is mainly leveraging Intel for price with TSMC, and we see a very small chance Nvidia would ever risk leading-edge manufacturing with Intel with an earliest revenue date by 2028,” Danely wrote. “We continue to believe Intel should exit the foundry business.”

Of the 44 analysts who cover Intel, nine had buy ratings, 28 had hold ratings and seven had sell ratings, along with an average target price of $36.91, according to FactSet data.