This is an important step for Samsung as it tries to close the gap with SK Hynix, its long-time competition and the top supplier of advanced memory used in Nvidia’s accelerators. 

Delays in Samsung’s earlier HBM programmes weighed heavily on its earnings and share price last year. This time, the company is moving faster and with more at stake.

The initial production run is tied to months of qualification testing with Nvidia and AMD. Samsung cleared those tests late last year, opening the door to commercial shipments in early 2026. The company has not disclosed volumes, and it is unclear how quickly supply will scale.

Samsung declined to comment on the development. Nvidia did not respond to requests for comment.

Following the development, Samsung shares rose more than 2% in early trading, while SK Hynix shares slipped by nearly 3%. Investors are reading this as a sign that competition in the HBM market is about to increase.

HBM4 is the sixth generation of high-bandwidth memory, designed to handle the heavy data loads of artificial intelligence and high-performance computing. 

Compared with HBM3E, it gives higher bandwidth while using less power. That combination is important as chipmakers push systems harder to train and run ever-larger models.

For Nvidia, memory supply is now a strategic issue, not just a technical one. The company’s next platform, Vera Rubin, is already in production and will rely on HBM4 to achieve its performance targets. 

As Nvidia’s chief executive Jensen Huang said earlier this month, the platform is in “full production,” ahead of a launch later this year.

Until now, SK Hynix has carried most of that burden. It supplies the bulk of the HBM used in Nvidia’s current accelerators and has already locked in supply talks with major customers for next year. 

The company is also expanding capacity, with plans to start deploying wafers into its new M15X factory in Cheongju early next month. It has not said whether HBM4 will be part of that first output.

Micron is also pushing into the space, adding pressure to an already tight market. Demand for high-bandwidth memory is surging as data centres scale up, and analysts expect the market to grow sharply over the next few years.

Samsung’s entry into HBM4 production changes the balance. It gives Nvidia and AMD another qualified supplier, reduces the risk of shortages, and may help cool pricing pressures. It is a chance for Samsung to regain ground in a segment it cannot afford to miss.

Both Samsung and SK Hynix are due to report quarterly earnings later this week. Investors will be listening for any detail on HBM4 orders, production timelines and customer commitments.

The startup’s debut product, Chipmind Agents, is aimed at automating the repetitive, low-level work that ties up engineering teams and drags out design cycles.

Chipmind describes its agents as a new class of tools that work from a customer’s own, proprietary design data.

The agents are built to slot into existing engineering flows, learn a company’s toolchain and design hierarchy, and then carry out multi-step design and verification tasks autonomously, all while leaving final control with the human engineer. The company claims engineers can save roughly 40% of their time on routine chores.

The problem Chipmind targets is familiar to anyone who has worked in chip design: massive, customised EDA flows that simply were not built to talk to modern automation. Chipmind’s founders say they didn’t try to rip out those legacy systems.

Instead, they built a platform that prepares them for agentic automation and wraps intelligence around the existing stack.

“In the semiconductor industry, deep customisation and data protection are fundamental, but true design awareness is what separates a generic tool from an intelligent partner. Each company’s chip is a complex hierarchy with unique constraints, surrounded by a proprietary environment of tools and workflows,” said Harald Kröll, co-founder and CEO of Chipmind. 

“That is the reality we built for. Our ‘design-aware’ agents are engineered to holistically understand the entire chip context, not just the surrounding tools. We’ve found this deep awareness is the key that unlocks productivity, translating directly into significant time savings on the most complex tasks, all while integrating seamlessly into existing workflows.”

The startup’s origin is rooted in academic and industry experience. Co-founders Harald Kröll and Sandro Belfanti met at ETH Zurich during their PhDs and together have been involved in the development of more than 20 chips, from mobile modems to system-on-chip designs. Their experience, they say, exposed how much of chip engineering is precise but repetitive work.

“Anyone who’s spent time in chip development knows how much of the work is repetitive and time-consuming, demanding precision but not necessarily creativity,” said Sandro Belfanti, Co-Founder and CTO of Chipmind. 

“Throughout my career developing chips at top-tier semiconductor companies, I’ve often wished for a solution that could magically take care of those tedious tasks so I could focus on solving real engineering challenges.

“With Chipmind Agents, we’re finally bringing that solution to life: AI agents that can autonomously handle the boring parts, letting engineers focus on what truly matters: innovation.”

Chipmind’s first funding round was led by Founderful, with several semiconductor industry angels joining the table. The founders say the money will go to hiring engineers, speeding up product work and deepening relationships with strategic customers.

The launch arrives as chip design grows continually harder, demand for compute rises, design complexity increases, and simply adding headcount is not a realistic fix. Chipmind places itself as a practical bridge, a way to keep existing toolchains while automating the parts of the process that slow teams down.

Edouard Treccani, principal at Founderful, added: “In a world buzzing with AI every day, Chipmind stands out as a refreshingly real solution to a problem Harald and Sandro have spent 20 years deep in. From day one, they’ve built in close dialogue with the market, and the early feedback has been remarkably positive. Founderful is thrilled to be part of their journey!”

Chipmind is offering demos to semiconductor groups interested in testing its agents and says it will continue scaling its engineering team as it works with early customers.

U.S. shares rose 7% on Thursday, with a further 2.6% gain after hours, while Frankfurt-listed shares climbed 3.6% on Friday. Investors are betting that government backing could provide much-needed stability for the struggling chipmaker.

The potential investment was first reported by Bloomberg, noting discussions that followed an 11 August meeting between President Donald Trump and Intel’s Chief Executive Officer, Lip-Bu Tan. 

The talks reportedly focused on how Washington could accelerate domestic semiconductor manufacturing, with the delayed Ohio mega-fab project expected to be a central part of the plan.

The development comes amid one of Intel’s most challenging periods in decades. The company posted a $2.9 billion net loss in the second quarter of 2025, driven by $1.9 billion in severance costs and $800 million in asset impairments. 

A restructuring plan has seen 25,000 jobs cut and major chip fabrication projects in Germany, Poland, and Ohio scrapped or delayed.

Trump’s involvement with Intel has been far from smooth. Days before the reported stake talks, he called for Tan to resign over what he described as “highly conflicted” ties to Chinese firms. 

His comments followed a letter from Senator Tom Cotton, alleging Tan’s investments in over 100 Chinese technology companies, including at least eight linked to the People’s Liberation Army.

Tan’s past leadership of Cadence Design Systems has also resurfaced in political debate. In July 2025, the company admitted to illegally exporting chip design software to a Chinese military university, paying $140 million in fines.

Despite political issues, Intel aims to continue cooperating with the White House. “Intel is deeply committed to supporting President Trump’s efforts to strengthen U.S. technology and manufacturing leadership,” a company spokesperson said. 

“We look forward to continuing our work with the Trump Administration to advance these shared priorities, but we are not going to comment on rumours or speculation.”

Tan has also made it apparent that there will be no return to unchecked spending. “There are no more blank cheques. Every investment must make economic sense,” he said.

Intel has already secured nearly $8 billion in federal subsidies through the CHIPS and Science Act, placing it among the largest beneficiaries of U.S. semiconductor funding.

Analysts suggest a direct government stake could help stabilise Intel’s finances and restore competitiveness against rivals Nvidia, AMD, and TSMC.