The TV series Devs took as its premise the idea that a quantum computer of sufficient power could simulate the world so completely that it could project events accurately back into the distant past (the Crucifixion or prehistory) and predict the future. At face value somewhat absurd, the scenario supplied a framework on which to hang questions about determinism and free will (and less happily, the Many Worlds interpretation of quantum mechanics).

Quite what quantum computers will do for molecular simulations remains to be seen, but the excitement about them shouldnt eclipse the staggering advances still being made in classical simulation. Full ab initio quantum-chemical calculations are very computationally expensive even with the inevitable approximations they entail, so it has been challenging to bring this degree of precision to traditional molecular dynamics, where molecular interactions are still typically described by classical potentials. Even simulating pure water, where accurate modelling of hydrogen bonding and the ionic disassociation of molecules involves quantum effects, has been tough.

Now a team that includes Linfeng Zhang and Roberto Car of Princeton University, US, has conducted ab initio molecular dynamics simulations for up to 100 million atoms, probing timescales up to a few nanoseconds.1 Sure, its a long way from the Devs fantasy of an exact replica of reality. But it suggests that simulations with quantum precision are reaching the stage where we can talk not in terms of handfuls of molecules but of bulk matter.

How do they do it? The trick, which researchers have been exploring for several years now, is to replace quantum-chemical calculations with machine learning (ML). The general strategy of ML is that an algorithm learns to solve a complex problem by being trained with many examples for which the answers are already known, from which it deduces the general shape of solutions in some high-dimensional space. It then uses that shape to interpolate for examples that it hasnt seen before. The familiar example is image interpretation: the ML system works out what to look for in photos of cats, so that it can then spot which new images have cats in them. It can work remarkably well so long as it is not presented with cases that lie far outside the bounds of the training set.

The approach is being widely used in molecular and materials science, for example to predict crystal structures from elemental composition,2-3 or electronic structure from crystal structure.4-5 In the latter case, bulk electronic properties such as band gaps have traditionally been calculated using density functional theory (DFT), an approximate way to solve the quantum-mechanical equations of many-body systems. Here the spatial distribution of electron density is computationally iterated from some initial guess until it fits the equations in a self-consistent way. But its computationally intensive, and ML circumvents the calculations by figuring out from known cases what kind of electron distribution a given configuration of atoms will have.

The approach can in principle be used for molecular dynamics by recalculating the electron densities at each time step. Zhang and colleagues have now shown how far this idea can be pushed using supercomputing technology, clever algorithms, and state-of-the-art artificial intelligence.6 They present results for simulations of up to 113 million atoms for the test case of a block of copper atoms, enabling something approaching a prediction of bulk-like mechanical behaviour from quantum chemistry. Their simulations of liquid water, meanwhile, contain up to 12.6 million atoms.

For small systems where the comparison to full quantum DFT calculations can be made, the researchers find electron distributions essentially indistinguishable from the full calculations, while gaining 45 orders of magnitude in speed. Their system can capture the full phase diagram of water over a wide range of temperature and pressure, and can simulate processes such as ice nucleation. In some situations water can be coarse-grained such that hydrogen bonding can still be modelled without including the hydrogen atoms explicitly.7 The researchers say it should be possible soon to follow such processes on timescales approaching microseconds for about a million water molecules, enabling them to look at processes such as droplet and ice formation in the atmosphere.

For small systems where the comparison to full quantum DFT calculations can be made, the researchers find electron distributions essentially indistinguishable from the full calculations, while gaining 45 orders of magnitude in speed. Their system can capture the full phase diagram of water over a wide range of temperature and pressure, and can simulate processes such as ice nucleation. The researchers say it should be possible soon to follow such processes on timescales approaching microseconds for about a million water molecules, enabling them to look at processes such as droplet and ice formation in the atmosphere.

Both of these test cases are helped by being relatively homogeneous, involving largely identical atoms or molecules. Still, the prospects of this deep-learning approach look good for studying much more heterogeneous systems such as complex alloys.8 One very attractive goal is, of course, biomolecular systems, where the ability to model fully solvated proteins, membranes and other cell components could help us understand complex mesoscale cell processes and predict the behaviour of drug candidates. One challenge here is how to include long-range interactions such as electrostatic forces.

Its a long way from Devs-style simulations of minds and histories, which will perhaps only ever be fantasies. But one scene in that series showed what might be a more tractable goal: the simulation of a growing snowflake. What a wonderful way that would be to advertise the simulators art.

1. Jia et al., arXiv, 2020 http://www.arxiv.org/abs/2005.00223 (submitted, ACM, New York, 2020)

2 C C Fischer et al, Nat. Mater., 2006, 5, 641 (DOI:10.1038/nmat1691)

3 N Mounet et al, Nat. Nanotechnol., 2018, 13, 246 (DOI: 10.1038/s41565-017-0035-5)

4 Y Dong et al, npj Comput. Mater., 2019, 5, 26 (DOI:10.1038/s41524-019-0165-4)

5 A Chandrasekaran et al, npj Comput. Mater., 2019, 5, 22 (DOI:10.1038/s41524-019-0162-7)

6 L Zhang et al, Phys. Rev. Lett., 2018, 120, 143001 (DOI:10.1103/PhysRevLett.120.143001)

7. L Zhang et al, J. Chem. Phys., 2018, 149, 034101 (DOI:10.1063/1.5027645)

8. F-Z Dai et al., J. Mater. Sci. Technol., 2020, 43, 168 (DOI:10.1016/j.jmst.2020.01.005)

Go here to see the original:
Molecular dynamics used to simulate 100 million atoms | Opinion - Chemistry World

Research team: (from left) Associate Professor Stefano Palomba, Dr Alessandro Tuniz, Professor Martijn de Sterke. Photo: Louise Cooper

Light is emerging as the leading vehicle for information processing in computers and telecommunications as our need for energy efficiency and bandwidth increases.

Already the gold standard for intercontinental communication through fibre-optics, photons are replacing electrons as the main carriers of information throughout optical networks and into the very heart of computers themselves.

However, there remain substantial engineering barriers to complete this transformation. Industry-standard silicon circuits that support light are more than an order of magnitude larger than modern electronic transistors. One solution is to compress light using metallic waveguides however this would not only require a new manufacturing infrastructure, but also the way light interacts with metals on chips means that photonic information is easily lost.

Now scientists in Australia and Germany have developed a modular method to design nanoscale devices to help overcome these problems, combining the best of traditional chip design with photonic architecture in a hybrid structure. Their research is published today in Nature Communications.

We have built a bridge between industry-standard silicon photonic systems and the metal-based waveguides that can be made 100 times smaller while retaining efficiency, said lead author Dr Alessandro Tuniz from the University of Sydney Nano Institute and School of Physics.

This hybrid approach allows the manipulation of light at the nanoscale, measured in billionths of a metre. The scientists have shown that they can achieve data manipulation at 100 times smaller than the wavelength of light carrying the information.

This sort of efficiency and miniaturisation will be essential in transforming computer processing to be based on light. It will also be very useful in the development of quantum-optical information systems, a promising platform for future quantum computers, said Associate Professor Stefano Palomba, a co-author from the University of Sydney and Nanophotonics Leader at Sydney Nano.

Eventually we expect photonic information will migrate to the CPU, the heart of any modern computer. Such a vision has already been mapped out by IBM.

On-chip nanometre-scale devices that use metals (known as plasmonic devices) allow for functionality that no conventional photonic device allows. Most notably, they efficiently compress light down to a few billionths of a metre and thus achieve hugely enhanced, interference-free, light-to-matter interactions.

As well as revolutionising general processing, this is very useful for specialised scientific processes such as nano-spectroscopy, atomic-scale sensing and nanoscale detectors, said Dr Tuniz also from the Sydney Institute of Photonics and Optical Science.

However, their universal functionality was hampered by a reliance on ad hoc designs.

We have shown that two separate designs can be joined together to enhance a run-of-the-mill chip that previously did nothing special, Dr Tuniz said.

This modular approach allows for rapid rotation of light polarisation in the chip and,becauseof that rotation, quickly permits nano-focusing down to about 100 times less than the wavelength.

Professor Martijn de Sterke is Director of the Institute of Photonics and Optical Science at the University of Sydney. He said: The future of information processing is likely to involve photons using metals that allow us to compress light to the nanoscale and integrate these designs into conventional silicon photonics.

This research was supported by the University of Sydney Fellowship Scheme, the German Research Foundation (DFG) under Germanys Excellence Strategy EXC-2123/1. This work was performed in part at the NSW node of the Australian National Fabrication Facility (ANFF).

Read this article:
Light, fantastic: the path ahead for faster, smaller computer processors - News - The University of Sydney

Data science has become the hottest job of the century, but that does not mean it is for all. Not everyone can become data scientists as it requires continuous learning. Every other day, we witness innovation in the domain, thereby keeping up with the pace is demanding. Consequently, you can aspire to other rising technologies that will democratise rapidly in the coming years. Today, due to the lockdown, one has ample time to explore innovative technologies that will become popular as we move ahead.

Here is the list of technologies that you can explore:-

A combination of maths, physics and computer science, quantum computing has started seeing the light of the sun. Due to the rise of the data science landscape, the need for processing a colossal amount of data is increasing as organisations are rapidly collecting data. However, due to the dearth of required computational power, firms are struggling to process gathered information, thereby creating data silos. And since Moores law is starting to fail, we have to move away from dated hardware methodologies.

In 2019, Google claimed supremacy in quantum computing by benchmarking computation in 200 seconds, which would have taken the current worlds fastest computer 10,000 years. The technology proposed in the 1980s by Paul Benioff took decades for researchers to make a breakthrough. Now that everyone sees value in quantum computing, it is poised to rise in the coming years. Getting your hand on such technology during its nascent stage can open up new opportunities in the future.

Start exploring by enrolling in this course for free.

Blockchain has been at the forefront of the rising technologies for quite some time. Earlier deemed as bitcoin and other cryptocurrencies, today, it is used in a wide range of use cases to simplify the authenticity while keeping the security in check. Undoubtedly, it will revolutionise the finance industry, the state-of-the-art blockchain will be widely used in digital identity, supply chain management, healthcare, among others. According to a report, global blockchain technology will expand to reach $176 billion by 2025 and exceed $3.1 trillion by 2030. Blockchain in the future might become the go-to technology for almost all enterprise operations. Consequently, one should pursue a career in blockchain and innovate to streamline business processes.

Blockchain course here.

Augmented reality and virtual reality has the potential to revolutionise the whole world by providing immersive experiences in almost every use case. As remote working is becoming the new normal, the demand for VR and AR has doubled in the last two months as per Hemanth Satyanarayan, chief executive Imaginate. Among other technologies, mixed reality would be the next big thing. And according to Digi-Capital AR/VR market will reach around $65 billion revenue by 2024. The COVID-19 crisis has made humans realise how important these technologies are, however, to democratise companies will have to overcome numerous barriers associated with the AR/VR landscape. The requirement of high-performance computation, intuitive 3D interface design, and more, have been the reason why VR and AR have witnessed a slower growth. But, companies like Facebook is showing promise in making breakthroughs in the field to enable mass adoption of the technology.

Get started with these AR and VR courses by Google.

In the data-driven world, people are concerned about their privacy, which has led governments to make stringent rules and regulations for privacy protection. The introduction of various data protection rules such as GDPR and PDP across the globe has forced organisations to rethink their cybersecurity practices. Although PDP is yet to be enforced on organisations, firms are already changing their cybersecurity operations to immediately comply once the bill gets final approval. Currently, the Indian cybersecurity industry is valued at $6.7 billion, and with the rising digital growth, we are witnessing a renewed interest in cybersecurity.

Learn more about cybersecurity with this free course.

comments

More here:
Technologies That You Can Explore Other Than Data Science During Lockdown - Analytics India Magazine

We will know when someone opens his office door

John Martinis, who had established Googles quantum hardware group in 2014, has cleaned out his office, put the cats out and left the building.

Martinis says a few months after he got Googles now legendary quantum computing experiment to go he was reassigned from a leadership position to an advisory one.

Martinis told Wired that the change led to disagreements with Hartmut Neven, the longtime leader of Googles quantum project.

Martinis said he had to go because his professional goal is for someone to build a quantum computer.

Google has not disputed this account, and says that the company is grateful for Martinis contributions and that Neven continues to head the companys quantum project.

Martinis retains his position as a professor at the UC Santa Barbara, which he held throughout his tenure at Google, and says he will continue to work on quantum computing.

To be fair, Googles quantum computing project was founded by Neven, who pioneered Googles image search technology, and got enough cats together.

The project took on greater scale and ambition when Martinis joined in 2014 to establish Googles quantum hardware lab in Santa Barbara, bringing along several members of his university research group. His nearby lab at UC Santa Barbara had produced some of the most prominent work in the field over the past 20 years, helping to demonstrate the potential of using superconducting circuits to build qubits, the building blocks of quantum computers.

Googles ground-breaking supremacy experiment used 53 qubits working together. They took minutes to crunch through a carefully chosen math problem the company calculated would take a supercomputer 10,000 years to work out. It still does not have a practical use, and the cats were said to be bored with the whole thing.

Continue reading here:
Google's top quantum computing brain may or may not have quit - Fudzilla

Understanding advanced encryption standard on basic level doesnt require a higher degree in computer science or Matrix-level consciousness lets break AES encryption down into laymans terms

Hey, all. We know of security of information to be a hot topic since, well, forever. We entrust our personal and sensitive information to lots of major entities and still have problems with data breaches, data leaks, etc. Some of this happens because of security protocols in networking, or bad practices of authentication management but, really, there are many ways that data breaches can occur. However, the actual process of decrypting a ciphertext without a key is far more difficult. For that, we can thank the encrypting algorithms like the popular advanced encryption standard and the secure keys that scramble our data into indecipherable gibberish.

Lets look into how AES works and different applications for it. Well be getting a little into some Matrix-based math so, grab your red pills and see how far this rabbit hole goes.

Lets hash it out.

You may have heard of advanced encryption standard, or AES for short but may not know the answer to the question what is AES? Here are four things you need to know about AES:

The National Institute of Standards and Technology (NIST) established AES as an encryption standard nearly 20 years ago to replace the aging data encryption standard (DES). After all, AES encryption keys can go up to 256 bits, whereas DES stopped at just 56 bits. NIST could have chosen a cipher that offered greater security, but the tradeoff would have required greater overhead that wouldnt be practical. So, they went with one that had great all-around performance and security.

AESs results are so successful that many entities and agencies have approved it and utilize it for encrypting sensitive information. The National Security Agency (NSA), as well as other governmental bodies, utilize AES encryption and keys to protect classified or other sensitive information. Furthermore, AES is often included in commercial based products, including but limited to:

Although it wouldnt literally take forever, it would take far longer than any of our lifetimes to crack an AES 256-bit encryption key using modern computing technology. This is from a brute force standpoint, as in trying every combination until we hear the click/unlocking sound. Certain protections are put in place to prevent stuff from like this happening quickly, such as a limit on password attempts before a lockdown, which may or may not include a time lapse, to occur before trying again. When we are dealing with computation in milliseconds, waiting 20 minutes to try another five times would seriously add to the time taken to crack a key.

Just how long would it take? We are venturing into a thousand monkeys working on a thousand typewriters to write A Tale of Two Cities territory. The possible combinations for AES 256-bit encryption is 2256. Even if a computer can do multiple quadrillions of instructions per second, then we are still in that eagles-wings-eroding-Mount-Everest time frame.

Needless to say, its waaaaaaaaaaaaaaaaaaay (theres not enough memory on our computers to support the number of a letters that I want to convey) longer than our current universe has been in existence. And thats just for a 16-byte block of data. So, as you can see, brute forcing AES even if it is 128 bits AES is futile.

That would likely change, though, once quantum computing becomes a little more mainstream, available, and effective. Quantum computing is expected to break AES encryption and require other methods to protect our data but thats still a ways down the road.

Manage Digital Certificates like a Boss

14 Certificate Management Best Practices to keep your organization running, secure and fully-compliant.

To better understand what AES is, you need to understand how it works. But in order to see how the advanced encryption standard actually works, however, we first need to look at how this is set up and the rules concerning the process based on the users selection of encryption strength. Typically, when we discuss using higher bit levels of security, were looking at things that are more secure and more difficult to break or hack. While the data blocks are broken up into 128 bits, the key size have a few varying lengths: 128 bits, 196 bits, and 256 bits. What does this mean? Lets back it up for a second here.

We know that encryption typically deals in the scrambling of information into something unreadable and an associated key to decrypt the scramble. AES scramble procedures use four scrambling operations in rounds, meaning that it will perform the operations, and then repeat the process based off of the previous rounds results X number of times. Simplistically, if we put in X and get out Y, that would be one round. We would then put Y through the paces and get out Z for round 2. Rinse and repeat until we have completed the specified number of rounds.

The AES key size, specified above, will determine the number of rounds that the procedure will execute. For example:

As mentioned, each round has four operations.

So, youve arrived this far. Now, you may be asking: why, oh why, didnt I take the blue pill?

Before we get to the operational parts of advanced encryption standard, lets look at how the data is structured. What we mean is that the data that the operations are performed upon is not left-to-right sequential as we normally think of it. Its stacked in a 44 matrix of 128 bits (16 bytes) per block in an array thats known as a state. A state looks something like this:

So, if your message was blue pill or red, it would look something like this:

So, just to be clear, this is just a 16-byte block so, this means that every group of 16 bytes in a file are arranged in such a fashion. At this point, the systematic scramble begins through the application of each AES encryption operation.

As mentioned earlier, once we have our data arrangement, there are certain linked operations that will perform the scramble on each state. The purpose here is to convert the plaintext data into ciphertext through the use of a secret key.

The four types of AES operations as follows (note: well get into the order of the operations in the next section):

As mentioned earlier, the key size determines the number of rounds of scrambling that will be performed. AES encryption uses the Rjindael Key Schedule, which derives the subkeys from the main key to perform the Key Expansion.

The AddRoundKey operation takes the current state of the data and executes the XOR Boolean operation against the current round subkey. XOR means Exclusively Or, which will yield a result of true if the inputs differ (e.g. one input must be 1 and the other input must be 0 to be true). There will be a unique subkey per round, plus one more (which will run at the end).

The SubBytes operation, which stands for substitute bytes, will take the 16-byte block and run it through an S-Box (substitution box) to produce an alternate value. Simply put, the operation will take a value and then replace it by spitting out another value.

The actual S-Box operation is a complicated process, but just know that its nearly impossible to decipher with conventional computing. Coupled with the rest of AES operations, it will do its job to effectively scramble and obfuscate the source data. The S in the white box in the image above represents the complex lookup table for the S-Box.

The ShiftRows operation is a little more straightforward and is easier to understand. Based off the arrangement of the data, the idea of ShiftRows is to move the positions of the data in their respective rows with wrapping. Remember, the data is arranged in a stacked arrangement and not left to right like most of us are used to reading. The image provided helps to visualize this operation.

The first row goes unchanged. The second row shifts the bytes to the left by one position with row wrap around. The third row shifts the bytes one position beyond that, moving the byte to the left by a total of two positions with row wrap around. Likewise, this means that the fourth row shifts the bytes to the left by a total of three positions with row wrap around.

The MixColumns operation, in a nutshell, is a linear transformation of the columns of the dataset. It uses matrix multiplication and bitwise XOR addition to output the results. The column data, which can be represented as a 41 matrix, will be multiplied against a 44 matrix in a format called the Gallois field, and set as an inverse of input and output. That will look something like the following:

As you can see, there are four bytes in that are ran against a 44 matrix. In this case, matrix multiplication has each input byte affecting each output byte and, obviously, yields the same size.

Now that we have a decent understanding of the different operations utilized to scramble our data via AES encryption, we can look at the order in which these operations execute. It will be as such:

Note: The MixColumns operation is not in the final round. Without getting into the actual math of this, theres no additional benefit to performing this operation. In fact, doing so would simply make the decryption process a bit more taxing in terms of overhead.

If we consider the number of rounds and the operations per round that are involved, by the end of it, you should have a nice scrambled block. And that is only a 16-byte block. Consider how much information that equates to in the big picture. Its miniscule when compared to todays file/packet sizes! So, if each 16-byte block has seemingly no discernable pattern at least, any pattern that can be deciphered in a timely manner Id say AES has done its job.

We know the advanced encryption standard algorithm itself is quite effective, but its level of effectiveness depends on how its implemented. Unlike the brute force attacks mentioned above, effective attacks are typically launched on the implementation and not on the algorithm itself. This can be equated to attacking users as in phishing attacks versus attacking the technology behind the service/function that may be hard to breach. These can be considered side-channel attacks where the attacks are being carried out on other aspects of the entire process and not the focal point of the security implementation.

While I always advocate going with a reasonable/effective security option, a lot of AES encryption is happening without you even knowing it. Its locking down spots of the computing world that would otherwise be wide open. In other words, there would be many more opportunities for hackers to capture data if advanced encryption standard wasnt implemented at all. We just need to know how to identify the open holes and figure out how to plug them. Some may be able to use AES and others may need another protocol or process.

Appreciate the encryption implementations we have, use the best ones when needed, and happy scrutinizing!

Read more:
Advanced Encryption Standard (AES): What It Is and How It Works - Hashed Out by The SSL Store - Hashed Out by The SSL Store