Samsung has announced that it is aiming to start manufacturing 1.4nm semiconductor chips by 2027. The company unveiled the roadmap during its annual Samsung Foundry Forum event in San Jose, US on Monday. The South Korean behemoth recently began mass producing 3nm process technology and had previously revealed plans to move to 2nm by 2025.
Samsung reveals its advanced semiconductor production roadmap
Samsung is one of the world’s largest semiconductor companies. While its in-house Exynos smartphone processors have been infamous for poor thermal management and battery efficiency, the firm has a good reputation in most other semiconductor areas. It is the world’s outright largest manufacturer of memory chips. The company also operates a foundry business, manufacturing semiconductor chips for other firms.
Earlier this year, Samsung started manufacturing 3nm chips (3GAE), becoming the world’s first foundry to do so. The company is employing the Gate-All-Around (GAA) transistor architecture in these solutions. The new architecture brings performance and power improvements over the FinFET (Fin field-effect transistor) architecture that its 4nm or bigger chips are based on. The GAA transistor architecture also allows for smaller chips.
Over the next few years, Samsung will improve the performance and power efficiency of its 3nm GAA solutions. It plans to introduce second-gen 3nm chips (3GAP) in 2024, and follow up with Plus iterations (3GAP+) in 2025. The company is touting the second-gen solutions to be 20 percent smaller than the first-gen ones, allowing for smaller and more power-efficient chips.
Samsung to produce 2nm chips by 2025 and 1.4nm by 2027
Alongside improving its existing solutions, Samsung will also work on more advanced semiconductor chips. As said earlier, the company has already revealed that it will be ready with 2nm chips by 2025. It has now announced that those chips will feature backside power delivery. Currently, transistors in a semiconductor chip can receive power or communicate through only one side of the chip. The new technology will enable simultaneous power and communication through both sides. It will bring a substantial performance boost.
In 2026, Samsung plans to introduce its second-gen 2nm chips, followed by 1.4nm solutions in 2027. By this time, its production capacity for advanced semiconductors will expand by more than three times compared to this year. The company didn’t reveal what improvements these advanced solutions will bring, but as semiconductors shrink in size, you can expect more performance and power improvements.
At the recently concluded Samsung Foundry Forum event, the company also announced that it is improving 2.5D/3D heterogeneous integration packaging technology. It will start mass-producing 3D packaging X-Cube with micro-bump interconnection in 2024. The bump-less X-Cube will be available in 2026.
Samsung is also improving its automotive and telecommunication solutions
Samsung also makes embedded non-volatile memory (eNVM) solutions for automotive customers. Its current offerings are based on 28nm technology. The company plans to move to 14eNVM solutions by 2024 (via). And in the telecommunication sector, Samsung is working on 5nm RF (Radio Frequency) chips. It has already moved from 14nm to 8nm RF solutions.
“With significant market growth in high-performance computing (HPC), artificial intelligence (AI), 5/6G connectivity, and automotive applications, demand for advanced semiconductors have increased dramatically, making innovation in semiconductor process technology critical to the business success of foundry customers,” Samsung said in a press release.
The company will also hold the Samsung Foundry Forum conference in Europe (on October 7 in Munich, Germany), Japan (October 18 in Tokyo), and South Korea (October 20 in Seoul).