Want to Win a Chip War? You're Gonna Need a Lot of Water

Want to Win a Chip War? You're Gonna Need a Lot of Water

By Wired, July 21, 2023

The chip industry’s thirst for water springs from the need to keep silicon wafers free from even the tiniest specks of dust or debris to prevent contamination of their microscopic components.

Credit: Bill Varie/Getty Images

Building a semiconductor factory requires enormous quantities of land and energy, then some of the most precise machinery on Earth to operate. The complexity of chip fabs, as they are called, is one reason why the US Congress last year committed more than $50 billion to boost U.S. chip production in a bid to make the country more technologically independent.

But as the U.S. seeks to boot up more fabs, it also needs to source more of a less obvious resource: water. Take Intel's ambitious plan to build a $20 billion mega-site outside Columbus, Ohio. The area already has three water plants that together provide 145 million gallons of drinking water each day, but officials are planning to spend heavily on a fourth to, at least in part, accommodate Intel.

Water might not sound like a conventional ingredient of electronics manufacturing, but it plays an essential role in cleaning the sheets, or wafers, of silicon that are sliced and processed into computer chips. A single fab might use millions of gallons in a single day, according to the Georgetown Center for Security and Emerging Technology (CSET)—about the same amount of water as a small city in a year.

Chip companies hoping to take advantage of the CHIPS and Science Act, last year's federal spending package aiming to boost US chip manufacturing, are now constructing new water processing facilities alongside their fabs. And cities trying to attract new factories funded by the legislation are studying the potential impact on their water supplies. In some places it may be necessary to secure the water supply; in others, new infrastructure must be installed to recycle water used by fabs.

From Wired

View Full Article  

More Details, but Not Enough

Interesting point, but do not fully understand.    Posting to be revisited.

More Details, but Not Enough

By Gregory Goth, Commissioned by CACM Staff, March 29, 2023

Nearly two years since the publication of the paper in Nature, Google has not yet fully open-sourced the data or code on which its claims were based.

The contentious discussion over the validity of Google researchers' claim that machine learning agents could achieve superhuman results in creating plans for computer chips entered a new, more public phase Tuesday (March 28), with a leading researcher in design automation finding the Google technology did not perform as its authors claimed in a paper published nearly two years ago in Nature.

The dispute around the Nature paper's claims has bubbled for nearly a year in prepared public statements and GitHub code repositories and FAQ sections; researchers directly involved in the situation have declined to speak extemporaneously for the public record. Even some subject matter experts have not wished to speak openly, given Google's dominant position in its ability to distribute research resources to academic computer scientists. However, Tuesday's presentation by Andrew Kahng, a prominent University of California, San Diego researcher in the field of electronic design automation (EDA), at the 2023 ACM/IEEE International Symposium on Physical Design, could elevate the issue to a more open avenue of argument among industry and academic experts.

Briefly stated, the authors of the Nature paper claimed their reinforcement learning (RL) agents could revolutionize the labor-intensive task of floorplanning—the architecting of the incredibly intricate network of memory components (called macro blocks) and logic circuitry (standard cells) on a chip. "Our method generates manufacturable chip floorplans in under six hours, compared to the strongest baseline, which requires months of intense effort by human experts," the authors wrote.

Kahng served as a peer reviewer for the paper, and also wrote an encapsulation for the news and views section of the journal, quoting science fiction author Arthur C. Clarke's observation that any sufficiently advanced technology is indistinguishable from magic.

"To long-time practitioners in the fields of chip design and design automation, (lead author Azalia) Mirhoseini and colleagues' results can indeed seem magical," Kahng wrote.

How open is open?

Science is not magic, however, and the Google paper's claims took the research community by storm. At the conclusion of his summation, Kahng wrote, "We can therefore expect the semiconductor industry to redouble its interest in replicating the authors' work, and to pursue a host of similar applications throughout the chip-design process."

For researchers who presumably were interested in trying to replicate those results, the Google team noted at the end of the paper that "the data supporting the findings of this study are available within the paper and the Extended Data," and that "the code used to generate these data is available from the corresponding authors upon reasonable request."

US-China Chip War Continues

 A Space I am watching, regards the players and supply chain.  Point me to other relevant information.

US-China chip war: How the technology dispute is playing out    By Suranjana Tewari and Jonathan Josephs  BBC News  (excerpt) 

The US is rapidly ramping up efforts to try to hobble China's progress in the semiconductor industry - vital for everything from smartphones to weapons of war.

In October, Washington announced some of the broadest export controls yet - requiring licences for companies exporting chips to China using US tools or software, no matter where they're made in the world.  Washington's measures also prevent US citizens and green card holders from working for certain Chinese chip companies. Green card holders are US permanent residents who have the right to work in the country.

It is cutting off a key pipeline of American talent to China which will affect the development of high-end semiconductors.

Why is the US doing this?

Advanced chips are used to power supercomputers, artificial intelligence and military hardware.  The US says China's use of the technology poses a threat to its own national security.

Alan Estevez, undersecretary at the US Commerce Department announced the rules, saying his intention was to ensure the US was doing everything it could to prevent "sensitive technologies with military applications" from being acquired by China.  "The threat environment is always changing and we are updating our policies today to make sure we're addressing the challenges," he said.

Meanwhile, China has called the controls "technology terrorism".

Countries in Asia that produce chips - such as Taiwan, Singapore and South Korea - have raised concerns about how this bitter battle is affecting the global supply chain.

And there were three significant developments in the chip conflict over the past week.

More Chinese firms on 'entity list'

The Biden administration has added 36 more Chinese companies, including major chipmaker YMTC to Washington's "entity list".

It means American companies will need government permission to sell certain technologies to them, and that permission is difficult to secure.   The US restrictions have broad implications. Last week, UK-based computer chip designer Arm confirmed that it was not selling its most advanced designs to Chinese firms including tech giant Alibaba because of US and UK controls.

Arm said it was "committed to adhering to all applicable export laws and regulations in the jurisdictions in which it operates."  .... ' 

New Computing Architecture

At a school I attended, worth a look.   Technical

Researchers at the University of Pennsylvania Propose a New Computing Architecture Ideal for Artificial Intelligence (AI)

By Khushboo Gupta- October 5, 2022

Conventional computing architectures severely constrain artificial intelligence’s ability to improve technology. In traditional models, memory storage and computing occur in separate areas of the machine. This is why data must be transported from its storage area to a CPU or GPU for processing. The most significant disadvantage of this design is that this movement takes time, which reduces the performance of even the most potent processing units available. There is no avoiding lag when compute performance exceeds memory transfer. These delays become a severe issue when dealing with the massive amounts of data required for machine learning and AI applications. 

Researchers have focused on hardware innovation to achieve the necessary increases in speed, agility, and energy efficiency as AI software advances in sophistication and the rise of the sensor-heavy Internet of Things produces larger datasets. A team of researchers from the University of Pennsylvania’s School of Engineering and Applied Science, in collaboration with researchers from Sandia National Laboratories and Brookhaven National Laboratory, have created a new computing architecture based on compute-in-memory (CIM), which is ideal for AI. Processing and storage take place simultaneously in CIM systems, which helps to reduce energy consumption and eliminate transfer time. The new CIM design from the team stands out for containing no transistors. This design is specifically adapted to how Big Data applications have changed how computing works today.

Transistors limit the speed at which data may be accessed, even in a compute-in-memory architecture. They utilize more time, space, and energy than is ideal for AI applications since they require much wire in a chip’s overall circuitry. The transistor-free design by the team is distinctive since it is straightforward, quick, and uses less energy. The researchers clearly emphasize that the advancement is not limited to circuit-level design. Their earlier materials science research on a semiconductor known as scandium-alloyed aluminum nitride (AlScN) was the foundation for the new computing architecture. Ferroelectric switching is possible with AlScN, making it faster and more energy-efficient than other nonvolatile memory components. Another crucial feature is the material’s ability to be deposited at temperatures low enough to work with silicon foundries. This makes it possible for the architecture to be space-efficient, which is crucial for small chip designs. .... '    (much more, Computing Technical)