TSMC Prepares Next-Gen HBM4 Base Dies, Leveraging 12nm and 5nm Nodes


May 16, 2024 by our News Team

TSMC reveals plans to use N12FFC+ and N5 processes for HBM4 base dies, enabling advanced packaging methods and improved performance for high-performance processors in AI and HPC applications.

  • TSMC's advanced packaging methods will accommodate the wider 2048-bit interface of HBM4 memory stacks.
  • The use of N12 and N5 processes will give TSMC an advantage in the HBM4 manufacturing process.
  • TSMC's collaboration with EDA partners will ensure the certification of HBM4 channel signal integrity, IR/EM, and thermal accuracy.


HBM4 memory is set to bring several significant changes, and one of the most notable is the wider memory interface. Moving from a 1024-bit interface to an ultra-wide 2048-bit interface, HBM4 memory stacks will require chip manufacturers to adopt more advanced packaging methods to accommodate the increased width.

During TSMC’s European Technology Symposium 2024 presentation, the company shared details about the base dies it will manufacture for HBM4 using logic processes. TSMC plans to utilize variations of its N12 and N5 processes, which will give them an advantage in the HBM4 manufacturing process. This is because current memory fabs are not equipped to produce such advanced logic dies in a cost-effective manner, if at all.

For the initial wave of HBM4, TSMC will employ two fabrication processes: N12FFC+ and N5. While both processes serve the same purpose of integrating HBM4E memory with next-generation AI and HPC processors, they will be used differently to connect memory for high-performance processors in AI and HPC applications.

TSMC’s base die manufactured on the N12FFC+ process will be used to install HBM4 memory stacks on a silicon interposer alongside system-on-chips (SoCs). TSMC believes that their 12FFC+ process is well-suited for achieving HBM4 performance, enabling memory vendors to build 12-Hi (48 GB) and 16-Hi (64 GB) stacks with per-stack bandwidth exceeding 2 TB/second.

In addition, TSMC is optimizing CoWoS-L and CoWoS-R for HBM4, utilizing over eight layers to enable HBM4’s routing of over 2,000 interconnects with proper signal integrity. HBM4 base dies on N12FFC+ will play a crucial role in building system-in-packages (SiPs) using TSMC’s advanced packaging technology, CoWoS-L or CoWoS-R. These technologies offer interposers up to 8x reticle size, providing enough space for up to 12 HBM4 memory stacks. Presently, HBM4 achieves data transfer rates of 6 GT/s at currents of 14mA.

TSMC collaborates with EDA partners such as Cadence, Synopsys, and Ansys to certify HBM4 channel signal integrity, IR/EM, and thermal accuracy.

Furthermore, memory manufacturers will have the option of utilizing TSMC’s N5 process for their HBM4 base dies. N5-built base dies will offer even more logic, lower power consumption, and higher performance. The most significant advantage of this advanced process technology is the ability to achieve very small interconnect pitches, ranging from 6 to 9 microns. This allows N5 base dies to be used in conjunction with direct bonding, enabling HBM4 to be 3D stacked directly on top of logic chips. Direct bonding is expected to enhance memory performance, which will greatly benefit AI and HPC chips that require increased memory bandwidth.

TSMC is known to collaborate with SK hynix on HBM4 base dies, and it is likely that they will also produce base dies for Micron. However, it would be surprising to see TSMC working with Samsung, as the conglomerate already has its own advanced logic fabs through Samsung Foundry.

About Our Team

Our team comprises industry insiders with extensive experience in computers, semiconductors, games, and consumer electronics. With decades of collective experience, we’re committed to delivering timely, accurate, and engaging news content to our readers.

Background Information


About Samsung: Samsung, a renowned South Korean multinational conglomerate, has established itself as a global leader in various industries, including electronics, technology, and more. Founded in 1938, Samsung's influence spans from smartphones and consumer electronics to semiconductors and home appliances. With a commitment to innovation, Samsung has contributed groundbreaking products like the Galaxy series of smartphones, QLED TVs, and SSDs that have revolutionized the way we live and work.

Samsung website  Samsung LinkedIn

About SK hynix: SK Hynix is a prominent South Korean semiconductor company known for its innovative contributions to the global technology landscape. Specializing in the production of memory solutions, SK Hynix has played a vital role in shaping the semiconductor industry. With a commitment to research and development, they have continuously pushed the boundaries of memory technology, resulting in cutting-edge products that power various devices and applications.

SK hynix website  SK hynix LinkedIn

About Synopsys: Synopsys is a prominent American software company that specializes in electronic design automation (EDA) and semiconductor intellectual property. Founded in 1986, Synopsys provides cutting-edge tools and solutions for designing and testing complex integrated circuits and electronic systems. Their software aids in the development of semiconductors, electronic products, and software applications, playing a crucial role in advancing technological innovation across various industries. For more detailed information about their offerings and contributions, you can visit their official website at

Synopsys website  Synopsys LinkedIn

About TSMC: TSMC, or Taiwan Semiconductor Manufacturing Company, is a semiconductor foundry based in Taiwan. Established in 1987, TSMC is a prominent player in the global semiconductor industry, specializing in the manufacturing of semiconductor wafers for a wide range of clients, including technology companies and chip designers. The company is known for its cutting-edge semiconductor fabrication processes and plays a critical role in advancing semiconductor technology worldwide.

TSMC website  TSMC LinkedIn

Technology Explained


CoWoS: CoWoS, or Chip-on-Wafer-on-Substrate, is a recent advancement in chip packaging that allows for more powerful processors in a compact size. This technology stacks multiple chips on a silicon interposer, enabling denser connections and improved performance. Developed for high-performance computing, CoWoS promises faster processing, lower power consumption, and the ability to pack more processing power into smaller devices.


EDA: EDA stands for Electronic Design Automation, and it refers to a category of software tools and solutions used in the design and development of electronic systems and integrated circuits. EDA tools assist engineers and designers in creating complex electronic designs, from individual components to entire systems, by automating various aspects of the design process. These tools encompass a wide range of functionalities, including schematic capture, simulation, layout design, verification, and testing.


Foundry: A foundry is a dedicated manufacturing facility focused on producing semiconductor components like integrated circuits (ICs) for external clients. These foundries are pivotal in the semiconductor industry, providing diverse manufacturing processes and technologies to create chips based on designs from fabless semiconductor firms or other customers. This setup empowers companies to concentrate on innovative design without needing substantial investments in manufacturing infrastructure. Some well-known foundries include TSMC (Taiwan Semiconductor Manufacturing Company), Samsung Foundry, GlobalFoundries, and UMC (United Microelectronics Corporation).


HPC: HPC, or High Performance Computing, is a type of technology that allows computers to perform complex calculations and process large amounts of data at incredibly high speeds. This is achieved through the use of specialized hardware and software, such as supercomputers and parallel processing techniques. In the computer industry, HPC has a wide range of applications, from weather forecasting and scientific research to financial modeling and artificial intelligence. It enables researchers and businesses to tackle complex problems and analyze vast amounts of data in a fraction of the time it would take with traditional computing methods. HPC has revolutionized the way we approach data analysis and has opened up new possibilities for innovation and discovery in various fields.





Leave a Reply