Google CEO Sundar Pichai talks about the company’s third-generation artificial intelligence chips.
Source: YouTube screenshot
Not content with relying on standard chips that are in high demand, some of the world’s biggest tech firms are developing their own semiconductors.
Apple, Amazon, Facebook, Tesla and Baidu are all shunning established chip firms and bringing certain aspects of chip development in-house, according to company announcements and media reports.
“Increasingly, these companies want custom-made chips fitting their applications’ specific requirements rather than use the same generic chips as their competitors,” Syed Alam, global semiconductor lead at Accenture, told CNBC.
“This gives them more control over the integration of software and hardware while differentiating them from their competition,” Alam added.
Russ Shaw, a former non-executive director at U.K.-based Dialog Semiconductor, told CNBC that custom-designed chips can perform better and work out cheaper.
“These specifically designed chips can help to reduce energy consumption for devices and products from the specific tech company, whether it relates to smartphones or cloud services,” Shaw said.
The ongoing global chip shortage is another reason why big tech firms are thinking twice about where they get their chips from, Glenn O’Donnell, research director at analyst firm Forrester, told CNBC. “The pandemic threw a big wrench in these supply chains, which accelerated efforts to do their own chips.”
“Many already felt limited in their innovation pace being locked into chipmaker timelines,” O’Donnell said.
A.I. chips and more
At present, barely a month goes by without a Big Tech company announcing a new chip project.
Perhaps the most notable example came in November 2020 when Apple announced it was moving away from Intel’s x86 architecture to make its own M1 processor, which now sits in its new iMacs and iPads.
More recently, Tesla announced that it is building a “Dojo” chip to train artificial intelligence networks in data centers. The automaker in 2019 started producing cars with its custom AI chips that help on-board software make decisions in response to what’s happening on the road.
Baidu last month launched an AI chip that’s designed to help devices process huge amounts of data and boost computing power. Baidu said the “Kunlun 2” chip can be used in areas such as autonomous driving and that it has entered mass production.
Some of the tech giants have chosen to keep certain semiconductor projects under wraps.
Google is reportedly edging closer to rolling out its own central processing units, or CPUs, for its Chromebook laptops. The search giant plans to use its CPUs in Chromebooks and tablets that run on the company’s Chrome operating system from around 2023, according to a report from Nikkei Asia on Sep. 1. Google did not immediately respond to a CNBC request for comment.
Amazon, which operates the world’s largest cloud service, is developing its own networking chip to power hardware switches that move data around networks. If it works, it would reduce Amazon’s reliance on Broadcom. Amazon, which already designs a number of other chips, did not immediately respond to a CNBC request for comment.
Facebook’s chief AI scientist told Bloomberg in 2019 that the company is working on a new class of semiconductor that would work “very differently” than most of the existing designs. Facebook did not immediately respond to a CNBC request for comment.
Designing but not manufacturing
At this stage, none of the tech giants are looking to do all the chip development themselves.
“It is all about the design and performance of the chip,” Shaw said. “At this stage, it is not about the manufacturing and foundries, which is very costly.”
Setting up an advanced chip factory, or foundry, like TSMC‘s in Taiwan, costs around $10 billion and takes several years.
“Even Google and Apple are reticent to build these,” O’Donnell said. “They’ll go to TSMC or even Intel to build their chips.”
O’Donnell said there’s a shortage of people in Silicon Valley with the skills required to design high end-processors. “Silicon Valley put so much emphasis on software over the past few decades that hardware engineering was seen as a bit of an anachronism,” he said.
“It became ‘uncool’ to do hardware,” O’Donnell said. “Despite its name, Silicon Valley now employs relatively few real silicon engineers.”
The slump in stocks can partly be traced to a turnaround in sentiment regarding artificial intelligence. Tech behemoths such as Nvidia, Broadcom and Oracle slumped, with the last losing more than one-third in value since it rocketed 36% in September.
Investors, it seems, are growing worried over the high valuations of tech names, as well as the gigantic amount of capital expenditure they are committing to — with some, like Oracle, having to take on debt to fulfil those obligations.
Uncertainty over an interest rate cut in December is also putting a downer on Wall Street. It’s a coin toss as to whether the U.S. Federal Reserve will ease monetary policy then, according to the CME FedWatch tool. That’s a huge difference from a month ago, when traders were pricing in a 95.5% chance of a December cut.
Not having October’s employment and inflation numbers, and possibly never getting them, means the Fed lacks visibility into the state of the economy — and whether it should try to support the labor market or continue reining in inflation.
After all, flying blind makes it hard to see where you’ll land. As of now, that applies both to the Fed and investors trying to navigate the still-hazy ambitions of tech companies.
What you need to know today
And finally…
Tan Su Shan, chief executive officer of DBS Group Holdings Ltd., speaking at the Singapore Fintech Festival in Singapore, on Nov. 12, 2025.
“The proliferation of generative AI has been transformative for us,” DBS CEO Tan Su Shan told CNBC on the sidelines of Singapore Fintech Week. She adding that the company was experiencing a “snowballing effect” of benefits thanks to machine learning.
Tan expects AI adoption to bring DBS an overall revenue bump of more than 1 billion Singapore dollars (about $768 million) this year, compared to SG$750 million in 2024. That assessment is based on about 370 AI use cases powered by over 1,500 models throughout its business.
LISBON, Portugal — Top tech executives told CNBC they’re concerned about a bubble forming in the artificial intelligence sector, underscoring growing unease within the industry over soaring valuation.
In recent weeks, markets have been reckoning with the notion that too much capital is pouring into the AI boom, clouding the outlook on revenue and actual profit and putting high valuations into question.
Up to now, warnings around overstretched valuations have mostly come from investors and leaders in the world of finance. Goldman Sachs’ David Solomon and Morgan Stanley’s Ted Pick have warned of potential corrections as valuations of some major tech firms reached historic highs.
The concerns have been crystallized by famed ‘Big Short’ investor Michael Burry, who this week accused major AI infrastructure and cloud providers, or ‘hyperscalers’ of understating depreciation expenses on chips. Burry warned that profits at the likes of Oracle and Meta may be vastly overstated. He recently disclosed put options that bet against Nvidia and Palantir.
However, CEOs of companies who are themselves developing AI, expressed their concerns this week during interviews with CNBC at the Web Summit tech conference in Lisbon.
“I think the evaluations are pretty exaggerated here and there, and I think there is signs of a bubble on the horizon,” Jarek Kutylowski, CEO of German AI firm DeepL, told CNBC on Tuesday.
The sentiment was echoed by Picsart CEO Hovhannes Avoyan.
“We see lots of AI companies raising … tremendous valuations … without any revenue,” Avoyan told CNBC on Tuesday, adding that it is a “concern.”
The market values smaller startups with “just some noise and vibe revenue,” he said, referring to companies being backed even though they have minimal sales.
Vibe revenue is a play on “vibe coding,” a term that refers to using AI to code without needing deep technical expertise.
AI demand growing
Even with concerns over valuations, the technology industry remains bullish on the long term potential of AI.
Lyft CEO David Risher said there are reasons to be optimistic given the potential impact of AI but acknowledged the risks.
“Let’s be clear, we are absolutely in a financial bubble. There is no question, right? Because this is incredible, transformational technology. No one wants to be left behind.”
Risher went on to argue that there is a difference between the financial bubble and the industrial outlook.
“The data centers and all the model creation, all of that is going to have a long, long life, because it’s transformational. It makes people’s lives easier. It makes people’s lives better… On the other hand, you know, the financial side, it’s a little risky right now.”
The tech CEOs also addressed their outlook on AI demand for 2026 from businesses, as investors look for any clues as to what this will look like.
“I think there’s a lot of demand, and there’s a lot of interest. I think everybody understands that AI can do magical things to businesses, and… we can all operate on another level when it comes to efficiency,” Kutylowski said.
Still, businesses are “strugging in adopting” AI. “We’re going to get further, but I don’t think we’re that we’re going to be in a place where we can say, like every enterprise, every organization, has it figured out totally,” Kutylowski said.
DeepL’s core product is an AI translation tool but it recently launched a more general purpose “agent” designed to be able to carry out tasks on behalf of employees.
Francois Chadwick, the chief financial officer of Cohere, a company that is also focused on enterprise AI, told CNBC on Tuesday that “demand is definitely there.”
$4 trillion capex outlook
Despite the concerns over overstretched valuations and huge capex spend, the investment into artificial intelligence doesn’t appear to be slowing down. A report from venture capital group Accel released this week showed that the buildout of new AI data center capacity is forecast to reach 117 gigawatts by 2030 which translates into about $4 trillion worth of capital expenditure over the next 5 years.
About $3.1 trillion worth of revenue is required to pay back that capex, according to the Accel report.
Already this year, there have been a slew of deals worth billions announced by the likes of Nvidia and OpenAI as they look to develop data center capacity around the world in a bid to keep up with demand.
Philippe Botteri, a partner at Accel, said that three major factors will drive that revenue — more powerful AI models that require capacity to be trained, the use of new AI services and the “agentic revolution in the enterprise.”
“Agentic” is often a term used to describe a type of AI tool that can automatically carry out tasks on behalf of users.
But not everyone believes that the large amount of spending is necessary.
Ben Harburg, managing partner at Novo Capital says the figures being discussed by large tech firms for future investment may be overblown.
“We hear these crazy headline numbers about how much energy is going to be needed, how many chips are going to be needed, although, again, I think that there is probably more of a bubble brewing there than on kind of the front end, the actual product front,” Harburg told CNBC on Tuesday.
“I think we’re starting to realize that there’s been probably over exuberance around data centers. Even Sam [Altman], I think, would privately admit that they need fewer chips than they originally set out, they need less capital than they originally set out. They need less energy than they originally set out.”
Tan Su Shan, chief executive officer of DBS Group Holdings Ltd., speaking at the Singapore Fintech Festival in Singapore, on Nov. 12, 2025.
Bloomberg | Bloomberg | Getty Images
SINGAPORE – Amid fears of an artificial intelligence bubble, much has been made of recent reports suggesting that AI has yet to generate returns for companies investing billions into adopting the tech.
But that’s not what the chief executive of Southeast Asia’s largest bank is seeing — she says her firm is already reaping the rewards of its AI initiatives, and it’s only just the beginning.
“It’s not hope. It’s now. It’s already happening. And it will get even better,” DBS CEO Tan Su Shan told CNBC on the sidelines of Singapore Fintech Week, when asked about the promise of AI adoption.
DBS has been working to implement artificial intelligence across its bank for over a decade, which helped prepare its internal data analytics for recent waves of generative and agentic AI.
Agentic AI is a type of artificial intelligence that relies on data to proactively make independent decisions, plan and execute tasks autonomously, with minimal human oversight.
Tan expectsAI adoption to bring DBS an overall revenue bump of more than 1 billion Singapore dollars (about $768 million) this year, compared to SG$750 million in 2024. That assessment is based on about 370 AI use cases powered by over 1,500 models throughout its business.
“The proliferation of generative AI has been transformative for us,” Tan said, adding that the company was experiencing a “snowballing effect” of benefits thanks to machine learning.
A major area in which DBS has applied AI is in its financial services to institutional clients, with AI used to collect and leverage data for clients in order to better contextualize and personalize offerings.
According to Tan, this has resulted in “faster and more resilient” teams. The CEO believes that these uses of AI have contributed to a recent uptick in the bank’s deposit growth as compared to competitors’.
The company also recently launched a newly enhanced AI-powered assistant for corporate clients known as “DBS Joy,” which assists clients with unique corporate banking queries around the clock.
ROI concerns
Despite Tan’s strong convictions about AI, recent evidence suggests that many companies are struggling to turn their AI investments into tangible profits.
MIT released a report in July that found 95% of 300 publicly disclosed AI initiatives, encompassing generative AI investments of $30–$40 billion, had failed to achieve real returns.
However, at least in the banking sector, there are signs that the tides are turning.
While DBS doesn’t differentiate spending in generative AI from other in-house investments, other major banks have recently offered this comparison.
JPMorgan Chase CEO Jamie Dimon stated in an interview with Bloomberg TV last month that the bank is already breaking even on its approximately $2 billion of annual investments in AI adoption. That represents “just the tip of the iceberg,” he added.
Those expectations are shared by DBS, which plans to continue to accelerate its AI development to become an AI-powered bank.
The ultimate goal, according to Tan, is for its generative AI to develop into a trusted financial advisor for clients, including retail users who are expected to interact with personalized AI agents through the DBS banking app.
The bank already has over 100 AI algorithms that analyze users’ data to provide them with personalized “nudges,” such as alerts on incoming shortfalls, product recommendations, and other insights.
Continued AI investments
While DBS may already be reaping rewards from its AI adoption, Tan acknowledged that it will require continued investments, not only in capital, but in the time needed to reskill employees.
The company has launched several AI reskilling initiatives across departments this year and has even deployed a generative AI-powered coaching tool to support these efforts.
This will help the company automate mundane work and refocus its staff on building and maintaining human-to-human relationships with customers, rather than reducing headcount, Tan said.
“We’re not freezing hiring, but it does mean reskilling. And that’s a journey. It’s a never-ending journey … a constant evolution.”