Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

TSMC Huang Hansen: transistors can reach 0.1 nm in 2050, hydrogen atomic scale

2025-04-05 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > Internet Technology >

Share

Shulou(Shulou.com)06/03 Report--

Phillip Wong, head of research and development at TSMC, said at a recent Hot Chips conference in Silicon Valley that Moore's Law is still valid and in good condition, and that with the progress of new technology, the transistor will reach 0.1nm by 2050, about the size of a hydrogen atom.

"there is no doubt that Moore's Law is still valid and in good condition, it is not dead, it is not slowing down, it is not sick."

Philip Wong (Huang Hansen), head of research and development for TSMC, said at the Hot chips conference that he showed TSMC's vision for chip technology, saying that by 2050, the transistor would be reduced to the hydrogen atomic scale, or 0.1nm.

Huang Hansen became vice president of corporate research for TSMC in August last year. Before that, he was a professor in the Department of Electrical Engineering at Stanford University, specializing in new forms of storage technology research, and has a high reputation in academic circles.

The title of Huang Hansen's speech at the Hot Chips conference is: "what will the next semiconductor process node bring us?" He also details the development of other chip technologies, such as storing memory directly above the processor, which he expects to improve performance

Being very optimistic about Moore's Law, TSMC put forward three major directions of chip development.

Moore's Law predicts that the number of transistors that can hold on an integrated circuit doubles about every 18 months, and its performance doubles (that is, more transistors make it faster).

However, how to put billions of transistors in a single chip in the most cost-effective way has become the biggest challenge in chip manufacturing, so in recent years, many people think that Moore's Law approaches the physical limit and begins to slow down.

Intel has been trying to develop advanced processes, but the price of individual transistors does not continue to fall across the industry. This limits the new manufacturing process can only be used for high-end, high-cost chips. The good old days of the chip industry are gone forever, when the clock speed of the chip increased, but the power consumption was not affected.

So it is not surprising that there are pessimists in the chip industry.

Huang Hansen predicts that the processor will be made up of 3D stacks of different chip components, which are usually separate at present. This means that the chip has a smaller size and higher performance.

However, TSMC, as the leader of foundry foundry, is very optimistic. Huang Hansen said Moore's Law was making good progress and boldly predicted progress by 2050, although he did not provide any detailed plans.

David Kanter, an analyst at Real World Technologies, is more cautious. Since TSMC is now tied with Intel, rather than after Intel, TSMC has to assume more leadership responsibility and invest more in research and development, so it is not surprising to hear that the company is so optimistic. But when it comes to the progress of chips, Huang Hansen avoids some practical issues, such as the slowdown in shrinking transistors and the increased cost of making the latest generation of products.

Fundamental improvement

"We look forward to seeing more innovation in different directions, which will provide sustained benefits to the industry." "this is what we care about," Huang Hansen said. "

Huang Hansen said that the components of chip technology are shrinking to a very small size.

With regard to the future technology route, Philip Wong believes that transistors such as carbon nanotubes (1.2nm scale) and 2D layered materials can make transistors faster and smaller; at the same time, phase change memory (PRAM), rotation moment transfer random access memory (STT-RAM), etc., will be packaged directly with the processor to reduce the size and speed up data transfer; in addition, there is 3D stacking packaging technology.

Specifically, Huang Hansen puts forward some views on the future direction of development:

The new technology will make transistors faster and smaller. A technology that has been considered for a long time, carbon nanotubes, is now becoming feasible. The other is a material called 2D layered material, which provides similar enhancements by making it easier for electrons to flow through the chip. Some new memory technologies will be built directly into the processor rather than being connected as separate chips. This fast connection will greatly improve performance because the logic circuit on the chip (the part that processes the data) will get the required data more quickly, so you don't have to have too much idle time. 3D stacking technology will mean that today's isolated computer processor functions can be sandwiched in multiple layers and connected to high-speed data paths.

"in these systems, multi-tier logic and memory are integrated in a fine-grained manner, and connectivity is the key," Huang said.

Analyst Nathan Brookwood said that while Huang Hansen is very concerned about technologies such as carbon nanotubes, he does not think TSMC itself will bet on any particular new technology at this stage.

However, Huang Hansen stressed that in addition to hardware, software algorithms also need to catch up. Once this is achieved, advances in chips will provide better computing devices. This is crucial, Huang Hansen said. "the social demand for advanced technology is endless."

Next, Xin Zhiyuan brings the complete PPT of Huang Hansen's keynote speech at Hot Chip 2019, with an edited interpretation.

TSMC Hot Chips Conference speech (with PPT)

Moore's Law is about component density, which is the main driving force of high-performance computing.

From the logarithmic graph, Moore's law is not only not dead, but also alive very well, the transistor density is still increasing, and will continue to increase in the foreseeable future, as for clock speed and operating efficiency and other new attributes that people are equally concerned about, actually beyond the scope of Moore's law.

In the era of AI and 5G, the problem of "memory wall" has become increasingly prominent, the demand for the flow and transfer of massive data is getting higher and higher, and memory access determines the energy efficiency of computing.

Deep neural network needs a lot of memory capacity, and the problem of memory shortage will be more prominent in the future. More SRAM is needed on the chip, but it's never enough. What matters is what kind of memory.

Most of the existing systems are 2D and 2.5D, using TSV, and we need to take a step forward and move into 3D.

The next step is Beyond 3D, which realizes the multi-layer integration of logic and memory, and realizes the high-density TSV process on the nanometer scale, that is, the "N3XT" system.

Next-generation memory needs to be high-bandwidth, high-capacity, and on-chip.

The research shows that the memory with the above conditions can increase the system-level benefit by nearly 2000 times, of course, it is difficult to achieve with the existing technology. It is difficult to build high-performance transistors on the upper layer because it requires a high temperature of 1000 degrees and the memory layer melts.

To achieve the ideal system mentioned above, an ultra-thin equipment layer and a lower manufacturing temperature are needed.

In recent years, transistor technology has made a lot of progress, such as excessive metal design of 2D layer materials, 1D carbon nanotube design and so on. These materials are very thin, which greatly reduce the channel width of transistors, but still maintain a high mobility level.

Realize the integration of memory and logic platform in 3D architecture, so that the progress of transistor and manufacturing technology becomes a continuous unity.

In order to achieve this goal, it is impossible to fight on our own. This requires the close cooperation of system engineers and developers, the closer communication of hardware equipment manufacturing technology and requirements, and the establishment of closer ties between academia and industry.

Https://www.toutiao.com/i6729340752591077902/

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

Internet Technology

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report