Intels stunning fumble in the execution of its plan for 7 nm nanochips, rocked stock markets. The incident brought into dramatic relief the crucial role of nanotechnology for the future of the semiconductor industry.

In a single day of trading, the company lost a whopping forty billion dollars indicating how much keeping pace with nanotechnology will affect its prospects. By contrast, AMD and Taiwan Semiconductor Manufacturing rose over fifty percent after their most recent announcement of earnings.

The end of Moores Law has accelerated innovation in chip design aided by nanotechnology. (Read also: More Than Moore: 50 Years of Moore's Law.) As chips size becomes infinitesimally small, memory, processing, storage, and sensors, are assembled into single system-on-a-chip,

Join nearly 200,000 subscribers who receive actionable tech insights from Techopedia.

Carbon nanotubes, the material used for such systems, have higher speeds, conduct electricity better, and generate less heat.

Sri Savamedam senior vice-president, CMOS technologies, at Belgium-based imec, a global R&D and product development company in nano and digital technologies industries, told us:

New materials innovations are needed to lower back-end resistance. Additionally, novel low-temperature semiconductors like IGZO ( (Indium Gallium Zinc Oxide) will find use in 3D DRAMs. 2D materials like WS2 will find use in new transistor channels. Finally, novel magnetic memories will gain currency,

The commercialization of nanotechnology is in the early days and expected to expand considerably in the next few years. We are working on several variations of IGZO class of materials (different phases, composition,) and expect to see improvements within the next year, Sri Savamedam concluded.

Over the medium term, crystalline 2D materials bring electrical, optical, and magnetic properties that improve performance. 2D materials (like WS2, MoS2) as channels are important to extend the CMOS roadmap beyond Si, SiGe, and Ge channel, Sri Savamedam informed us.

Much work is needed to grow these materials at a high quality using 300mm tools. It will likely take 2-3 years before we have breakthroughs in 300mm tools for these materials.

The Internet of Things and Edge computing have increased the demand for miniaturized connected devices. For example, in the Covid-19 context, microfluidics MEMS (Micro-Electro-Mechanical Systems) devices are capable of point-of-care diagnosis to detect the virus without delay sans a visit to a lab. These devices are particularly useful for repeated testing that the chronically ill need. In the Covid-19 context, they are also useful for rapid testing. (Read also: 5 Top Health Tech Trends.)

Existing methods of testing are overly expensive, slow, and unable to scale for pandemic situations. The golden standard of diagnosis is a molecular analysis (detecting viral RNA) by either PCR or next-generation sequencing. These techniques are sensitive (able to detect low quantities of the virus), and specific (the test results will not be affected by confounding factors such as the presence of other viruses). Current methods of molecular analysis need expensive infrastructure (lab, personnel, and equipment) to perform the tests, Peter Peumans, CTO Health Care Solutions, imec, told us in response to email questions.

The success of a molecular test depends critically on the pre-analytical steps (such as sampling and transport). Nanotechnology has the potential to address both challenges by bringing molecular tests to the point-of-need by making them more compact, easier to use, faster, and cheaper.

One such innovation is the company miDiagnostics, which was spun out of imec in 2015. miDiagnostics is using imecs silicon chip know-how to test at the point-of-need, Peumans informed us.

Cepheid is another such device in the Covid world, which binds specific nanoparticles like graphene to test for pathogens at the DNA level. Due to pinpointed targeting with nanoparticles, tests are far more specific and sensitive. By contrast, currently, most lab results are rife with false positives and false negatives besides taking about a week to complete.

Brain-machine interface is another application that is getting a boost from nanotechnology. Due to the nanosensors tiny size, the number of them used for picking brain signals increases by double-digit multiples, Gene Locklear, Senior Artificial Intelligence Research Scientist at Entrust Solutions, told us. He went on:

The processing of this many signals with AI is a forbidding challenge. Techniques from emergent learning are used to aggregate data before it can be analyzed. The methods are akin to flocks of birds that fly and return in concert. They have implicit rules by which the subordinate birds take cues from the lead bird without explicit guidance.

Reinforcement learning is a key tool which can model for the collective behavior and continuously learn from the data to achieve a predefined objective. When out of the lab, human users confirm the validity of the results, he added.

We spoke to Dr. Matthew Putman, Co-founder & CEO of Nanotronics, who worked on nanodevices in a university laboratory, before launching his venture ten years ago, to commercialize and scale them on an industrial scale. Dr. Mathew Putman told us:

Rapid advances in machine learning, deep learning, and reinforcement learning, automation, and imaging enable us to spot and correct errors to achieve quality standards for scalable and feasible nanotechnology.

For example, the traditional semiconductor lithography technologies are subtractive, i.e., they cut pieces from a larger slab of material. By contrast, nanotechnology is additive,i.e., it builds from bottom-up by joining nanomaterials to make a larger piece. AI and imaging technologies help to join nanomaterials precisely to achieve quality standards, Dr. Putnam explained. (Read also: Nanotechnology: The Biggest Little Innovation in Technology.)

In the past, the means for checking quality were limited to an atomic microscope or electron microscopy. You could see only an incredibly tiny part, a single nanoparticle, of a nano product or nano equipment. It was analogous to finding a needle in a haystack but not the larger formation or the product as a whole, Dr. Putnam recalled.

AI algorithms like reinforcement learning define the objective function or the properties for a product like graphene. One piece of the puzzle is the observed needle or the first nanoparticle. Accrued machine knowledge helps determine the remaining nanoparticles to produce graphene and align them precisely to achieve the desired quality. We reserve hardware for manual correction of only pinpointed errors, Dr. Putnam elucidated.

The precision that AI brings to the process is critical not only for manufacturing but also for the solutions possible with nanotechnology. The specificity and sensitivity of DNA level diagnosis of a disease like Coronavirus are contingent on its accurate sequencing, and the pathogen is identified right. Graphene, a nano product, binds with the DNA at the fault lines between nanoparticles for diagnosis. We feed the objective functionthe DNA sequenceof a pathogen that the algorithm verifies, Dr. Putnam concluded.

Machine intelligence does not preclude human creativityimaging technologies only augment it. Individuals and teams visualize nanoparticles networks, otherwise beyond their intuitive capabilities.

Teams collaborate based on the situational awareness gained from the visuals to solve problems machines cannot. Imaging provides a macro-image (which is itself micro in nanotechnology) but not molecular. Due to the limitations of physics, you cannot capture images of particles any smaller than that. Now there are computational tricks to construct product formation at a molecular level. You can now classify topologies previously undiscovered when viewed with a microscope or the human eye, Dr. Putnam informed us.

Nanotronics uses virtual reality and augmented reality to help their staff select the right algorithmic models and recalibrate them when the expected results fall short. Dr. Putnam observed:

We use 3D technologies to overcome human intuitions limitations and immerse in data otherwise hard to grasp and manipulate. Virtual reality transports them to an alternate world where they see the same imaging and alter it to visualize alternative scenarios. Augmented Reality helps direct the staff to move through this second life type world to make deft moves. Concurrently, we train the AI system as we put together the data and visualize it.

While the early viable uses of nanotechnology, in the semiconductor industry, are now proven, and lately, point-of-care diagnostics have shown promise, further progress hinges on much greater precision in the operating technologies. The room for error in the more advanced applications like DNA Origami or molecular scaffolding protein and catalytic engines for drug discovery is non-existent, so their viability is in the far horizon, Dr. Putnam surmised based on Nanotronics experience across several industries.

The near-term growth of uses has some prominent industries. There have been successes, such as sensors for smartphones. Nano is working for neuromorphic chips including sometimes for photonics, and genomic sequencing, RUV, IG Stack and new types of MEMS devices and sensors are candidates for near-term growth, Dr. Putnam concluded.

The complexity and the invisibility of nanotechnologys structure inhibited its growth. By removing these barriers with artificial intelligence and imaging, a new era of precision manufacturing and usage is poised for exponential growth driven by industries created with its help.

Follow this link:
Viable Use Nanotechnology: The Future Has Arrived - Techopedia