General decoding algorithm; data transmission can be carried out without heating.

General Decoding Algorithm Researchers at MIT, Boston University, and Maynooth University use a method called guessing random additive noise decoding (GRAND).

The coded data transmitted through the network is susceptible to interference from interference signals. Error correction codes use hashes, allowing the decoding algorithm to determine the possible original data based on the structure of the hash.

In contrast, GRAND works by guessing the noise that affects the message and using noise patterns to infer the original information. GRAND generates a series of noise sequences in the order in which they may appear, subtracts them from the received data, and checks whether the resulting codeword is in the codebook. Although noise appears random in nature, it has a probabilistic structure that allows algorithms to guess what it might be.

“In a way, it’s similar to troubleshooting. If someone drives their car into the store, the mechanics don’t start by mapping the entire car to the blueprint. Instead, they first ask: “It’s most likely to go wrong What's the place? "Maybe it only needs gasoline. If this doesn't work, what's the next step? Maybe the battery is dead?" said Muriel Médard, a professor in the Department of Electrical Engineering and Computer Science at MIT.

The GRAND chip uses a three-layer structure, starting with the simplest possible solution in the first stage, and processing longer and more complex noise patterns in the subsequent two stages. Each stage runs independently, which increases the throughput of the system and saves power.

The device is also designed to seamlessly switch between two codebooks. It contains two static random access memory chips, one can break the decoded words, the other loads a new codebook, and then switch to decoding without any downtime.

Since the codebook is only used for verification, it is suitable for legacy codes and codes that have not yet been introduced.

"For reasons I'm not sure about, people are in awe of coding as if it is black magic. The process is mathematically bad, so people just use the code that already exists. I hope this will be discussed again. Discussion, so it’s not so standards-oriented, enabling people to use existing code and create new code,” Médard said.

In the test, the GRAND chip can effectively decode any moderately redundant code up to 128 bits with a delay of about 1 microsecond.

Médard said that developing chips means rethinking preconceived notions about hardware design. "We can’t go out and reuse what we’ve done. It’s like a complete whiteboard. We have to really think about each component from the ground up. This is a journey of rethinking. And I think when we make the next chip, we Will realize that certain things in the first chip are out of habit or assumption that we can do better."

Next, the team plans to reconfigure the GRAND chip for soft detection, experiment with longer and more complex codes, and optimize energy efficiency.

Data transmission without heating. Researchers from the Australian National University, Swinburne University of Technology, and Carl von Ossietzky Universität Oldenburg suggested the use of atomic-level thin semiconductors as an energy-saving data transmission method.

The material is based on a single layer of transition metal dichalcogenide crystal tungsten disulfide (WS2) embedded in an optical microcavity. Early research shows that it does not heat up, so it wastes less energy.

"Computers have used about 10% of all available electricity in the world. This figure is accompanied by huge financial and environmental costs, and due to the increasing demand for computing, it is expected to double every 10 years," a doctoral student from the United States Matthias Wurdak said. School of Physics Research, Australian National University. "A lot of energy used by computers is wasted because the electricity used to power the computers heats the equipment while performing tasks."

The researchers explained that in TMDC, “the combined electron-hole pairs (excitons) are stable at room temperature and interact strongly with light. When TMDC is embedded in an optical microcavity, the excitons can interact with cavity photons Hybridization forms excitonic polarons, inheriting useful properties from their composition." They were able to demonstrate polaron capture and ballistic propagation spanning tens of microns at room temperature.

The team next plans to integrate the technology into transistors.

"There are many other options for future research, including the development of energy-efficient sensors and lasers based on this semiconductor technology," said Professor Elena Ostrovskaya of the Australian National University.

Name* (Note: This name will be displayed publicly)

Email* (this will not be displayed publicly)

Some market segments are normalizing, and other market segments may be affected in 2022.

Threats are growing and expanding, but what is considered sufficient may vary by application or user. Even so, it may not be enough.

Large-scale expansion campaigns target various chips, but export controls limit the growth of the lead.

How doping atoms make qubits last longer.

The CTO of JCET talked about the slowdown in Moore's Law and the growing interest in new packaging methods and small chips.

From a design perspective, some things will get better, while others will get worse.

New and faster memory designs are being developed, but their future is uncertain.

Sharing resources can significantly increase utilization and reduce costs, but this is not a simple change.

The new level of system performance brings new trade-offs.

Manufacturing technologies that cannot provide sufficient analog design capabilities are not commercially viable processes. But how good must it be?

$3.5B for 35 companies scattered around the world, including IC manufacturing equipment, AI, ADAS, and batteries.

Intelligence, not just functions, is evolving in all areas from watches and glasses to clothing, but there are still some obstacles to overcome.

It is not easy to include an interposer in today's design, but as wrinkles are eliminated, new tools, methods, and standards will make it accessible to the general public.