“We've continued to see a lot of excitement with these large fleets […] The larger you are, the larger your footprint is of your operation, you get the largest benefit from the safety technology, and from the efficiency gains.”

There was a lot to catch up on with Intel on Thursday night. Following a report in July in which CEO Lip-bu Tan gave mixed messages about his willingness to pursue leading-edge chip technology, the U.S. government in August announced it took a ten percent stake in the company, and then Nvidia last month said it would take a four percent stake in the company.

The results and outlook show the same struggle for Intel as they have been for a while now.

The company’s revenue for chips to the data center and artificial intelligence, the key metric, continue to be uneven from quarter to quarter, with this quarter’s sales declining a point from a year earlier. Tan continues to argue for the appeal of “AI PC” to boost personal computer microprocessors, although there’s no sign that AI PCs are a hit with consumers.

Aside from a partnership with Nvidia, Tan has not really outlined how Intel will stage a comeback in AI, where its products are no match for Nvidia’s.

The remarks about the company’s most cutting-edge tech, “14A,” continue to be vague, with Tan saying that “We remain active engaged with potential external customers and are

encouraged by the earlier feedback which help us to drive and inform our decisions.” Waiting for your customers to give you the signal is not necessarily a great place to be with advanced technology.

It’s becoming more and more the case that fans of Tesla look to the future for some kind of deliverance. The present state of the business is underwhelming, and the promise of the “Optimus” robot is a pleasant place to escape.

The third-quarter report Wednesday evening was anti-climactic because the important part had been announced three weeks ago, which were the September-quarter deliveries, which were the highest in company history, but also showed a trend of slowing growth for deliveries. The stock initially traded down at the open, but have gained a couple points Thursday afternoon, at $447.34. 

The best that can be offered by the bulls given the slowing trend is that this quarter, “there is broad agreement volume is set to decline sequentially,” but, “the magnitude of the drop might not be as bad as feared,” as writes Deutsche Bank’s Edison Yu, who has a Buy rating on the stock and a $440 price target.

There was also no forecast by the company for 2026, which leaves nothing to be excited about the rest of this year or next year from a vehicle standpoint.

So much for playing defense. Back in July, at the last earnings call, Netflix was a defensive stock, seen as relatively insulated in a time of tariff-driven uncertainty.

That seems to have run its course. The shares are down by eight percent Wednesday following the company’s report Tuesday evening, which featured the first earnings-per-share shortfall in almost two years, and a revenue number that was roughly in line with consensus but that is actually being counted a miss.

My sense is 2026 is going to be harder for Netflix based on the need to show revenue growth, something hinted at by this report.

One analyst, Brian White of Monness, Crespi, Hardt, who rates the stock Neutral, characterized the report as “a lackluster third quarter” and “an uninspiring fourth-quarter outlook.” Remember that Netflix this year stopped reporting its subscriber numbers, which means that when financial results don’t dazzle, there’s an element of uncertainty that makes it worse.

The week that ended October 17th brought the TL20 group of stocks to consider to a 293% gain since inception on July 15th of 2022. It's also stretched valuation somewhat. It’s been a month and a half since the latest rebalancing of the TL20, the sixth rebalancing, on September 2nd — formally, the reference date is August 29th, the preceding Friday.  

You can see that most of the twenty have seen some expansion of their valuation multiple — they’re more expensive now. I don’t think this expansion is fatal to any of the names, but something to keep in mind as an elevated risk for them. 

Some stocks have exceeded their target as they rose in price and some have fallen below. That’s a problem that I’ll tackle probably at year’s end, and, generally, a good problem to have.

Since the sixth rebalancing of the TL20 at the beginning of September, the group is up eighteen percent through last Friday’s close, triple the performance of the Nasdaq Composite in that time.

The below table shows the key points. The stocks are up an average of fifty-three percent this year, and an average of twenty-one percent since the rebalancing. Stand-out performers include Micron Technology, up seventy percent since the rebalancing; ASML, up thirty-nine percent, and Vertiv Holdings, up thirty-six percent.

In a market awash in artificial intelligence enthusiasm (or, maybe, artificial enthusiasm), consider being a little more choosy.

That’s the advice Monday morning from Barclays’s Tom O’Malley, who shifts ratings on multiple chip stocks following what he calls the “unprecedented wave of compute announcements,” referring to OpenAI’s deals with Nvidia, Advanced Micro Devices, Oracle, and Broadcom over the past month.

O’Malley is “getting more selective,” he writes, to “consolidate risk” to the chip companies. Some stocks are “pricing in much of the full benefit of AI deployment,” he observes, and there is bound to be “some 'selling the news' as the bar is high.”

O’Malley cut his rating on Marvell Technology, Astera Labs, and Lumentum to Equal Weight from Overweight, as the expectations are too high.

“Today, the numbers are insane,” said C.C. Wei, CEO of Taiwan Semiconductor Manufacturing, on Thursday morning’s conference call with analysts following an upbeat third-quarter report. 

The insane numbers Wei was referring to are the surge in artificial intelligence chip plans. As good as the company’s report was — and it was very good — the lingering question from the call is whether Taiwan Semi or anyone else really has a handle on how real those plans are.

Since last we heard from Wei, in mid-July, his biggest customer, Nvidia, inked a hundred-billion-dollar deal with OpenAI, and the latter inked similar deals with two other Taiwan Semi customers, Advanced Micro Devices and Broadcom. There are a lot of promises of future demand, as yet unfunded, suddenly crowding Wei’s list of priorities.

Wei is convinced, or, he sounds convinced, that all the future orders represented by what are now over a trillion dollars worth of commitments by OpenAI are real.

One of my favorite shows on which to be a guest is Simon Erickson’s 7Investing Podcast. 7Investing is his stock-picking subscription service, and he brings together excellent questions in his podcast with a nice gift for keeping the conversation flowing.

In this latest installment, Simon grilled me on all things quantum computing, following my feature on quantum two weeks ago.

First Oracle, then Nvidia, then AMD, then Broadcom, and now, ASML is the latest company basking in the glow of OpenAI’s big plans.

Shares of the chip-equipment giant closed up almost three percent Wednesday at $1,009.81 (for the American Depository Receipts), and helped to lift shares of fellow equipment makers such as KLA, after ASML announced in the morning a third-quarter report dripping with AI enthusiasm.

It was a bounce-back report, if you will, a relief from the sour tone in mid-July, when ASML’s CEO, Christophe Fouquet, had told the Street he was “praying for growth” amidst the uncertainty of tariffs.

Three months later, artificial intelligence is boosting the outlook, although we can all question just how sustainable that is.

Broadcom shares surged almost ten percent on Monday to close at $356.70 after the company became the latest beneficiary of the seemingly endless largesse of OpenAI, which agreed to buy ten gigawatts of AI chips over from Broadcom over a three-year period, similar to its three weeks ago with Nvidia. 

It’s interesting the stock got such a pop as this deal was not entirely a surprise given that The Financial Times reported last month that OpenAI was the unnamed fourth big customer for AI chips to which Broadcom CEO Hock Tan had alluded during the company’s earnings call.

The total commitments by OpenAI now are pretty staggering. The company currently has two gigawatts worth of chips, but its commitments to Broadcom, Nvidia and to Advanced Micro Devices will multiply that more than ten-fold, to twenty-six gigawatts.

“If we run some numbers about this deal it will lead to billions in incremental revenue,” writes Daiwa Securities’ Louis Miscioscia.

The week that ended October 10 ended on a sour note for markets with renewed U.S.-China trade tensions pushing the Nasdaq Composite Index down 2% and the Standard & Poor’s 500 index the same.

What continues to dominate the tech scene is discussion of the astounding slew of deals by OpenAI with just about everyone — Nvidia, Advanced Micro Devices, Oracle, etc. 

I mentioned on last week’s podcast that there was a danger of OpenAI being a kind of structurally weak point in the global economy as they do deals that have been regarded as “circular” — OpenAI gives buys from the vendors and also is in some way financed or linked financially to the vendors, so that money the reasoning for a deal is not really sound.

That subject is taken up by Richard Waters of the Financial Times in a long piece over the weekend, “How OpenAI put itself at the centre of a $1tn network of deals,” in which he quotes various investor types on the risks of the circular deals. One of the facts that is important that comes out of the piece is the potential for systemic financial risk as debt financing is put into data center projects that are propelled by AI.

This is a busy week for those corporate dog & pony shows that companies hold in between earnings season, with Dell and Autodesk both holding their annual briefing for analysts, and OpenAI holding its developer conference in San Francisco, all three of which generated a lot of coverage the past twenty-four hours. 

DELL STILL HAS TO PROVE AN AI PROFIT

With these analyst meetings, it’s all about keeping track of what was said before, at the last meeting, and what’s being said now.

In the case of Dell, the update from management is encouraging, especially as regards profitability, which has been a concern given the company’s rush into selling low-margin artificial intelligence server computers.

Do we need to start referring to OpenAI by the phrase “Too big to fail” like we used for the banks in the Great Recession of 2008?

The last podcast, September 22nd, followed the news Nvidia would invest a hundred billion dollars in OpenAI in order to help the latter to purchase Nvidia chips. 

The latest bubble-seeming incident is Monday’s announcement that OpenAI will purchase perhaps as much as $110 billion worth of GPUs from Nvidia rival Advanced Micro Devices. 

This deal comes with a strange extra clause: OpenAI will take a ten percent stake in AMD. For AMD, it means a huge lift in the company’s revenue potential. 

But now a lot is riding on a company that as The Financial Times points out, is on the hook for twenty-three gigawatts of capacity incoming years, over a trillion dollars in commitments.

Shares of Advanced Micro Devices are the latest to get the “OpenAI Bounce,” as I call it, surging almost thirty percent after the company said Monday morning it has struck a deal for OpenAI to deploy six gigawatts worth of AMD AI chips in return for a “performance-based warrant” for a hundred and sixty million shares of AMD stock, which at Monday’s price of $208.71, is worth about thirty-four billion dollars. 

This is the latest commitment by OpenAI to billions in chip capacity following its deal two weeks ago to buy a hundred billion dollars worth of chips from Nvidia, and a week prior to that, reportedly, with Oracle to use three hundred billion dollars worth of AI capacity in Oracle’s cloud.

This may be one of the most striking AI deals so far, in that the warrant is is “directly tied to increasing AMD stock price milestones, with the final tranche vesting at a price of $600 per share,” said AMD CEO Lisa Su on a conference call this morning. 

OpenAI, in other words, is on the hook to boost AMD’s share price to get value from the warrants it is receiving. 

You’re probably well aware that this time of year, early October, is the season for incessantly following the “lead times” of Apple’s iPhone. That’s the time it takes, on average, for a phone ordered online to be delivered. It’s a rule of thumb the Street uses to gauge how popular the latest iPhones are, in this case, the iPhone 17 models introduced two weeks ago.

Jefferies & Co.’s Edison Lee opines that things are going pretty well, but he still cut his rating on Apple stock on Friday to Underperform from Hold, after concluding that the new design introduced this time, the “Air,” is “the least popular by our tracking,” and that, “this new form factor has not helped AAPL driven iPhone sales.” 

More important, Lee is looking not this year’s crop, but next year, to a rumored “foldable” iPhone, something that Apple watchers such as Bloomberg’s Mark Gurman have been writing about for a long time now. 

It’s expected by some a foldable iPhone may show up next year, and if so, it would be a first from Apple in the foldable category, going head to head with models from Samsung Electronics and Google, among others. 

It would appear there’s great appetite among retail investors for curated stock picking, especially when it comes to artificial intelligence.

The Dan IVES Wedbush AI Revolution ETF, owned by Wedbush Securities, which you can trade under the “IVES” ticker, informed by the equity research of longtime Wedbush analyst Dan Ives, hit three quarters of a billion dollars in value a week ago after starting with just a million dollars four months ago.

Since inception, through Friday morning, the fund is up thirty percent, nicely topping the seventeen percent of the Nasdaq Composite.

“We expected it to be very successful but this is a lot quicker than I anticipated,” says Cullen Rogers, portfolio manager for the IVES. “I think there’s a confluence of events there, which is the general market performance, and AI buzz, but, the retail acceptance of it has been fantastic.”

Enthusiasm for quantum computing comes and goes — this year, it’s way up. Stocks of the most prominent names, IonQ, D-Wave Quantum, and Rigetti Computing, have, on average, doubled in price.

But, the fundamental challenge for all quantum companies hasn’t changed in two decades. They are all trying to get to “scale,” and it’s not clear when, if ever, any of them will.

I’m emphasizing this point because the scaling issue is fundamental to all computing; without scaling, nothing is meaningful.

Scaling means that as a computer gets more complex — basically, bigger — the amount of work the computer can do increases in proportion. The main example is the integrated circuit, which became more and more powerful over six decades, increasing dramatically what could be done, to the point that you now have a supercomputer in your pocket with today’s phones.

Quantum is not there yet. All the companies have shown interesting machines, but none have proven they can scale those machines the way the traditional semiconductor scaled.

What’s the value of a company with no revenue?

Barclays analyst Christine Cho writes Monday that it all depends on the headlines.

Cho initiates coverage this week of Oklo and NuScale, two nuclear power hopefuls, giving Oklo an Overweight rating, and giving NuScale an Equal Weight rating. Her main argument is that the news cycle is what will drive the stock prices.

“Generally,” Cho writes, “we think that the macro news, such as policy or trade updates we get from the Administration (which tend to be more positive than not), and headlines around how the world is short power, will be the largest drivers to stock price reaction while announcements for any binding agreements should also act as a positive catalyst.”

A week ago, we heard about Nvidia’s plans to dump a hundred billion dollars into OpenAI, one of Nvidia’s biggest customers, in return for an enormous commitment to buy data center chips over some unspecified period of time. It seemed to me, and to a lot of others, a bit of “circular” business reminiscent of the DotCom days, and a red flag in that respect.

But the market is undeterred. Nvidia dipped a bit following that report, but is higher Monday afternoon at $182.09 amidst continued confidence in the AI trade.

One undaunted enthusiast is Blayne Curtis of Jefferies & Co., who on Monday reiterates his Buy rating on the stock and raises his price target to $220 from $205.

Curtis offers one of the first increases that I’ve seen in estimates based on the OpenAI deal. 

The SPAC is back!

So-called “special-purpose acquisition companies,” or, SPACs, also known as blank-check companies, were the rage until the 2022 plunge in the market. A blank-check firm issue shares to the public, then uses the money raised in that offering to go on the hunt for some private company to buy, thereby using a back-door process to take companies public that might not have otherwise come to market.

SPACs never went away, but enthusiasm died down a bit. Now, it seems there’s a resurgence of SPACs around artificial intelligence. 

Bloomberg’s Bailey Lipschultz relates Friday that “The SPAC ecosystem is growing again,” noting that “So far this year, nearly 100 blank check firms have pooled $19.3 billion, surpassing the total volume for 2023 and 2024 combined, according to data by SPAC Research.”

The poster child for the SPACs, Chamath Palihapitiya, saw his latest blank check, amusingly named American Exceptionalism Acquisition Corp., listed on NYSE under ticker “AEXA,” make a strong debut Friday morning, rising over eight percent. The offering, underwritten by Santander, raised $300 million to go after “businesses operating in the energy production, artificial intelligence, decentralized finance and defense industries.”

“The number seven [vendor] buying the number two, the dynamics of this are intriguing to see how you make it successful.”

Everything that sounds questionable about the artificial intelligence trade in recent days — the decision by Nvidia to invest $100 billion into OpenAI for example — sounds more more convincing, more realistic coming from Micron Technology CEO Sanjay Mehrotra. 

Tuesday evening, Mehrotra extolled the benefits of AI. “Artificial intelligence is also a powerful productivity driver for Micron,” he said. “We are using AI throughout the company across product design, technology development, manufacturing.”

As one example, “In manufacturing, we have driven a five-times increase in [semiconductor] wafer images analyzed in the past year,“ he said, improving the amount of defect-free chips Micron produces.

Mehrotra has good reason to trumpet AI: it is dramatically lifting the business. He presided overly a quarterly earnings report by that was pitch perfect: The company not only exceeded expectations, it exceeded the already high preview of its results it had offered last month. And demand for both DRAM chips and NAND flash memory are rising faster than previously expected this year.

The week ended September 19th was a good one for tech, the Nasdaq Composite Index rising 2%, the Standard & Poor’s 500 rising 1%, and much of it had to do with the big announcement by Nvidia that it will take a four-percent stake worth $5 billion in Intel, its arch-rival, and that the duo will make custom chips for the data center and for laptop computers based on Intel’s x86 technology.

Intel ended the week up 23%, giving it a return this year of 43%. Nvidia ended the week down fractionally, but its return this year is still very good at 37%.

However, the spending by Nvidia is raising questions. The Intel deal was followed on Monday, September 22nd by Nvidia’s announcement it will make an investment of a hundred billion dollars over some unspecified period of time in OpenAI, makers of ChatGPT, contingent on the latter buying and deploying at least ten gigawatts’ worth of Nvidia chips.

Nvidia is making news as much with its largesse these days as its chips.

Monday, the company announced a plan with ChatGPT creator OpenAI for the latter to deploy “at least” ten gigawatts of Nvidia’s chips, in return for which Nvidia will invest a hundred billion dollars in the startup, 

The news follows Nvidia’s announcement last week it will take a ten-billion-dollar stake in chip rival Intel.

Nvidia shares were up one percent on the news, at $178.97.

To put that in perspective, New York City pulls five gigawatts of power on an average day, according to those in the data center business who track these kinds of mammoth usages. Ergo, the two companies are planning to deploy two very large metropolitan areas’ worth of AI chips.

The deployment of ten gigawatts constitutes “millions of GPUs for OpenAI’s next-generation AI infrastructure,” said the two companies in prepared remarks. 

Intel shares are up twenty-three percent Thursday afternoon after it announced with Nvidia that the latter will take a five-billion-dollar stake in Intel, equivalent to four percent at Wednesday’s close, and the two will develop custom chips for artificial intelligence servers and a new breed of laptops.

The stock bounce gives Intel a fifty-three percent gain for the year, topping Nvidia’s thirty percent to date. Nvidia shares rose four percent, while ARM Holdings, perceived as put out in the cold by this, is down four percent. Advanced Micro Devices, now the odd one out for the x86 world, declined by one percent.

The Street reaction has so far been very upbeat for both companies, although I suspect that the balance of the account is squarely in Nvidia’s favor.

IonQ, one of the three publicly traded “pure plays” in quantum computing, along with Rigetti Computing and D-Wave Quantum, last Friday held its “analyst day” to brief sell-side analysts on its technology roadmap and financial goals. 

Most of the points made by CEO Niccolo de Masi and his team were the same points that de Masi made in my interview with him last month; you could say that our chat was a preview of what was heard last week. 

You can catch the whole day’s presentation in the archived recording on the IonQ investor Web site. You can also download the deck of accompanying slides. 

The talks were very well received, with the shares closing up by eighteen percent on Friday. At a recent $68.27, IonQ shares are up sixty-three percent this year.

At a high level, the company reiterated its outlook for this year’s revenue, $82 million to $100 million. And de Masi again compared IonQ to IBM as the only company that he contends comes even close to competing with IonQ. Analysts so far have not made any dramatic changes to their estimates for IonQ’s financials.

We are not in a bubble of artificial intelligence, but that doesn’t mean things can’t get perilous in the AI trade.

So implies Bernstein analyst Mark Newman in an initiation of coverage on Monday spanning a hundred and twenty-three pages in which Newman reflects on the AI payoff for Apple, Dell and other vendors of IT hardware such as server computers, disk drives and smartphones.

Newman thinks the opportunity for hardware vendors is “massive,” though he also sees the risk there could be a period of “digestion,” which brings to mind thoughts of the DotCom collapse.

“We see sustainable increase in spending as AI expands the addressable market for IT and creates a potentially massive opportunity for IT hardware to return to growth,” writes Newman.

Newman’s report is filled with lots of great detail, reflecting his many years covering the hardware beat. His contribution is to put some serious numbers around the AI trade and where it is going.

Newman has three main points.

We have come to the end of earnings season, with some significant news from two of the last companies to report, Oracle and Adobe, really the first companies of a new earnings season, they just report very early.

Oracle’s stock soared by 36% on Wednesday after its Tuesday-evening report contained the stunning disclosure that it had more than quadrupled its annual backlog of business, its “remaining performance obligation,” or, RPO, promising sharply increasing cloud revenue for the next five years.

Oracle is up 75% for the year through Friday’s close. The pop in Oracle stock gave the newly rebalanced TL20 group of stocks to consider a forty-percentage-point gain since the rebalancing, now up 273% for the year through Friday’s close.

Almost immediately, questions arose about that backlog. 

Adobe shares are rising three percent in early trading at $361.90 following Thursday evening’s fiscal third-quarter earnings report, and, if it holds, this would be the first time in five quarters the stock hasn’t sold off. 

The drama of Adobe is by now well understood. The worry on the part of investors is that the company will see its franchise eroded as more and more average users, people not trained on its Photoshop and Premiere tools, turn to generative artificial intelligence to do their design work.

The stock has been an absolute dog as a result, down forty percent this year and the same amount over a twelve-month period.