(Image by Garik Barseghyan from Pixabay )

The quantum tipping point that fabled moment when quantum technologies break through to commercial adoption at scale has beenquestioned in a previous diginomica report.

According to speakers at Davos this year, a more likely scenario is the gradual emergence of vertical use cases in which a quantum advantage finding hidden correlations between data points, for example will present itself. Plus, algorithms that allow the querying of massive data sets more quickly than classical models.

Even so, investors of every size are beginning to back quantum technologies, especially those that relate to enterprise data. And they are doing so despite the lack of strategic, long-term thinking among potential customers.

According to McKinsey, quantum investment exceeded $1.4 billion in 2021, while the UK government launched a 10-year, 2.5 billion ($3.2 billion) funding drive in March.

One of the challenges, says David Magerman, a founding partner at New York City-based investment house Differential Ventures, is that in a world of AI hype and a rush to associate with popular technologies like GPT, some enterprise users prefer a quick hit of tactical advantage to a long bet with an uncertain payback date.

A sugar rush trumps healthy eating for the future, perhaps? Magerman says:

I come from the financial industry. And while there are a lot of problems that quantum computing will solve significantly better [than classical systems], the financial industry is very much about the next five minutes.

All the spending that goes on research in finance is about solving problems today. There's such a competitive environment that they can't afford to be too forward-looking. We can't get the financial industry to devote many resources to quantum until production solutions are imminent. Then theyll spend a lot of money playing catch-up.

But every dollar they spend on forward-looking research gives their competitors a chance to beat them today.

A frustrating problem for tech innovators who need financial backing. One that applies across many sectors, though in areas such as cybersecurity and communications, where quantum-safe technologies are vital, and industries such as materials science and pharmaceuticals, early investment is a more immediate priority.

Yet being first out of the (quantum) gate is what drives Differential Ventures, a seed fund that backs promising early-stage companies in quantum, plus AI, machine learning, and data science. That said, seed-investment in the US can mean writing cheques for a few million dollars, as well as smaller, less risky sums.

Hotspots include bridging the gap between quantum technologies and classical environments, says Magerman. For example, his company recently led a $6.1 million seed extension round in Agnostiq, the Toronto-based distributed computing start-up that is building Covalent, an enterprise quantum and high-performance computing platform.

His company is also backing more personal artificial intelligence (AIs that query personally identifiable data, or PID), and their necessary flipside: private AIs, which keep PID safe from acquisitive AIs online. That certainly keeps the options open!

He explains the Differential philosophy:

A lot of people promote fantastical solutions that are oversold and aren't realistic. We'll avoid those companies. And those that are pitching things that aren't that hard, even if they're interesting applications. Once those ideas hit the mainstream, they are hard to protect from the competition, so we avoid those investments too.

Really, its based on my experience as a data scientist.

So, how common is demand for seed funding in quantum computing, an area often based in research labs and university spinouts, where high-risk, long-term, big-ticket VC backing would seem to be the order of the day? Massive funds that can afford to have a speculative under-performer in their portfolio when the other wins might be massive?

He says:

We typically avoid hardware companies, because they tend to require much more capital. And not only to start out: a quantum hardware company might raise a few million dollars in their first investment round, but they're going to require probably hundreds of millions to get to product.

But there are a lot of quantum software companies that typically want the seed rounds we look to invest in. They can get to product and get to revenue at scale, which means we can continue to invest in them before they get too big for us.

For example, Agnostiq recognized early on that the big problem impeding the growth of quantum research was orchestration tools that could integrate classical computing workloads with quantum computing resources. So, they built a product that allows companies to do high-performance computing orchestration.

So, does Magerman believe that a tipping point is approaching, or is it more a case of softly, softly? Because on the face of it, a $42 million seed fund (at present) could not afford to wait too long for commercial payback in such an uncertain space?

He explains:

I wouldn't say it's fast approaching, but I would say it's inevitable. And I think it will approach more quickly based on more intellectual and financial resources being devoted to solving the problem.

The subtext is the limits of Moores Law being reached, at which point quantum may offer a promising way forward: a new upward slope of increasing speed and power, assuming the challenge of noise in quantum circuits can be, if not solved, then at least made workable at scale.

He says:

The end of Moore's Law, plus the supply-chain issues that we had with silicon during COVID are among the factors contributing to the increased attention on quantum, accelerating the solutions to problems that prevent it from scaling.

For so long there has been this belief that, if you can continue to scale and get more power out of classical computing chips by engineering them with more densely packed circuitry, making them faster, and simply building more of them there is less pressure to develop alternative architectures and hardware.

But now we're seeing that we're becoming more limited in computing power [in the near future] yet we have more and more data, and more and more computing needs, especially with AI there'll be a lot more pressure to accelerate the development of quantum hardware.

There are still unsolved problems it is still too error prone and difficult to scale but these are practical problems, not theoretical ones. I wouldn't bet against the ingenuity of humans, especially when there's economic forces pushing for a solution!

So, the voices saying a tipping point wont happen were wrong?

He says:

I think the tipping point is more about coming up with the killer algorithm that is currently intractable. One that would take, effectively, infinite computing time in a classical environment, but which quantum computing could solve in a reasonable amount of time.

So, the two questions are: one, can you actually implement these algorithms in a quantum computing environment? Can you solve that problem? And, two, does the precise solution to the optimization perform substantially better, in practical problems, then the heuristic approximation?

Because it could be that the heuristic approximation is good enough and that getting the exact mathematical solution doesn't give you meaningfully better real-world performance.

So, the value of quantum will be finding the real-world problem where the heuristic solutions we are using today are inadequate and the quantum solution is substantially better.

So, what should enterprise decision-makers do in this push me, pull you environment of conflicting messages, not to mention the tactical pressure to innovate now, rather than place longer-term bets?

He says:

Theres a big education process that needs to happen, and that's going to need both a bottom-up and a top-down approach.

First, you have to get quantum researchers inside these corporate research groups, and give them the opportunity to experiment with solutions and promote those up the food chain to senior management.

And the top-down is getting management to make the hiring decisions, to bring in people who understand quantum technology well enough to do good research and prepare the company for quantum at scale.

Wise words from a brave investor who is prepared to be there in the early days of new ventures, even as the big guns of IBM, Google, Microsoft, et al, gear up in the background.

Read the original:
Quantum commercialization softly, softly towards the inevitable future, says backer - Diginomica

04.26.23

***WATCH: Senator Murrays remarks and questioning***

Washington, D.C. Today, at a Senate Appropriations Commerce, Justice, Science, and Related Agencies subcommittee hearing on President Bidens FY24 budget request for the Department of Commerce, Senator Patty Murray (D-WA), Chair of the Senate Appropriations Committee, spoke about the need to build on investments that keep the U.S economy competitive, support innovation, and strengthen trade partnerships.

When we compete with our adversaries its not just a matter of who spends the most on defense, said Senator Murray. Its a matter of who has the strongest supply chains, who supports workers and growing businesses most, who makes sure our communities have the high-speed internet they need to do just about anything in the 21st Century, and who is on the cutting edge of technologyadvanced manufacturing, green energy, quantum computing, and more.

In particular, Senator Murray underscored just how critical it would be that the Biden administration have the resources necessary to effectively implement the bipartisan CHIPS and Science Act and how the investments of the Inflation Reduction Act would also help build a stronger clean energy economy and boost American manufacturing.

The Department is also using resources we provided in the CHIPS and Science Act to bring manufacturing jobs in things like quantum computing and chip manufacturing to states like minebecause weve got to make sure the jobs of the future are here in Americanot shipped overseas, stated Senator Murray. And the Inflation Reduction Act is actually ushering in a boom in green manufacturing for batteries, and electric vehicles, and more.

We know China is making investments in many of these same industries and trying to get the upper hand. So if we want to not just compete, but to lead the world, we cannot let our investments fall behind or fall victim to partisan pot shots, Senator Murray continued.

In her questioning, Senator Murray asked Secretary of Commerce Gina Raimondo how the National Oceanic and Atmospheric Administration (NOAA) plans to continue aiding salmon recovery in Washington state and protect important ecosystems in the Pacific Northwest from invasive species like European green crabs. Senator Murray and Congressman Derek Kilmer recently joined tribal leaders, environmental advocates, and experts to trap European green crabs and discuss the threat they pose to Washingtons states native species.

###

Continue reading here:
Majority News Release | Majority News Releases | News | United ... - United States Senate Committee on Appropriations

This article has been reviewed according to ScienceX's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

trusted source

proofread

Flowchart and schematic of a probabilistic (p-bit) stochastic nanomagnet network for scalable, energy-efficient computing of combinatorial optimization problems: a giant spin Hall effect (GSHE) layer converts input voltage into a spin current (top left); a free circular nanomagnet with low thermal barrier serves as the free layer (top right). Credit: Zhu, Xie, and Bermel, doi 10.1117/1.JPE.13.028501.

According to computational complexity theory, mathematical problems have different levels of difficulty in the context of their solvability. While a classical computer can solve some problems (P) in polynomial timei.e., the time required for solving P is a polynomial function of the input sizeit often fails to solve NP problems that scale exponentially with the problem size and thus cannot be solved in polynomial time. Classical computers based on semiconductor devices are, therefore, inadequate for solving sufficiently large NP problems.

In this regard, quantum computers are considered promising as they can perform a large number of operations in parallel. This, in turn, speeds up the NP problem-solving process. However, many physical implementations are highly sensitive to thermal fluctuations. As a result, quantum computers often demand stringent experimental conditions such extremely low temperatures for their implementation, making their fabrication complicated and expensive.

Fortunately, there is a lesser-known and as-yet underexplored alternative to quantum computing, known as probabilistic computing. Probabilistic computing utilizes what are called "stochastic nanodevices," whose operations rely on thermal fluctuations, to solve NP problems efficiently. Unlike in the case of quantum computers, thermal fluctuations facilitate problem solving in probabilistic computing. As a result, probabilistic computing is, in fact, easier to implement in real life.

Shedding much-needed light on this potential alternative, a group of researchers have now demonstrated the capabilities of probabilistic computing by simulating stochastic nanodevice networks to solve specific NP problems. The study, led by Professor Peter Bermel from Purdue University, is published in the Journal of Photonics for Energy (JPE).

The researchers used the "Ising model," a canonical model for simulating a wide variety of physical as well as mathematical problems. Originally devised to model the interactions of magnetic dipole moments of atomic spins, its energy operator, namely the "Hamiltonian," can also represent NP problems. Essentially, solving an NP problem amounts to solving the corresponding Ising Hamiltonian. Probabilistic computing devices made of networks of optical parametric oscillators (OPOs) and stochastic circular nanomagnets with low thermal barriers have been used to solve such problems.

The researchers implemented one such nanomagnet network using existing fabrication methods. They then used it to solve the Ising Hamiltonians of four NP-complete problems (problems with no efficient solution algorithm) from number theory associated with combinatorial optimization. These included number partitioning, exact cover, binary integer linear programming, and integer linear programming.

The simulation results of the first three problems with 3, 3, and 6 probabilistic bits (p-bits) strongly agreed with the theoretical solution (Boltzmann law) of the Ising model. The researchers observed a similar agreement between modeling and theory in the simulations of five different exact cover problems with 3, 6, 9, 12, and 15 p-bits. This indicated the potential for scaling up probabilistic computing frameworks.

According to Bermel, "in probabilistic computing, efficient scaling with problem size is the key to make it a robust, relevant alternative to classical computing techniques. Both modeling and experiments will be needed to confirm which approaches are most promising."

While the simulation results reported show robust results for all p-bits (from 3 to 15), the researchers suggest that parallel algorithms could help further scale up the simulation capability. In cases where parallelism is not feasible, transitioning from nanomagnet to OPO networks could facilitate efficient problem solving. The system can be mapped on an OPO network and easily implemented with current fabrication techniques, such as the CMOS technology. This, in turn, can lead to the development of stochastic nanomagnets with low energy barriers for probabilistic computing.

JPE Editor-in-Chief Sean Shaheen of University Colorado Boulder remarks, "Developing unconventional forms of computing hardware is becoming increasingly important as AI and scientific/enterprise computing accelerate in scale, at a pace that brings significantif not urgentconcerns about their energy consumption and carbon footprint."

"This work by Zhu, Xi, and Bermel provides a practical pathway to a hardware platform that solves an important class of NP-complete problems. By creatively harnessing networks of nonlinear optical devices to carry out Ising computing, the work demonstrates a scalable, energy-efficient approach that has the potential to vastly outperform conventional hardware in solving computationally complex problems."

More information: Jie Zhu et al, Numerical simulation of probabilistic computing to NP-complete number theory problems, Journal of Photonics for Energy (2023). DOI: 10.1117/1.JPE.13.028501

Excerpt from:
Solving computationally complex problems with probabilistic computing - Tech Xplore

In Hindi, the word for revolution is Kranti. This kranti for India has a larger meaning. KRANTI is Knowledge, Research, and New Technology in India.

The worlds largest free market democracy has become the strongest adopter and creator of emerging technologies. More than a billion Indian citizens and consumers are getting direct benefits as they leapfrog over several legacy challenges.Since 2014, the fourth industrial revolution has swept the country where a billion plus people are getting last mile delivery of products and services both from the government and an entrepreneurial private sector. This scale of last mile delivery using tech platforms has no precedent in the history of democracy in the world. Technologies like AI, blockchain, drones, IoT and 3D manufacturing are changing Indias economic landscape.A few statistics here would help. Food delivery apps have crossed 2 million food deliveries in a day. Indias digital payments infrastructure United Payments Interface (UPI) crossed 8.7 billion transactions in March. UPI allows various money-financial transaction applications and is important for those who have not been part of the formal banking system. Over $4.5 billion of welfare support has been given to 160 million citizens with transparency. Despite the odds, over 115 tech based unicorn companies have a combined valuation of over $350 billion.All this has been made possible by a technological revolution which has taken root in India. In Hindi, the word for revolution is Kranti. It has been associated mainly with Indias fight against British occupation, but now it reflects the deep and widening spread of the Fourth Industrial Revolution. This kranti for India has a larger meaning. KRANTI is Knowledge, Research, and New Technology in India. Indian industry is now being incentivized to invest in research and knowledge.In recent years, India has begun to innovate and technologize to solve the unique challenges of the country. The fourth industrial revolution democratized access to many emerging technologies. Indian innovators seized this opportunity to create solutions which could help domestic businesses to improve their products and services rapidly. Artificial intelligence, internet of things, drones, virtual and augmented reality and automation brought efficiency and transparency to businesses. The charge of the digital brigade has allowed India to claim success as a tech creator. The number of tech patents in India have seen a steady rise. Industry body Nasscom says that Indian companies have filed 1,38,000 tech patents in India and over 9,500 patents in the US during 2015-2021. More than 60% of these patents were filed by Indian companies and start-ups, while 17% of the tech patents were filed by individual inventors/academia research outfits. These patented technologies have been deployed in various sectors such as finance, health and logistics. Reports suggest that India already has 127 patents for 6G technology which will underpin Web 3 and metaverse technologies.Much of the rise has been made possible by public private collaboration and the Jandhan Aadhar and Mobile trinity model created by the Narendra Modi government. New models have allowed last mile delivery of products and services. While the private sector has innovated for consumers and government backed projects brought welfare schemes and financial inclusion for citizens. A Niti Aayog report says that digital platform based gig economy will give livelihood to 23.5 million workers by 2029-30 from just 7.7 million today.Cutting edge technologies are getting attention too. The government has approved the setting up of a National Quantum Mission with a funding of over $730 million. India joins the league of countries which are investing in emerging technologies which will shape business and innovation. Quantum technologies include applications like computing, communication, simulation and sensing. These can be used in sectors like materials science, energy, drug discovery and defence.While quantum computing has been around for decades, the last few months have seen tremendous interest from investors and government. Quantum computing uses principles of physics where sub-atomic Qubits can store and process information in a way that traditional computers cant. The scale and speed of quantum computing is far ahead of what even a supercomputer can achieve. Some tasks that a classical computer may take a week to complete can be done in just one second with quantum computing.

AGRICULTUREIndia has brought emerging technologies to the farm sector. A recent report by World Economic Forums (WEF) Centre for Fourth Industrial Revolution captures this trend. Using Technology to Improve a Billion Livelihoods report that the use of drones in agriculture can create a $100 billion GDP boost and help millions get better livelihoods. Drones are poised to be an effective tool to support farmers reduce their operating costs and efforts, while at the same time optimizing their input use. There are multiple uses for drones, including surveying, seeding, spraying, pollination, etc. that are at different stages of technology and business model maturity, the report says.There are several areas where drones are making their presence felt. These include seed propagation; pesticide spraying; yield prediction; land records; insurance assessment; and crop monitoring.Several policies of the government are now promoting the manufacture and use of drones not just in agri but other sectors. A mission on agri mechanization will support the use of drones, while the Drone Shakti programme will promote drone-as-a-service across the country. The PLI and export promotion incentives are expected to bring $50 million of investment in the drone production.

METAVERSEIn India companies like Tech Mahindra and Infosys have launched industrial metaverse solutions for domestic and global businesses. Tech Mahindra is geared up to harness the power of Metaverse to bring imagination to life and offer futuristic experiences now. Network of the future, 5G, will form the backbone to provide enhanced experience to customers in the metaverse universe, the company says.With the roll-out of 5G in India, metaverse applications will be accelerated across the industrial spectrum. High speed connectivity and low latency is critical for any real-time application and usage. Companies using metaverse-based solutions will also have to change their internal business and manufacturing processes. If many of the design and quality check issues are avoided with digital design, many physical steps may become redundant.

BHASHINIIndias efforts on bringing the digital revolution to languages is being recognized and lauded. The Bhashini project led by the Ministry of Electronics and IT aims to build a National Public Digital Platform for languages to develop services and products for citizens by leveraging the power of artificial intelligence and other emerging technologies. As millions of Indians join the digital mainstream, they will need information and services in their own languages. Companies that want to deliver their products and services to Indias growing consuming class understand the need to use local languages. The Bhashini project is bringing together industry, academia and innovators to create an open repository. Translation applications will be created between several languages. From Tamil to Assamese. Or from Punjabi to Kannada.Institutions and individuals can contribute in the four pillars of language understanding: Suno (to listen); Bolo (to speak); Likho (to write); and Dekho (to read). Since Bhashini is a public project, everyone can contribute but no one company can have a monopoly to use or license. The open source framework will help it grow with crowd contributions

DIGI YATRA, DIGILOCKER, DIGI EVERYTHINGMore than 160 million citizens of India have now signed up and loaded important documents and certificates on the Digilocker platform. Seeking, saving and managing important documents of identity, tax and assets in a digital format has freed citizens from the tyranny of paper files. Launched by the Ministry of Electronics and Technology in 2016, the Digilocker platform had a slow start. Till 2020 the number of registered users on the platform were about 38 million. The number of users more than doubled in just one year to over 91 million by 2021. These users have stored and saved over 5.5 billion documents on the platform.In another first, several airports in India are now accepting face recognition based DigiYatra app for flight passengers. Users can link their identity cards saved on Digilocker to the Digiyatra app.The Digital India mission has become all pervasive and is spawning solutions for 1.4 billion people. Cutting edge technologies to low tech solutions applied is scale is bringing efficiency, transparency and affordability in the services they need. This is the new Kranti of India.

*Pranjal Sharma is the author of India Automated and Kranti Nation.

Read the original here:
India embraces the technology Kranti - The Sunday Guardian Live - The Sunday Guardian

RSA Conference Adi Shamir, the cryptographer whose surname is the "S" in "RSA", thinks folks need to stop worrying about quantum computing breaking encryption algorithms.

Speaking on the annual cryptographers' panel at the RSA Conference in San Francisco this week, he opined that in the 1990s he saw three big issues appear on the security industry's radar: AI, cryptography, and quantum computing. Two out of three had delivered, he said, and quantum computing has yet to show promise and won't for decades to come.

99 percent of encrypted messages are junk, he opined. Requests for lunch meetings or banal chat; waste of time to decrypt, and there's so much of it.

The idea that such missives would be a top cracking priority isn't realistic, he reminded the audience. And while important messages might be decoded decades on, the signal-to-noise ratio is going to make throwing a quantum machine at the job a poor way to find real secrets.

He wasn't alone in his skepticism. British mathematician Cliff Cocks, who developed public-key cryptography years before session host Dr Whitfield Diffie and his colleagues came up with the same idea, was somewhat cutting about stories that the Chinese have developed quantum systems to crack current encryption systems.

The Chinese system may work well on very small data sets, he opined, but there's "no evidence whatsoever" that it would work on a larger scale. That said, Anne Dames, IBM zSystems Distinguished Engineer and Cryptographic Technology Architect, argued China's efforts are as good a reason as any to update your public-private keys just to be on the safe side. The longer and more secure the keys the better she opined. There's no harm in using quantum-resistant algorithms, either, we note.

The RSA cryptographer's panel in San Francisco today

"Quantum computers, even if they don't exist today, will do in the next 30-40 years, so we will need to switch keys," she advised, saying the current concerns over quantum cryptography reminded her a lot of blockchain hype.

That said, all the encryption in the world isn't going to help you defend against insider threats. It's been ten years since an IT contractor called Edward Snowden managed to walk off with the NSA's crown jewels, and the latest Pentagon leak is alleged to have involved a guy showing off classified information on Discord to impress friends. This showed the systems we use are still critically weak, Diffie argued.

Shamir argued Snowden was a short-term and long-term disaster for the NSA, and diminished America's influence by exposing directly long-suspected practices - such as the presence of backdoors in commercial products - for which no evidence had previously been available. Quantum computers breaking encryption could deliver similar revelations, Shamir opined, but it's a way off doing so.

Continue reading here:
You can cross 'Quantum computers to smash crypto' off your list of existential fears for 30 years - The Register