stub Smaller, Faster, Better - 3D Processors to Underpin Tomorrow's Data Transfer Needs - Securities.io
Connect with us

Computing

Smaller, Faster, Better – 3D Processors to Underpin Tomorrow’s Data Transfer Needs

mm

Published

 on

Securities.io is not an investment adviser, and this does not constitute investment advice, financial advice, or trading advice. Securities.io does not recommend that any security should be bought, sold, or held by you. Conduct your own due diligence and consult a financial adviser before making any investment decisions.

University of Florida researchers are set to transform wireless communication with a novel semiconductor technology for manufacturing processors, significantly enhancing global data transfer efficiency. Published in the journal Nature Electronics, this innovation comes at a time when the demand for high-speed data transfer is rising dramatically. 

Already, the expansion of the Internet, with the number of Internet users surging significantly and global Internet traffic rising several times, has caused the demand for data centers and data transmission to grow at an extremely rapid pace.

On top of it, over the past couple of years, Artificial Intelligence (AI) has further led this demand to skyrocket. The multi-billion dollar industry of AI is changing our lives and growing at a fast pace. It is being used in everything from education, art, and entertainment to healthcare, automobiles, and much more. 

AI applications require considerable computational power to process and analyze vast amounts of data, which drives the need for more powerful and efficient hardware in data centers. 

Moreover, there is a growing demand for high-speed data transfer within data centers and across networks, as AI systems often rely on large datasets for training and continuous learning. And the more complex these AI systems, the greater the need for higher network bandwidth and improvements in higher-capacity switches, routers, and fiber optic cables to ensure efficient data movement.

The recent rise of Generative AI applications, in particular, is taking AI into new and interesting areas. As they enter the mainstream, they are calling for real-time data processing to provide quick, on-the-spot responsiveness. As such, the current data infrastructure needs a complete overhaul to meet the rising AI demands.

So, against this backdrop, scientists have successfully transitioned from planar to three-dimensional (3D) processors. 

Less than a month ago, researchers from the University of Stuttgart, Germany, along with many companies, achieved “Imaging circuits in three dimensions.” As per the study, the researchers added a diamond plate with nitrogen-vacancy (NV) centers to the area of the circuit to be studied. They then placed the entire setup in a wide-field fluorescence microscope, and the magnetic fields generated affected the spins of the NV centers, which were measured using optically detected magnetic resonance. It was able to detect currents as weak as 10 μA μm−2 with sub-micrometer spatial resolution.

So, now, making use of semiconductor technology to propel wireless communication into a new dimension is helping bring in a new era of efficiency in data transmission. The innovation is made by the team of Roozbeh Tabrizian, Ph.D., who's an associate professor at the Department of Electrical and Computer Engineering at the University of Florida.

The 3D processor, according to Tabrizian, marks a monumental moment in the evolution of wireless communication as we become progressively dependent on data exchange in real-time. He said:

“The ability to transmit data more efficiently and reliably will open doors to new possibilities, fueling advancements in areas such as smart cities, remote health care, and augmented reality.” 

Click here to learn why 2023 was a breakout year for artificial intelligence.

Evolution of Wireless Communication 

The transmission of information from a sender to the receiver without any wired connection was first achieved in the late 19th century when Heinrich Hertz demonstrated the transmission of electromagnetic waves through space in the 1880s. The unit of frequency was named ‘hertz' in his honor. Then, many years later, Guglielmo Marconi transmitted signals over long distances using radio waves in 1895. It wasn't until many decades later, in the 1970s, that the precursor to modern-day mobile phones was developed.

So, when it comes to wireless communication, it is all about the transmission of information over a distance without the help of any form of electrical conductors such as wires and cables. This form of communication is about connecting and communicating between two or more devices using a wireless signal through wireless devices and technologies. Some forms of this communication network involve Mobile, Bluetooth, Broadcast radio, Wi-Fi, and Infrared communication.

This form of communication usually involves the transfer of data using electromagnetic waves. This mechanism carries signals anywhere between a few meters, such as a TV remote control, and a thousand kilometers, like satellite communication. There is no physical mode of communication here. 

Meanwhile, electromagnetic waves include Visible Light, Gamma Rays, X-rays, Ultraviolet Rays, Infrared Rays, Radio Waves, and Microwave Rays, which pass on the signal.

Wireless communication offers several advantages, the primary being flexibility and convenience. This form of communication enables people to communicate regardless of their location. This flexibility allows people to carry their devices with them anywhere, requiring no physical connection, enabling us to connect with anyone, anywhere, anytime.

Because this method doesn't involve the use of connection wires, network cables, or the setup of elaborate physical infrastructure, wireless connections are more effective than their wired counterparts.

Improvements are also seen in terms of speed and accuracy here. Moreover, wireless technology offers better accessibility, especially in remote areas where laying out ground lines is a difficult and complex task. This easy accessibility offers constant connectivity, allowing people to respond to emergencies relatively quickly.

Given the many benefits of wireless communication, wireless technology has evolved expeditiously to meet the rising user requirements. Advancements in technology have led to higher data transmission rates and improved quality of services.

As internet connectivity becomes an essential part of our personal and professional lives, researchers, scientists, companies, and governments have been developing and investing in ways to make data transfer faster and more efficient. Earlier this year, scientists at the University of Oxford in the UK developed magnetic whirls in membranes to enable data transfer at kilometers per second. This advancement is expected to pave the way for a new generation of superfast computing platforms.

According to Hariom Jani, who's doing a postdoc at Oxford University's Department of Physics:

“Silicon-based computing is much too energy-inefficient for the next generation of computing applications such as full-scale AI and autonomous devices.” 

Researchers fabricated extremely thin crystalline hematite membranes, which combine advantageous qualities of both 2D and 3D materials and are easily transferable. These flexible membranes can be twisted in different forms without breaking, a quality utilized to shape magnetic whirls in 3D. Integration with this technology is expected to allow future computers to work like the human brain.

Another interesting development in the field took place two years ago when researchers at the University of Birmingham's School of Engineering revealed a new beam-steering antenna to enhance the efficiency of data transmission. Moreover, it opens up a range of frequencies for mobile communications that have been inaccessible to currently used technologies.

A Game-Changing Processor

As of now, the data in our mobile devices is converted into electromagnetic waves that are communicated back and forth among users worldwide. Here, spectral processors or filters are responsible for moving the data across different frequencies. Spectral processing breaks audio into tiny, discrete units, allowing for targeting it at a very fine level.

The way to understand this is through traffic lights, which ensure an efficient flow of traffic through a city. However, there's only a certain level of traffic that can be handled by a city's infrastructure. If the volume of cars keeps on increasing, it will create a problem.

Now, “we're starting to reach the maximum amount of data we can move efficiently,” stated Tabrizian. He explained that planar processors, which wireless communications have traditionally relied on, are “no longer practical as they limit us to a very limited span of frequencies.”

In the semiconductor industry, planar is a manufacturing process used to build single components of a transistor and then connect them together. It is how silicon-integrated circuit chips are built. An integrated circuit (IC), also known as a microchip, is a small electronic device made up of multiple interconnected electronic components like transistors, resistors, and capacitors etched onto a small piece of semiconductor material.

Semiconductors are an integral part of our world today as they are the ones that have helped revolutionize technology. These are materials with conductivity between conductors and insulators. They are used in the manufacturing of diodes, transistors, and ICs for their low cost, power efficiency, compactness, and reliability. It is semiconductors that provide us with electronic devices, from radios and computers to medical diagnostic equipment and much more. 

Now, with the technological advancements and the advent of AI, the substantial increase in demand requires considerably more filters at various frequencies to move data. Tabrizian said:

“Think of it like lights on the road and in the air. It becomes a mess. One chip manufactured for just one frequency doesn't make sense anymore.” 

Tabrizian and his teammates at the Herbert Wertheim College of Engineering used the CMOS technology fabrication process to build the three-dimensional nanomechanical resonator. CMOS or Complementary Metal Oxide Semiconductor comprises NMOS transistors that have N++ regions at the source and drain terminal and a p-type substrate and PMOS transistors that have two P++ regions and an n-type substrate. CMOS is used for its low power dissipation and low operating currents.

Last month, we noted in our “Computing at the Speed of Light with Silicon-Photonics” article that researchers at the University of Pennsylvania developed a chip that uses light waves instead of electricity to perform the intricate math required to train AI. 

This new chip can facilitate much faster processing speed while reducing the energy consumption of devices, paving the way for the new generation of AI development. This shows the increasing impact of and demand for AI, leading everyone to take notice and work towards making it even better.

Coming back to 3D processors, it can deliver enhanced performance while occupying less space. They also have indefinite scalability, which makes them extremely powerful in accommodating growing demands. On this, Tabrizian said:

“By harnessing the strengths of semiconductor technologies in integration, routing, and packaging, we can integrate different frequency-dependent processors on the same chip. That's a huge benefit.”

By integrating diverse frequencies on a single chip, this new type of spectral processor is “truly a game changer,” said David Arnold, associate chair for faculty affairs in Florida University's Department of Electrical and Computer Engineering. This “new approach for multi-band, frequency-agile radio chipsets,” Arnold noted, solves a huge manufacturing challenge while allowing designers to “imagine entirely new communication strategies in an increasingly congested wireless world.” 

This innovation is the result of five years of work. The team of researchers, including Faysal Hakim, Troy Tharpe, Nicholas Rudawski, and Tabrizian, first started working on this new approach to the processor in 2019. 

The study received funding from the Defense Advanced Research Projects Agency (DARPA). A research agency of the US Department of Defense (DOD), DARPA invests in and develops emerging technologies for the military's national security. Arnold stated:

“Put more simply, our wireless devices will work better, faster, and more securely.” 

Companies 

Now, let's take a look at the companies that can benefit from this research:

#1. NVIDIA Corporation 

The California-based chipmaker is known for its graphics processing units (GPUs). NVIDIA's tech is used in data centers and AI applications, which can benefit from improvements happening in data transfer efficiency. Recently addressing a question on how many more chip factories (or semiconductor fab, short for fabrication) are needed to support the expansion of the AI industry, Nvidia CEO Jensen Huang said:

“We're going to need more fabs.” 

However, they're also improving algorithms and AI processing tremendously.

finviz dynamic chart for  NVDA

With a market cap of $2.188 trillion, the shares of the company have been trading at $869, up 76.75% year-to-date (YTD). The company has a revenue (TTM) of $60.9 bln and an EPS (TTM) of 11.93, P/E (TTM) of 73.38, and ROE (TTM) of 91.46%. The company also pays a dividend of 0.02%. In the Feb. earnings call, CEO Huang talked about collaborations with healthcare, biology, financial services, robotics, and autonomous vehicles companies, as well as AI and LLM developers.

#2. Intel Corporation 

Another leading company in the semiconductor industry, Intel, could also benefit from advancements in processor technology. The corporation is currently hopeful of obtaining a massive $3.5 billion, to be dispersed over a period of three years, in funding from the US government to manufacture advanced semiconductors for the nation's military programs. 

Intel already has longstanding relationships with the DOD and the Department of Energy (DOE), which involved the production of prototype multi-die chips and processors at the core of DOE's Aurora Supercomputer.

finviz dynamic chart for  INTC

With a market cap of $187.42 billion, the company's shares have been trading at $44.33, down 12.44% YTD. The company has revenue (TTM) of $54.22 bln and an EPS (TTM) of 0.38, P/E (TTM) of 116.92, and ROE (TTM) of 1.63%. The company also pays a dividend of 1.13%.

#3. Samsung Electronics

South Korea's tech giant, Samsung, produces a range of components for wireless communication devices and has a market cap of 368.9 billion. The company's shares have been trading at 1375, down 8.28% YTD.

According to the company's results for the fiscal year 2023, Samsung reported KRW 258.94 trillion (over $197 bln) in annual revenue and KRW 6.57 trillion ($5 bln) in operating profit.

Conclusion 

As we discussed above, the need for faster data transfer continues to rise every day, especially when AI is here and advancing like crazy. In such an environment, it is important that research continues to introduce improvements in transmitting data, allowing us to have a more enriching experience. 

Gaurav started trading cryptocurrencies in 2017 and has fallen in love with the crypto space ever since. His interest in everything crypto turned him into a writer specializing in cryptocurrencies and blockchain. Soon he found himself working with crypto companies and media outlets. He is also a big-time Batman fan.