Three former Meta and Google silicon executives on Monday announced they’ve raised a total of $100 million to build technology they say will reduce cloud companies’ spend on data center buildouts.
Called Majestic Labs, the startup’s co-founders are Ofer Shacham, Sha Rabii and Masumi Reynders, all of whom spent years working together leading silicon products at Meta and Google. Majestic’s patent-pending silicon design architecture includes 1,000 times the memory of a typical enterprise-grade server, the co-founders told CNBC.
Majestic develops its entire chip system, much like Nvidia, but each of the startup’s servers may replace up to 10 of today’s leading racks, they said.
The company closed its $71 million Series A funding round in September, led by Bow Wave Capital. Among its others investors is Lux Capital. Prior to Monday’s announcement, the co-founders had been quietly working on the startup since late 2023.
Majestic’s funding announcement comes as major tech companies raise their capital expenditures, primarily for data center infrastructure. Alphabet, Meta, Microsoft and Amazon each lifted their guidance for capital expenditures in October, and they collectively expect that number to reach more than $380 billion this year.
While the majority of large language models and AI workloads have relied on Nvidia’s graphics processing units, or GPUs, more companies are entering the fold. Google last week announced Ironwood, its latest tensor processing unit, or TPUs, which artificial intelligence startup Anthropic plans to use for its Claude model.
Sha Rabii is Majestic Labs President. Rabii used to lead Google’s silicon efforts.
Google
However, memory capacity remains a challenge for companies that have large amounts of data to process, which the Majestic co-founders said they hope to address.
“Nvidia makes excellent GPUs and has driven incredible AI innovation,” Shacham, Majestic’s CEO, told CNBC. “We’re not trying to replace GPUs across the board — we’re solving for memory-intensive AI workloads where the fixed compute-to-memory ratio becomes a constraint”
Majestic is going after hyperscalers and large companies that run AI models, including those from the financial and pharmaceutical industries, the co-founders said.
The technology underpinning the company includes patent-pending architectures that allow Majestic to collapse multiple racks worth of conventional equipment and memory into a single server. Majestic claims that will allow for a smaller footprint and requires less power and cooling, helping the company’s clients reduce their data center costs.
Prototypes of Majestic’s box servers will be available for some customers in 2027, and the startup has already begun discussing pre-orders, the co-founders said. Majestic declined to share its clients.
The company has fewer than 50 employees — half of which are based in Tel Aviv, Israel, while the rest are based in Los Altos, California. Majestic said it plans to grow each location and seek additional funding in the coming year.
Always in the periphery
The trio met at Google, and together, they helped stand up the team working on Google’s TPUs.
“We’d been building AI for DARPA, for Google, for Meta, but suddenly AI became this ubiquitous thing that everybody needs,” Shacham said. “Every large company needs AI, and that was a good time for me to say ‘How about we go and do that?'”
The first to join the search company was Reynders in 2003, who went to Google as a senior corporate counsel. She’s now Majestic’s COO and manages the business side of the company.
At Google, her teams were focused on business development and product strategy focused on silicon. Reynders spent 15 years at Google, where she rose to director of product management and silicon.
Rabii, Majestic’s president, earned his doctorate from Stanford University, and in 2011, he sold his last chip design company, Arda Technologies, to Google, where he rose to senior director of engineering. Under Rabii, Google launched its video chip Argos, which is used in YouTube’s data centers.
Shacham sold his company, Chip Genesis, to Google in August 2013, when Rabii was leading Google’s silicon team. Chip Genesis’s technology was used broadly across many products including the AI accelerators in Google’s Pixel smartphones. At Google, Shacham rose to head of silicon design and implementation for consumer hardware.
Majestic Labs chief operating officer and co-founder Masumi Reynders
Sharyce Rains
The three of them left Google in 2018 for the company then known as Facebook. There, they built the Facebook Agile Silicon Team, known as “FAST,” within the company’s Reality Labs hardware division. Shacham rose to vice president and head of FAST.
When Meta laid off employees from Reality Labs in 2023, Shacham had to conduct the cuts, he said. At that point, Meta was trying to conserve cash, he said.
“Part of that was layoffs across the organization, and FAST was not excluded from that,” Shacham said. “It’s not a good place to be, not a good feeling to do.”
The three founded Majestic Labs after brainstorming about the biggest bottleneck problems for silicon and AI.
“We’ve been friends and colleagues for a long time, so this notion of working together and doing something exciting has always been in the periphery,” Reynders said.
As Majestic eyes growth and hiring in 2026, the co-founders said they’re tapping into their collective of rolodex of the more than 1,500 employees they’ve worked with at prior companies, including Meta and Google.
“There’s that trust they already have with us,” Rabii said of the former colleagues.
Slope, a lending startup that uses artificial intelligence to vet businesses, is partnering with Amazon starting Tuesday to provide a reusable line of credit to Amazon sellers, backed by a JPMorgan Chase credit facility, the company told CNBC exclusively.
The new relationship means eligible U.S. Amazon vendors can apply for and access capital directly through their Amazon Seller accounts with real-time approvals.
Slope was co-founded by CEO Lawrence Lin Murata, who said said he saw the ups and downs of running a small business while he was growing up in São Paulo.
Lin Murata helped his parents at their family’s toy shop, which they’ve been running for more than three decades. As he gained more insight into the finances of the business, he said he realized that cash flow was a large pain point for his parents and other small businesses.
That led him to start Slope, an AI-powered lending platform backed by OpenAI CEO Sam Altman and JPMorgan Chase, with co-founder Alice Deng.
“Leveraging AI, we’re able to underwrite these businesses, and we’re able to handle all the complexity of assessing the risk for a business,” Lin Murata said. “At the same time, [we’re] providing a very easy, real-time experience to them.”
The lines of credit will start at an 8.99% APR, according to Slope, and require vendors to be in business for at least one year with more than $100,000 in annual revenue. Once approved, Amazon sellers can draw from the line as needed and choose a term ranging from three months to a year to align repayment with their inventory cycle. Scope did not disclose the financial aspects of its deal with Amazon.
“Most people don’t realize that sellers, independent sellers, are kind of the backbone of Amazon and e-commerce in general,” Deng told CNBC. “More than 60% of Amazon’s sales are driven by independent sellers.”
Deng said Slope is filling a gap with the new partnership. Currently, Amazon sellers can use some third parties to access capital, though Deng said those initiatives are more focused on smaller sellers, while Slope is focused on mature sellers, some of whom reach hundreds of millions of dollars in revenue and require bank-grade financing.
Deng said when Amazon did its own lending around four years ago, the total addressable market was between $1 billion and $2 billion. With Slope taking over the program, the company expects that number to grow.
“We’re excited about our work with Slope, which expands the financing tools available to Amazon selling partners,” an Amazon spokesperson told CNBC. “Whether they are just starting out or looking to grow, access to sufficient capital is a critical need for small business owners, and we’re always evaluating new ways to empower sellers to thrive in the Amazon store.”
With Slope’s new deal, sellers can take a few minutes directly on Amazon Seller Central to apply for capital and get approved almost instantly, using proprietary Amazon performance data and Slope’s in-house large language model, Lin Murata said.
“That is one of the reasons why we’re able to give a more compelling offer than if you were outside of the Amazon dashboard,” Lin Murata said. “And then we give real-time decisions, so we analyze Amazon performance, data, and cash flow in real time.”
It’s a process that the Slope co-founders said is easier, faster and more integrated than having to apply for loans at banks as a small business. With the granular data that Amazon provides, like a breakdown of sales by product, they said the AI model is able to make a more informed decision on financing than a bank would based on overall financial documents.
With the new deal, Amazon joins a growing slate of Slope’s customers, which already include Samsung, Alibaba, Ikea and more.
Deng and Lin Murata said the company has trialed the new Amazon integration, and though the trial has been live for just a few weeks, the pair said it’s seen significant demand and applications growing 300% week over week.
“Going back to the initial inspiration of my parents, I think we want to be the credit intelligence layer for these businesses,” Lin Murata said. “Ultimately, what we’re really doing is helping these businesses grow by giving them fair, affordable, fast and very easy access to different forms of financing.”
The U.S. has halted a technology trade deal with the U.K., after officials in Washington became frustrated with the pace of progress, the Financial Times reported on Tuesday.
Announced in September during President Donald Trump’s state visit to the U.K., the “technology prosperity deal” is a sweeping agreement aimed at encouraging collaboration between the countries on tech like artificial intelligence, nuclear fusion, and quantum computing.
At the time, Trump said that the deal would “ensure our countries lead the next great technological revolution side by side.” U.K. Prime Minister Keir Starmer said that the agreement was a “generational step change in our relationship with the U.S.” that would deliver “growth, security and opportunity up and down the country.”
Talks were suspended by the U.S. last week, the FT reported, quoting unnamed British officials.
When asked to comment on the report, a U.K. government spokesperson told CNBC: “Our special relationship with the US remains strong and the UK is firmly committed to ensuring the Tech Prosperity Deal delivers opportunity for hardworking people in both countries.”
The agreement would establish AI-enabled research programs in areas including the development of models and datasets in mutual priorities such as AI for biotechnology, precision medicine for cancer and rare and chronic diseases, and fusion energy, the two countries said in September.
It came as the U.K. signed deals totalling £31 billion ($41 billion) with U.S. tech firms like Microsoft, Nvidia, Google, OpenAI, and CoreWeave to build out the country’s AI infrastructure. The U.S. is the U.K.’s largest trading partner.
The U.S. Department of Commerce has been approached for comment.
The logo of an Apple Store is seen reflected on the glass exterior of a Samsung flagship store in Shanghai, China Monday, Oct. 20, 2025.
Wang Gang | Feature China | Future Publishing | Getty Images
A shortage of memory chips fueled by artificial intelligence players is likely to cause a price rise in smartphones in 2026 and a drop in shipments, Counterpoint Research said in a note on Tuesday.
Smartphone shipments could fall 2.1% in 2026, according to Counterpoint, versus a previous outlook of flat-to-positive growth.
Shipments do not equate to sales but are a measure of demand as they track the number of devices being sent to sales channels like stores.
Meanwhile, the average selling price of smartphones could jump 6.9% year-on-year in 2026, Counterpoint said, in comparison to a previous forecast of a 3.6% rise.
The continued build-out of data centres globally has hiked demand for systems developed by Nvidia, which in turn uses components designed by SK Hynix and Samsung — the two biggest suppliers of so-called memory chips.
However, a specific component called dynamic random-access memory or DRAM, which is used in AI data centers, is also critical for smartphones. DRAM prices have surged this year as demand outstrips supply.
For low-end smartphones priced below $200, the bill of materials cost has increased 20% to 30% since the beginning of the year, Counterpoint said. The bill of materials is the cost of producing a single smartphone.
The mid and high-end smartphone segment has seen material costs rise 10% to 15%.
“Memory prices could rise another 40% through Q2 2026, resulting in BoM costs increasing anywhere between 8% and over 15% above current elevated levels,” Counterpoint said.
The rising price of components could be passed on to consumers and that will in turn, drive the rise in the average selling price.
“Apple and Samsung are best positioned to weather the next few quarters,” MS Hwang, research director at Counterpoint, said in the note. “But it will be tough for others that don’t have as much wiggle room to manage market share versus profit margins.”
Hwang said this will “play out especially” with Chinese smartphone makers who are in the mid-to-lower end of the market.
Counterpoint said some companies may downgrade components like camera modules, displays and even audio, as well as reusing old components. Smartphone players are likely to try to incentivize consumers to buy their higher-priced devices too.