I’ve never used an Oura ring, the electronic health and wellness finger band, but I’ve met people who swear by it. It has lots of sensors on the inner edge of the ring that monitor things such as how you’ve slept and your heart rate. Some customers featured by the company testify to the ring having saved their lives by, for example, sending them to a doctor when low heart rate was detected.

Now, Oura, a thirteen-year-old company originally founded in Finland and now based in San Francisco, is going public, aiming to raise half a billion dollars. Given the devotion I have heard people express, I was surprised to learn in its prospectus for the IPO that the “retention,” the number of those who pay $5.99 for a membership, versus those who leave, averaged over twelve months, is 87%. (Membership lets you get analysis of the data collected by your ring to form a larger health profile of yourself. The company says, vaguely, it uses AI to shape analysis of the data.)

An 87% retention rate means that 13% of people who sign up to pay $5.99 for membership churn away over twelve months. That’s a fairly high rate of churn for a subscription business. Most software-as-a-services businesses have a retention rate on the order of 97% or so; even The Technology Letter manages about 97%. That low rate of retention includes cases where Oura is able to get someone to come back to the product after they quit, known as a “winback.” The real rate of churn is presumably therefore something higher.

I dropped shares of Everpure (formerly Pure Storage) from the TL20 group of stocks to consider in March, and what a mistake: the stock is up 110% since then, better than fourteen of the twenty names, including a jump Thursday of 18% to $129.92, following an upbeat “analyst day” presentation to the analysts covering the company.

How could I have been so wrong?

When I rebalanced in March, the stock at the time of that rebalancing was down 8% for the year, and it had merely tracked the Nasdaq in the preceding twelve months. I concluded that the stock couldn’t work because investors would only appreciate Everpure if they won a second deal with a cloud giant on top of their existing deal with Meta Platforms.

Shares of “connectivity” specialist Amphenol are down half a point Wednesday at $82.44 despite a favorable initiation of coverage by Bernstein analyst Varun Govindaraj, who assigned an Outperform rating and a $99 price target. Amphenol is one of the TL20 group of stocks to consider, having been added with the year-ago rebalancing. The stock is up 51% since then.

Govindaraj has two main points: the stock is expensive but justifiably so, and, two, the company will be just fine even though investors are worried about whether its electrical products will be made obsolete by fiber-optic-based connections.

I think he makes a decent case on both counts.

Elsewhere in data centers, UBS analyst Karl Keirstead Wednesday makes a vigorous defense of CoreWeave, the neocloud data center operator, assigning a Buy rating and a $120 price target, which would be a pretty decent return from a recent price of $87.73.

The past twenty-four hours have seen a steady stream of intense interest in a new app from Meta Platforms, called “Muse,” with many predict epic payoff for the entire technology industry.

Muse was first unveiled two weeks ago, and, at the time, I noted that the early response from stock analysts was very favorable, and that the expectation was that Muse is among a bevy of “personal AI” things that will become a solid trend in tech going forward. These kinds of apps will take on tasks for a user such as navigating to a Web site automatically and placing an order. (Think of personal shoppers, etc.)

The fervor has now reached escape velocity since Muse was identified in the past week as the most popular free download from Apple’s App Store.

WHAT IT MEANS FOR META

First priority is assessing the impact for Meta, and the enthusiasm among Meta stock analysts is growing by the day.

AMSC’s chairman and CEO Daniel Patrick McGhan once again sits down with me to talk about the business of supplying semiconductor factories and mines and utilities around the world. Trillions of dollars are being spent on energy and upgrading the electrical grid and on moving supply of critical materials from China to the Western hemisphere. The opportunity for the company is selling larger and larger orders, often by showing up when existing customers need more from AMSC for a semiconductor fab, say. Having reached $90 million per quarter in revenue and two years of continued profit, McGhan says “the company is now really designed and built” to sustain growth and profitability.

A week ago, I met with management of one of the best-performing software stocks this year, Backblaze, whose shares are up 182% at Friday’s close of $13.16.

The performance is in no small part because in late June, Backblaze won the equivalent of the AI lottery: a multi-year deal to supply AI hosting firm CoreWeave worth $335 million.

Backblaze is getting paid by CoreWeave to manage the hard disk drives that store customer data in the cloud. Nineteen-year-old Backblaze has spent years refining ways to arrange files across multiple disk drives. It’s intellectual property makes movement of data fast and efficient, making best use of drives and networks. That has made Backblaze newly relevant in an age of AI that needs to store a lot more stuff.

In New York, I met with Backblaze co-founder and CEO Gleb Budman and CFO Marc Suidan following the company’s first-ever “analyst day” presentation. The analyst day is an event where company management pitches the stock analysts on the company’s strategy and how it will pay off financially.

Probably, upside from here hinges on satisfying a couple of concerns. Can a relatively small company, $950 million in market cap, continue to serve giants such as CoreWeave without being dis-intermediated in an area, storage, that is traditionally a commodity? And, what about all the concerns of the AI euphoria going away?

We are now past most of corporate earning season for technology companies, which means, it’s time for … analyst day season!

The analyst day meetings, which usually take up an entire day, usually at a hotel, with mini-muffins and abundant coffee, are for management of tech companies to regale stock analysts with rosy projections.

I call these “Dog & Pony” shows because the companies’ management preen and try to be pretty and get the award for most promising.

We had a bunch of these this week, and I thought I’d give you the highlights. The big picture is that every company is talking about how the raging AI trade is going to mean bright things for the business for years to come.

CIENA’S BRIGHT PROSPECTS

First, the good news. Ciena, the dominant vendor of fiber-optic gear, continues to dazzle. The company’s stock was already very solid coming into Wednesday’s presentation at its analyst day event, held in Ottawa, Canada, where Ciena has its R&D facility. It gained a couple points on the event to close at $344.25 on Thursday, a 47% gain this year.

Micron Technology and SanDisk, and competitors Samsung Electronics and SK Hynix, have all done superbly well this year as supply of DRAM and NAND flash chips have failed to keep up with raging demand, driving up prices of the chips dramatically.

Can demand keep going higher? The companies themselves refuse to tell us. Management at Micron and the others are no longer forecasting total demand in their market because they can’t even meet current demand.

Into the breach, analysts at Citigroup this week took a stab at measuring supply and demand.

Chip analyst Peter Lee and team conclude — surprise! — demand is headed much, much higher.

The team forecast, for example, that the total number of bits of “high-bandwidth memory,” or, HBM, the fastest DRAM, used in AI, that the world will need is going to surge by 62% in 2027 versus this year’s level of demand, and another 69% in 2028.

Folks, I wish I could say that a highly intelligent, well-coordinated effort to regulate artificial intelligence in the best interests of society is about to take place. I suspect that is not the case, however.

There is no leadership in Washington, D.C., and the companies that are calling for a slowdown of some kind, including Anthropic (CEO Dario Amodei’s blog post) are really so full of self-congratulation that they will not pursue anything on their own.

‘EDGE’ CHIP MARKET COMES INTO FOCUS

Speaking of AI chips, four years ago, I was writing for ZDNet about the market for so-called edge computing, running artificial intelligence on battery-powered devices. That was a very hot market for venture-backed startups until the market imploded. Most of those companies failed to make progress and went away or were acquired. Assets of Flex Logix, one of many startups I interviewed, which made a form of “field-programmable gate array,” sold its key assets to Analog Devices, one of the TL20 stocks to consider, in 2024. Hailo Technologies was bought by Microchip Technologies in July. Untether AI of Toronto, Canada, filed for bankruptcy in October of last year after Advanced Micro Devices hired some of its key people.

(0:05) We’ve pretty much come to the end of earnings season, and I’ve spent the last week tallying the software names that stand out.

(1:21) From an initial list of 460 companies I’ve narrowed it down to almost 60 that saw a “pop” following their reports and are therefore worth considering.

(3:11) The point of the list is to divine which ones could be winners, and there are two basic groups one can identify in software names, “infrastructure” and cyber-security.

(3:51) Infrastructure includes names such as Snowflake, one of the TL20 group of stocks to consider. Its shares jumped 17% following its September 2nd earnings report. After selling off 8% since then, the stock is still up 50% for the year. Snowflake is a good example of an infrastructure winner.

Oracle had fallen 24% since its June report, through Thursday’s close of trading, but the stock is up 6% Friday morning at $161.75, a reversal of the sell-off that was seen back in June. The key factor is that profit margin hasn’t eroded as much as expected, and the company didn’t raise its forecast for spending or for how much debt it will raise, even as the business of hosting GPU chips continues to grow fast.

No news is good news, in other words!

Back in June, the company’s CFO, Hilary Maxson, had told analysts to brace for a decline in gross profit margin. While it was lower this quarter, at 61%, it didn’t fall below 60%, as some expected.

She even had good things to say about margins moving forward, which is that as some of Oracle’s cloud contracts - most of which is now renting GPUs to people who want AI computing — come up for renewal, a lot of that can come in at a higher price, actually helping margins.

The unveiling in April by AI giant Anthropic of its “Mythos” software program, and its companion cyber-security initiative, “Project Glasswing,” was a landmark for the cyber-security industry. At the time, stocks of traditional security vendors swooned. Investors assumed that by automatically detecting cyber flaws, Mythos would make traditional cyber-security firms obsolete.

The opposite has happened: it has made cyber-security firms more relevant than ever.

“This is the hottest market,” says Jay Chaudhry, chief executive of cyber-security vendor Zscaler, in an interview he and I had last week following the company’s fourth fiscal quarter financial report on September 3rd.

“Mythos was probably the biggest event I have seen in cyber-security in a long time,” says Chaudhry. “Every board meeting, and at the level of finance ministers” in countries around the world, “they were talking about it,” trying to gauge the consequences.

Shares of Apple ended fractionally lower on Wednesday at $315.34 as the company formally unveiled the long-awaited “foldable” iPhone with a starting price of $2,000, the iPhone Duo.

The limp reaction of the stock is in contrast to the fairly positive reaction of analysts covering the event.

The most important element for the Apple bulls was the price, which was lower than the $2,100 starting price that bears such as Jefferies’s Edison Lee hypothesized last month.

The fully-loaded model, I should note, with two terabytes of storage, is in fact as expensive as Lee predicted, $3,100. $3,100 for a phone!!

At any rate, a lower entry price is perceived as a a good thing because it may lift the volume of sales of the device. Bear in mind that Apple offers two different payment options to ease the price. The standard financing for the $3,100 model comes out to $90 per month for three years while the recently introduced “Upgrade” plan for Apple devices comes out to $93 per month for two years.

OpenAI’s announcement this week of its token spend looks a little like science, which is the impression that’s intended, but it’s real yet another self-congratulatory press release. At least it’s good news for the chip makers such as Nvidia.

Software stocks are making a comeback, lifted by the artificial intelligence trade. We’ve seen that in two places in particular: stocks of companies selling “infrastructure” software, meaning, tools to build things, and cyber-security stocks. 

Wednesday, I met with a company doing both. Software maker JFrog is known principally as a tool for programmers to manage what’s called a “binary” file. When a programmer is done writing code, they push the compile button, and their instructions are converted into a machine-readable file, a collection of ones and zeros, hence, binary. The binary is what the microprocessor executes to run the program. 

JFrog keeps those binary files organized and protects them from mischief and corruption as programmers are working.

In February, JFrog, along with many software names, looked to investors like a casualty of multiple new software offerings from Anthropic and OpenAI. A consensus took shape that traditional tools would be rendered obsolete as the AI giants just re-created it all with “agents.” Stocks swooned.

What a difference six months has made. JFrog stock is up 113% since then at a recent $87.60.

We got another ample serving of software firms reporting Thursday evening, and I’m delighted to see that Samsara, a name that was in the TL20 up until this past March, has gotten a strong response to its healthy report, up almost 14% in pre-market trading at $44. Price targets are rising to as high as $55 this morning. ‍ ‍

Samsara makes software and hardware sensors that allows fleets of vehicles to be remotely directed (more here.)

However, Samsara has mostly gotten favorable responses in past to earnings only to see its stock slip later on, as I have noted before. The shares have been trailing the market this year, up just 8% prior to last night’s report.

The cloud computing giants, Amazon, Microsoft, and Alphabet, have been joined by a second-tier cohort of companies that are chasing opportunity where the giants leave crumbs on the table.

Some of those crumbs are now sizable. By serving up AI capabilities, for example, CoreWeave, the best known of the so-called neoclouds, has amassed a backlog of $104 billion, it said last month. Its larger competitor, Nebius Group, has $40 billion of backlog. These have overnight become substantial operations.

And some think they will take an even more prominent role.

“We would like to go win more neoclouds,” says Rajiv Ramaswami, CEO of software maker Nutanix, in a chat we had last week. “I think they are a great alternative to hyperscalers in terms of making GPU capacity available to the enterprise easily.”

It’s another funny evening of earnings Wednesday similar to Tuesday night, with even decent reports being sold off. That includes Broadcom, down 1%, and Hewlett Packard Enterprise, which is benefitting from artificial intelligence just like Dell Technologies but not seeing the same stock gain tonight.

The star of the evening, I’m happy to report, is TL20 stock to consider Snowflake, its shares up 23% in late trading at $376.60, adding to a 39% year-to-date gain prior to the report.

BROADCOM’S CEO HELPS THE NEEDY

The hyperbole in AI is really something else these days.

Back in June, there were two issues of concern for Broadcom. One was pressure on gross profit margin, the other was CEO Hock Tan’s soliciting hedge funds Apollo and Blackstone to finance his customers Anthropic and OpenAI.

UPDATE: Palo Alto Networks’s after-hours gain Tuesday night has melted away, the stock down 2.5% at $353.07 in early trading. That’s in spite of analysts raising their price targets as high as $450. Catharine Trebnick of Rosenblatt Securities, who has a Buy rating on the stock, attributes the sell-off to worries about gross profit margin. She notes that the company’s “Cortex” product, which is the root of its security management practice, and a showcase for Palo Alto’s AI efforts, has a lower gross profit margin than the traditional application firewall business at Palo Alto.

Bulls advise staying the course. Remember, Palo Alto had nearly tripled this year heading into the report.

Shares of Dell Technologies have given up a little of their gains overnight, but the stock is still up 8% in early trading at $460.55. Price targets on Dell this morning are going as high as $650 at Bernstein, which would be a 41% gain from here.

The really big gainer of the evening turns out to be GitLab, whose shares are now up 22% in early trading at $55, with price targets zooming to as high as $70 at Canaccord Genuity.

We continue to be in earnings season proper, and this week features an all-star cast of winners this year, the average return to date being a 42% gain. See the table below for the full lineup.

Dell reports this evening, up an amazing 248% this year as the AI data center race has put major fuel into its server sales, and the profit picture for those servers, despite lower profit margin than conventional servers, has turned out much better than many people thought, myself included. 

Dell competitor Hewlett Packard Enterprise, also riding the wave of AI servers in a way similar to Dell, reports Wednesday evening. Its shares are up 114% this year.

Also Tuesday night, Palo Alto Networks, the biggest of the cyber-security vendors at $311 billion in market cap, will report amidst a very healthy climate for security names, as we saw just last week from reports by peers CrowdStrike and Okta. Palo Alto is up a perfectly respectable 96% this year. 

It’s a slow Monday — for me, at least — and, so, I’m reaching back into some of last week’s high-level commentary, some of which is useful for assessing stock prospects.

RENAISSANCE OF MEMORY CHIPS

I’ve been a fan of Micron Technology for a while now, based on the contention that the technology of memory chips, including both DRAM and NAND flash, are essential to processing artificial intelligence. I think that view is becoming conventional wisdom more and more.

However, I think most investors don’t realize just how fervent, how incredibly active this market for memory is, a veritable renaissance of the field.

There’s just so much going on in the memory-chip market between Micron, its nearest competitors, Samsung Electronics, SK Hynix, Japan’s Kioxia Holdings, Sandisk, and also whatever Nvidia and its partners Taiwan Semi and Intel are up to on any given day.

In a thirty-four-page report last week, Bernstein analyst Mark Li goes into some of the very technical details of all the different kinds of memory chips, including something that Nvidia is pushing called “storage next,” which is a sort of hybrid of the best qualities of DRAM and of NAND.

Years ago, Nvidia’s co-founder and CEO outlined what I would call “The Total System,” where computers talk only to computers, with no humans in the loop. It’s one of the reasons I believe he’s looking beyond OpenAI and Anthropic and other AI leaders to the enduring value of chips.

In addition to Nvidia’s stellar report Wednesday evening and Salesforce’s big announcement with Anthropic, covered this morning, we continue to have several software names reporting Thursday night, and it continues to be a return to favor for the software group.

The results are in line with my general observation in this week’s podcast, which is that “infrastructure” software, including tools to build and manage applications, and tools for cyber-security, continue to be the best opportunity in software.

Also this evening, shares of chip maker Marvell Technology, which competes with Broadcom, are down 6% in late trading. I think that’s essentially profit-taking. The report and outlook were strong, but the shares had been up 184% since the start of the year.

Wednesday evening’s earnings conference call with Nvidia, CFO Colette Kress and CEO Jensen Huang was a little bit like the May, 2023 report that stunned the world.

That report three years ago gave investors the first sense that artificial intelligence was becoming an arms race that would hugely reward Nvidia.

Wednesday night, CFO Kress forecast the company’s sales to grow seventy percent in 2027 (actually the fiscal year that ends in January of 2028.) Kress never makes annual forecasts, only quarterly, so, the analysts on the call were surprised.

Price targets Thursday morning are soaring to as high as $515 from Raymond James’s Simon Leopold. “Will Nvidia be a greater than $1 trillion and $25 per share company in 2029?” Pierre Ferragu asks.

And the hyperbole is flying in these reports. Just the titles: “Remember the titan” (Bernstein); “A second Nvidia moment” (Rosenblatt); “A year of visibility” (Wedbush); “The Emperor Strikes Back” (Jefferies.)

(0:10) My recurring theme is a “renaissance” in semiconductors, and, I'm gonna say it once again: it's all about the chips — still.

(0:40) Consider that one of the best performing things this year is the Philadelphia Semiconductor Index, the “SOX,” the benchmark index of chip stocks, up 63% this year.

(0:50) The chip-heavy TL20 group of stocks to consider, curated by yours truly, is only up 53% through Monday, though I’m not complaining.

(1:24) An important recent example of the chips: I took a trip last week to Silicon Valley to see the newest offering from Cerebras Systems, which competes with Nvidia in AI chips. 

(3:01) The Cerebras “Wafer-Scale Engine” is a chip the size of an entire semiconductor wafer. They also build entire computer systems around that chip.

After all that’s happened with Nvidia in ninety days — offering its own balance sheet to back OpenAI’s dreams; inducing financial types to “securitize” the sale of AI chips; SpaceX saying that it will commit to Nvidia as its sole AI chip supplier;   continued raging demand for all chips; and CEO Jensen Huang touting his favorite stocks such as Marvell — this Wednesday evening brings another earnings installment from the company.

The question, once again, with Nvidia, is whether even very favorable results can do anything for a stock stuck in neutral.

Nvidia shares closed down three points Monday, at $208.48, now declining seven days in a row. The stock has gained just under 13% for the year, basically tracking the market despite continued positive reports from the company. 

Elon Musk has achieved a reputation as making the impossible come true. However, the larger his ambition, the more his promises seem to be pushed far out into the future.

There are no “Optimus” robots selling to customers, and the world is not filled with Tesla robo-taxies, despite high expectations for both.

Therefore, it’s reasonable to cast a skeptical eye on the latest outsized promise from Musk, the contention his Starlink broadband service inside of SpaceX will become the U.S.’s fourth largest cell phone operator next year.

To date, Starlink is a constellation of “low-earth-orbit” satellites that deliver either broadband connections to a customer premise, or a mobile service that fills in dead spots around the world — specifically, 7.4 million mobile users in thirty countries as of the end of June.

The question mark is Musk’s ambition to go way beyond that nascent mobile service.

I’ve been on the road a few days, and have finally had a chance to settle back into the corporate earnings groove. Thankfully, this is a less-intense week of reports. (See the chart down below.)

Two TL20 names, Analog Devices and Keysight, delivered excellent reports but their stocks fizzled. This is the equivalent of a Broadway show or a box office feature that gets rave reviews from the critics but fails to catch on with the general audience. However, both stocks have performed very well this year, so, it’s perhaps not a surprise that they didn’t get a pop their numbers.

Alibaba, the Chinese e-commerce giant, is an important read on the AI trade because Alibaba’s cloud computing division is such a large and fast-growing part of the business and it includes many offerings that are linked to AI. Chinese AI models have been increasingly competitive with the top U.S. offerings for several years now.

The company reported June-quarter results Thursday morning.

Monday and Tuesday, I was in San Francisco attending the unveiling of a new AI computer by Cerebras Systems, a company that has fascinated me for almost a decade.

I had a chance to meet with CEO and co-founder Andrew Feldman and his team on Monday in a small group press session, and Tuesday was the general reception for media, customers, partners and others. I heard there were three thousand RSVPs for the festivities, which took place at The Midway, a concert venue on the San Francisco Bay waterfront. At Tuesday’s event, there was a line around the block to get in hours before the start, which is a nice indicator of the momentum the company has.

The big challenge for Cerebras for its next computer was to maintain its edge in speed against The Leading Brand, Nvidia’s GPUs. It seems to me they accomplished that, and also set themselves up for meaningful improvement in profitability by boosting the economics of their products.

Things keep getting weirder in the artificial intelligence market. 

Following Nvidia’s announcement last week that they are partnering with BlackRock and other financial firms to securitize AI chip rentals, the company on Monday announced it will underwrite a portion of a massive data center in Ohio that is to be used by OpenAI.

This is a rather astounding push by Nvidia to leverage its balance sheet — over $120 billion in cash and equivalents — to both drive chip sales and to create more distance between itself and competitors. Is it worth gambling with investors’ money to encourage future chip sales? I have my doubts.

According to Nvidia, the main obstacle to AI’s growth these days is “land, power and shell,” or, LPS, as CEO Jensen Huang refers to it in his blog post accompanying the announcement. (The shell is the physical structure of a data center, the building itself.) Huang aims to be the one to resolve that impasse.