The week started out slowly because of the Labor Day holiday — or maybe because a lot of CEOs and VCs were stuck in the mud at Burning Man — but there was plenty of news as the week wore on.
The gist: Arm readies its IPO for next week, antitrust rulings start to bite big tech, and enterprise tech earnings largely come in positive but there’s abundant caution by companies and investors alike.
Next week the drumbeat of tech conferences will continue, led by Salesforce Inc.’s Dreamforce in San Francisco, now possibly the city’s biggest tech conference — unless it loses it next year. Look for a flurry of artificial intelligence announcements, because what else matters?
You can get more analysis of the week’s news and more on this week’s theCUBE Pod podcast with analysts John Furrier and Dave Vellante. And don’t miss Vellante’s weekly Breaking Analysis coming Saturday, this one digging into how how generative AI is both a copilot and a competitor for UiPath.
The year’s big tech IPO almost certainly will happen next week: Arm will seek to raise $4.7B+ in its upcoming IPO A few thoughts on a bellwether IPO:
There’s an unusual pile of challenges facing Arm, but it’s still going to be watched for how investors respond as a sign of whether the IPO window will open wider. Its valuation still seems rich compared with other chipmakers, even given its advantageous approach to the market that enables lower cost and more innovation by farming out manufacturing to an ecosystem of fabrication facilities.
And even if investors are hoping the IPO is something of a proxy for the generative AI opportunity — something Arm likes to emphasize, including a video of Nvidia CEO Jensen Huang in its roadshow presentation, even though his deal to buy Arm was called off last year on regulatory concerns — it’s not clear how much Arm will benefit from the gen AI boom.
“It’s a lot more challenging for Arm to capitalize on the current trend for AI than a company like Nvidia because around 60% to 70% of Arm’s revenues are derived from mobile and the AI landscape is centered around cloud-based operations, rather than being heavily integrated at the device or edge level, where Arm is more prominent,” Albie Amankona, an analyst at global research firm Third Bridge, told me via email.
Dave Vellante, by the way, believes that view is absolutely wrong, and I’m inclined to take his word on this. Indeed, AI inference is increasingly performed at the edge for many reasons such as network latency and data governance issues, and that’s got to favor Arm. Moreover, Arm itself has a neural processing unit chip design and is working with Microsoft and others to implement inference at the edge, not to mention GPU designs.
Arm itself notes that generative AI poses potential risks to certain parts of its business, since AI models generally run on graphics cards, not the central processing units it makes. But the combination of Arm’s CPUs, NPUs and GPUs, not to mention the expense and power demands of Nvidia’s GPUs, could give Nvidia a run for its money at the edge, especially given the need for reduced power consumption on edge devices.
Amankona also thinks the growing adoption of RISC-V chip architecture presents an even bigger risk for Arm. “Our experts say RISC-V is already making inroads in mid and low-end applications, such as the embedded market, including automotive,” he said. “It seems inevitable that Android and Microsoft will eventually embrace RISC-V, as it offers them the opportunity to lower overall costs and maintain or even improve performance.” That said, Arm has a big head start on RISC-V and, of course, that huge volume lead by all the chipmakers in its ecosystem.
Investors also may be concerned that Arm just isn’t growing very fast. “Our experts are skeptical about the long-term sustainability of revenue growth and high margins of Arm,” Amankona said. “They expect a yearly revenue growth of 5% to 10% for the next five years, followed by a peak and subsequent contraction on a yearly basis. Additionally, margins are expected to decrease by a third over five years.”
SoftBank, which will still own the vast majority of Arm shares, isn’t getting nearly the return it would have from its Nvidia deal. But for Arm, raising $5 billion or so should provide a lot of resources to press its advantages, especially since, as a chip designer and not a manufacturer, it’s not capital-intensive. And its strategy of focusing on high-volume production may already have won the long-term battle. We’ll see next week how many investors buy into its views rather than those of skeptics such as Amankona.
Lots of antitrust news
Antitrust hammer could finally land on Amazon: FTC antitrust lawsuit against Amazon will likely come later this month after talks break down
And, from across the pond, on other tech giants too: EU designates six tech giants as gatekeepers under DMA law
Big tech got some wins too this week:
UK realizes it was ruining encryption and backed off: UK updates stance on controversial Online Safety Bill after criticism of encryption proposal And a settlement: Google tentatively settles antitrust lawsuit over Play Store practices
But the big kahuna is still to come: In Its First Monopoly Trial of Modern Internet Era, U.S. Sets Sights on Google I’m not so sure it’s a slam dunk for the DOJ. Is paying companies to make your product the default — not exclusive — enough to create or maintain a monopoly? Maybe, but not if Google search sucked. Apple and others wouldn’t then pay so much for it. Back in the 1990s, Microsoft arguably was even more dominant and used more hard-nosed tactics such as bundling, with little apparent benefit to consumers, while Google’s main product clearly is a benefit for most people. But antitrust is certainly a political animal, so who knows?
In other news
Chip shortages to continue: TSMC says chip packaging shortage is constraining processor supply
Amazon CEO Andy Jassy expands steam, his senior leadership team, to include Dave Brown, Amazon Web Services’ compute and networking head, AI chief Swami Sivasubramanian and Aicha Evans, head of its Zoox robotaxi unit. It’s a clear sign of what Jassy thinks is important — that last one perhaps a surprise given Zoox has been fairly quiet.
Growing slowly while burning cash is not a good combination no matter what the economic environment is: Report: Big-data company Alteryx holding talks about possible sale
Enterprise software and cybersecurity names turn in generally good results, but they’re still cautious on their outlook. That “macro environment” still bites.
Big opportunity in gen AI but big losses to pursue it for now — this could be a pattern we see in other companies in coming months: C3 AI shares sink as plans to invest in generative AI cut profit forecast
Late-breaking Friday: Nvidia debuts new software to boost AI model performance on its high-end chips
Tony Baer provides an exclusive look at his newest research report just out now: Data and analytics interrupt their programs for generative AI
This startup’s looking to create AIs that reason. It’s not entirely clear what that means yet, but it would represent a leap beyond today’s gen AI models: AI startup Imbue closes $200M funding round backed by Nvidia
Of course, a driven car is what injured the victim, but this won’t help convince people driverless cars are ready for the open road: Patient dies after Cruise vehicles allegedly block ambulance in San Francisco
If 120,000 developers are right, get ready for an explosion of AI apps: Modular makes its AI-optimized Mojo programming language generally available
Gen AI continues to get infused into mainstream software:
And into mainstream disinformation campaigns: Microsoft suspects China is now using AI in disinformation campaigns to sway US voters
Another way gen AI companies will make money to pay for all those cloud GPUs: Anthropic announces Claude Pro, a paid subscription for its generative AI chatbot
Some companies are looking to make gen AI less of an energy hog:
One of these days an autonomous trucking company will succeed. Right? Argo AI founders return with new SoftBank-backed autonomous truck startup Stack AV
David Strom digs into two new generative AI-driven cyberattack weapons: It’s the summer of adversarial chatbots. Here’s how to defend against them
Dave Vellante’s editorial interview with a cybersecurity IPO candidate: Cybersecurity trends, market expansion and IPO outlook: Arctic Wolf’s CEO weighs in
Onetime hacker and former Twitter security chief Mudge lands at CISA: Well-known security consultant ‘Mudge’ is once again on the move
And consolidation continues: Tenable to acquire cloud security startup Ermetic for $240 million
Elsewhere around the tech world
Instead of going big, a former longtime Amazonian is going home after getting summarily shown the door: Flexport CEO Dave Clark steps down after just one year in charge Returning founder and CEO Ryan Petersen makes it pretty clear the growth push got out of hand in his view.
Another Cube Conversation with Vellante: Automating financial crime detection: Cable Tech bridges the gap between traditional finance and cryptocurrency
Elon goes off the deep end again: Elon Musk says he might sue the Anti-Defamation League for … defamation
And social media in general is still a mess, as Strom finds. Could be an even more scary 2024 U.S. election: How Kremlin-backed social media campaigns continue to spread disinformation
Apple’s Wanderlust event, Sept. 12
Salesforce’s Dreamforce conference in San Francisco Sept. 12-14: I’ll be there and we’ll have all the news.
AI Hardware & Edge AI Summit in Santa Clara Sept. 12-14: We’ll see if I can clone myself, could be an interesting conference
Earnings next week: Oracle on Sept. 11 and Adobe on Sept. 14
Image: Bing Image Creator
Your vote of support is important to us and it helps us keep the content FREE.
One-click below supports our mission to provide free, deep and relevant content.
Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger and many more luminaries and experts.