Intel Needs To Surprise (NASDAQ:INTC)

Investment Thesis

The general consensus currently seems to be that there is a recession ongoing, which implies a slowdown, which is likely to impact Intel Corporation (NASDAQ:INTC). On the other hand, Taiwan Semiconductor Manufacturing Company Limited (TSM) (“TSMC”), which recently reported its Q2 results already, beat estimates and raised guidance, proving that semiconductors are quite recession-hardened given the, after many decades, still ongoing digitization of everything.

From that view, reporting better-than-expected earnings (expected reporting date: July 28) and guidance would be just one piece of the jigsaw puzzle that Intel needs to solve in order to get out of its declining stock price.

Background

In Q1, Intel reported more-or-less less expected results, although PC weakness (which indeed wasn’t unexpected) was compensated by the data center. Intel guided Q2 below estimates, but reaffirmed its full year outlook, citing stronger than expected data center.

Q2 Preview

Earlier this month, I downgraded Intel to neutral based on its uncertain product execution, while also upgrading TSMC on a deflated valuation despite superior revenue and earnings growth and long-term double-digit (high teens) guidance.

However, since I am long Intel, management may (or may not) have a chance to regain some investor trust when the company releases its Q2 numbers. The best way to do so would be by launching tangible (and highly competitive) products as well as by reporting strong revenue growth.

However, neither of those are likely to occur in the near term, since per Intel’s original schedule Sapphire Rapids had to launch in Q2, which didn’t happen, and given some CFO comments and other indications such as PC shipments, the likelihood for strong revenue growth also is pretty much nonexistent (although that is in line with management expectation to increase revenue growth to double digits only by 2025-2026).

So the next best thing Intel could do would be to provide guidance above analyst estimates, and also to provide (real) indications that Intel’s execution is actually improving, if not this generation (according to the rumors it is not), then the next generation (which as I did acknowledge in my downgrade article is what CEO Pat Gelsinger has actually been focusing on).

Overall, it is understandable why people are cautious about Intel’s earnings and potential guidance: there is simply no segment that could yet make up for the sluggishness of the PC. Still, at this point in Pat Gelsinger’s tenure, the focus should remain on the product side, which hasn’t shown any promising signs either this year, given the lack of a substantial Sapphire Rapids ramp:

In terms of products, Alder Lake has been a star, and it’s ramping comfortably ahead of our expectations there, which has reaffirmed the health of Intel 7. Sapphire Rapids that you called out, it was first peer PRQs this quarter, and many of the additional SKUs, PRQ in Q2 — in the second half of the year, and that’s why you might be getting some of that views of the more muted ramp there. But we delivered on exactly what we said. First quarter PRQs of Sapphire Rapids, and we’ll see strength in that as we go through the rest of the year. (…) This product will be extremely well respected, accepted and broadly deployed in the marketplace this year. (…) And our objective is to ramp this platform meaningfully faster than we did the Ice Lake platform. (…) And as I already indicated, we’re seeing a tremendous amount of SKUs and instance types across all of the OEMs as well as all of the hyperscalers in the marketplace, and we’re looking forward to those being broadly available in the second half of the year.

Arc GPU

It seems fair to say people likely hadn’t expected the Arc rollout to be so slow, as the Arc brand was announced nearly a full year ago already. It seems Intel continues to focus on the low-end, with the A300 line (which has just 128EUs), whereas the A700 series (with up 512EUs) will be “limited edition” according to Intel.

With regards to the A300 series, the first reviews seem to indicate that Intel’s drivers aren’t of optimal quality. Note, by the way, that Meteor Lake (the Intel 4 CPU for 2023) was supposed to have 192EUs (although a recent rumor indicated it might be only 128EUs). This means Meteor Lake’s integrated graphics should be at least on par with the discrete GPU that Intel is currently shoving into the market in order to reach its goal of 4 million shipments in 2022. In other words, this isn’t the GPU that will be challenging Nvidia (and AMD).

So overall, Alchemist is too little, too late. Since it is based on 7nm-class process technology, Intel launched this GPU in late 2020 already, around the same time that Advanced Micro Devices (AMD) and NVIDIA Corporation (NVDA) launched their 7nm-class GPUs.

Intel Foundry Services

The news mill around IFS, which was spinning quite fast up to and including at the investing meeting, has slowed down. Nevertheless, Intel had said it expected to announce a new whale customer in the second half of the year, which would add to Amazon (AMZN), Cisco Systems (CSCO) and QUALCOMM Incorporated (QCOM).

CHIPS Act

At the February investor meeting, Intel said it expected 10% capital offsets, but was aiming for 20-30%. Although ~10% offsets are quite in line with historical handouts (for example the $1B Intel gets from Israel for its upcoming $10B fab), the 20-30% is possible due to the CHIPS acts in the E.U. and U.S. Pat Gelsinger reportedly even went so far as to say that he would build more in the E.U. if the U.S. version didn’t pass.

As Intel’s competitors have observed, this implies that Intel could get several tens of billions of dollars in subsidies.

Mobileye IPO

Reportedly, the mid-2022 Mobileye IPO has been delayed. The obvious likely cause is the general stock market trend. Intel reportedly was seeking a $50B valuation for Mobileye (which could allow Intel to raise perhaps around $5B), but since P/S multiples have contracted by well over 50% since the announcement, upholding this valuation would make Mobileye one of the most expensive stocks out there with a P/S well into the double digits.

Ultimately, the IPO was just financial engineering. On the real engineering side, Mobileye said it is waiting for permits/legislation to start its driverless robotaxi service in Israel and Germany as first locations, with more locations likely to follow in 2023.

Disclosure

Despite my recent downgrade, I recently bought a few more shares, for the first time since I opened my initial position in mid-2020. I was willing to buy even more, but then I saw the stock had already rallied from ~$35 back to ~$40, perhaps due to the CHIPS act news.

For me, the most important point is the progress in process technology (especially compared to TSMC’s lack of progress in the next several years with both N3 and N2 being underwhelming nodes), which in a few years will become the rising tide that will lift all of the Intel business segment boats. Intel’s market cap is also much lower than TSMC.

Investor Takeaway

Given the current near-term product and earnings outlook, Intel would need to surprise in order to deliver anything noteworthy that could possibly push the stock higher. However, I fail to see how or where Intel could deliver such a surprise. Although such is evidently the nature of surprises by definition, most of the surprises in the last so many months have been delays or rumors of delays. Further delays are not what investors expected to happen from Pat Gelsinger’s return to Intel.

Specifically, Pat Gelsinger was very clear in the previous earnings call about the aim to ramp Sapphire Rapids faster than Ice Lake, with broad availability in the second half of the year. As a reminder, Intel shipped 1M units from Q1 to Q3, and then another 1M in Q4 last year as cloud instances became available around early Q4.

Investors should check if Intel backtracks on any of its Sapphire Rapids claims from last quarter. The engineering side continuing to not deliver could start to be seen as a red flag, although the caveat is that the process tech actually remains on schedule. Hence, for now, investors could remain cautiously optimistic that any Sapphire delays won’t spill over to Granite Rapids on Intel 3, which Intel said taped out last quarter.

Lastly, when it comes to surprises, Intel announced in February that it would announce a new marquee foundry win (“whale customer”) in the second half of the year. If Intel really wants to surprise, then it would have to announce one of the big three: AMD, Nvidia, and/or Apple (AAPL). Obviously, none of those would switch their full business to IFS overnight, but the concept that is especially popular in software and software as a service (“SaaS”) applies here as well: land and expand.