Daily Archives: April 2, 2017

The nine startups participating in Singularity Universitys accelerator program presented thisafternoon atMoffett Federal Airfield just outside Mountain View, CA.Singularity University, founded in2008 by Peter Diamandis and Ray Kurzweil, aims to make it more feasible for people to address hard science problems and those that require a global reach.

Startups backed by Singularity University have gone on to raise $194 million. Collectively, Modern Meadow, Matternet and Getaround have raised $122 million of that total figureto bring people animal free leather, automated last-mile delivery and more convenient car rental.

Monique Giggy, Directorof Singularity Universitys startup accelerator says the program received 460 applications this year. From there 20 startups received formal diligence alongside technical experts. Ultimately only nine companies were selected for this years batch.

This is the third and final generalizedaccelerator program that Singularity will be offering. Participating companies in the program received mentorship and support from corporations, researchers and the broader Singularity network.

Going forward, Singularity wants to engage with portfolio companies across a longer span of time than the existing eight week program. Giggy explained that future programs could focus narrowly on the FDA approval process or raising a Series A.

The startups that spoketoday are addressing tough challenges in healthcare and materials science, among other things.Here are all nine companies that presented and the problems they are tackling.

IotaSecurity Security solution for mobile banking

IotaSecurity is building a solution to address the security vulnerabilities faced by users of mobile banking services. Development recently concluded on the teams primary SDK and its being marketed directly at banks.Yaron Vorona,who formally worked at a D.C. think tank, is working on the startup alongsideDr. Richard Krueger.Krueger previously worked to scale Amazon Prime.

Metamason 3D printed masks for sufferers ofsleep apnea

Metamason aims to serve the millions of Americans suffering from sleep apnea. The startup is developing technology that will allow sufferers to scan their face and have a 3D printed customized and fitted mask printed for them. Ideally, a better experience would incentivize more patients to stick with treatment instead of abandoning it.

Deep BlocksBringing intelligence to real estate developers

Deep Blocks is building a tool for real estate developers to aid them in the planning process. The long term vision is to deepen machine learning capabilities to be able to optimize decision making for all operators within the real estate vertical. The startupaggregates data across disparate sources, including critical regulatory information. Right now its available in Miami but plans to expand to other locals in the near future.

Nanobinoids Using hemp to create better batteries

Nanobinoids is using hemp to unlock thepromise ofgraphene. Graphene could somedaypush existing battery technology into the 21st century, but its incredibly expensive.Sanvar Oberoi andJahan Pestonjamas are using theirhemp textiles background to engineercarbon nanosheets at a price the market can stomach at scale.

Ourotech Enabling a moreefficient cancer treatment process

Ourotech is working on a process that would enable doctors to better prioritize treatment options for cancer patients. Though the work is still in testing, it could someday make it possible for biopsies to be tested against available treatments. This is an improvement over existing methods that rank treatments for the overall population rather than a specific patient.

Braincare A minimally invasive way to monitor pressure inside the skull

Braincare has created a headband that can monitor pressure within the skull. It could offer patients a minimally invasive sensor for tracking critical information that is necessaryfor doctors working to protect the brain.

Flow Making it easy to givepresentations in virtual reality

Virtual reality is still a nascent technology, but Flow is trying to get ahead of the wave. The team is building software that would make it possiblefor anyone to design presentations optimized for multi-dimensional VR presentations. The immersive nature of VR makes it ideal for conveying ideas that fall flat (no pun intended) on old-school PowerPoints.

Golden Connectingparentsand children in the time of need

Caring for our parents (and grandparents) is difficult. The complexity and volume of information required to manage someone elses finances and other affairs often causes important information to get lost.Golden is working to aggregate this information and make it easy to sift through so that we dont miss out on cost-saving programs benefitting older Americans when we need them most.

Calorie Cloud Putting schools andbusinesses to work tobenefitthemalnourished

Calorie Cloud hostscorporate wellness challenges and school programs to get people active. But this startup isnt about increasing physical activity, its about caring for themalnourished. The more activity participants partake in, the more supportgoes to providing food for children who desperately need nutrients.

Read the original:

Here are the frontier startups that presented at Singularity University's third demo day - TechCrunch

Okay, I know the long time running joke is that Ashes of the Singularity is a benchmark and not a game, but it really is a full fledged game receiving constant updates and even an expansion that later combined both the original AotS and Ascension into one package. Okay, with that out of the way let us get to the real topic at hand here, Ryzen update. Ashes of the Singularity is now the first game to see a proper update to take advantage of the new AMD Ryzen architecture. This was much needed as even in our review of the Ryzen 7 1800x it fell short by a significant margin in this title. Just keep in mind going forward that we only took the CPU test results for that review.

taken from our video review

Testing the game was easy enough. We ran through three runs of the built in benchmark tool for GPU and CPU and took the averages of each AVG ALL, Normal Batches, Medium Batches, and Heavy Batches to see what variation we could measure between the entire run. We ran the game at 1080p with the High preset in DX12 mode for all of these tests, both GPU and CPU on both V2.10 and the new V2.20 build.

The results of the GPU tests pretty much say it all here, there has been a massive improvement in performance, better utilization of the CPU netted serious gains here. Overall average FPS saw a 19.4% improvement. Normal Batches saw the same 19.4% improvement. The Medium Batches portion of the test saw our best improvement with a 20.9% gain. Heavy Batches may have gained the least amount of the entire test but I dont know anyone who would balk at a still 18% gain over the games previous version.

The CPU benchmark here shows yet again that 20% uptick in performance gains. If you refer back to the chart from our Ryzen 7 1800x review this gain now puts the 1800x ahead of the Intel i7 7700k in this title, which is theoretically where it should be giving the games ability to scale with hardware as well as it does.

AMD did say there would be updates coming and that they were working with developers to help bring these to market, and it looks like the first delivery was a success. I know there are other sites reporting greater gains than what were showing here, but this is where we are with what weve tested and the results are good. This is a great step in the right direction for not only the companies involved, but the consumer invested in these platforms. If AMD continues to push this level of success then those early adopters are in for a real treat. Now I cant help but wonder was Ryzen the only architecture to see gains from this update. Sound off in the comment section if youre running a different setup with either an older FX based CPU or an Intel Core series and let us know if youre seeing gains as well.

Share Tweet Submit

Read more:

Ashes of the Singularity V2.20 Shows 20% Improvement On Ryzen - Wccftech

In a previous article, I introduced the recent open-sourcing of quantum computing software by DWave. DWave is the maker of a quantum computer being used and studied by a number of groups, including NASA and Google, and there are other quantum computers in the works too. Although the field is still young, recent progress has been making headlines.

If we can make practical quantum computers, they will be very powerfulbut to see why requires understanding what makes them different. In this article, Ill explain the underlying physics that makes quantum computing possible.

Quantum computers arent just a new, faster model of the computer in front of you. Theyre based on a completely different method of storing information and decision-making. Its like comparing a jet turbine to a propeller: they achieve the same purpose, but the complexity and power are vastly disproportionate.

Lets begin by reminding ourselves how digital computers work.

The basic ingredient is the binary digit, or bit, which may take only the values 0 or 1. In modern computers, bits take the form of tiny electrical switches called transistors. Transistors are in one of two states. When they are switched on, they conduct electrical current. This is the 1 state. When switched off, they are not conducting current. This is the 0 state.

In a physical computer chip, we might find a series of transistors in the following states: on, on, off, on. In binary, the mathematical language of computation, the series becomes 1101.

This might appear to be an inadequately crude method of communicating informationhow could we possibly convey the rich tapestry of the world using only this black-and-white mold? The first step is recognizing that bits can represent numbers in our traditional counting system. For example, 1101 represents the number 13 and 0110 represents the number 6.

In fact, these are the only ways we can represent 13 and 6 using bits, creating a unique translation dictionary between strings of bits and normal numbers. In this way, we can assemble arbitrarily large numbers by stringing together bits. The MacBook Pro uses a 64-bit processor to express every number up to 18,446,744,073,709,551,615.

(Check out this video to learn more about how binary works.)

But if computers could merely store numbers, we would not find them very useful. The reason computers have become ubiquitous is we can use these numbers to further represent many other things.

Take shades of gray: simply interpolate between pure black (0) and pure white (255, by convention). Colors can be decomposed into red, green, and blue components, each having their value interpolated up to 255. Logic operations, musical notes, letters in the alphabet, internet pages, online dating profiles and many other types of information may be expressed in the same way.

Modern computers use billions of transistors and multiple levels of code to produce high-def video and complex apps, but look closely enough, and the digital world reduces to a simple series of bits.

We need only look in our pocket to see that traditional computers are powerful. But there are some problems they're ill-suited to solve. This is where quantum computers come in. A quantum computer can solve a special set of problems many magnitudes of order faster than traditional computers.

What makes quantum computers so much faster? They can perform many calculations at once.

This is possible because the building blocks of quantum computers are not bits and transistors. They are qubits and physical components so small they operate by the rules of quantum physics. These components might literally be elementary particles, such as electrons, suspended in magnetic fields.

This is where the weirdness of quantum physics comes into play. The standard shorthand explanation says traditional bits can be either 1 or 0, whereas according to the rules of quantum physics, qubits can be 1, 0, or both at the same time.

This is what truly makes a quantum computer quantum. But lets dig into what that means a bit more.

To be clear, quantum computers do not offer more discrete states than a traditional computerthe states are still 1 and 0but there is no longer an exclusive choice between these states required until the very end of a calculation. This may seem paradoxicalhow can something be 1 and 0 simultaneously? And even if this is so, why is a choice required at the end?

To better understand how this is possible, imagine hiking with a magnetic compass.

During the day you navigate as you please and the terrain dictates, glancing at your compass and noting that your direction changes. You might begin walking east, then turn north, spin around to go south, before finally nearing northwest.But at the end of each day, you record only whether your encampment is north or south of your departure point that morning.

An example log might read Day 1: North. Day 2: North. Day 3: South. Day 4: North.

This two-choice answer belies your more elaborate trajectory containing all the other directions available to the compass. North represents 1 and south represents 0, but of course, there are many other intermediate choices which can be expressed. This is similar to a quantum calculation. During the calculation, a qubit may take any value, but in the final answer there is only a 1 or 0 logged.

So, the qubits initial statethe hikes trailheadis the problem its trying to solve written in binary. The qubits final statethe campsite or destinationis its part of the solution, also written in binary. And simplistically, we can think of the qubits interim state as a combination of 1 and 0, just as the other directions you moved throughout your hike were combinations of north and south.

The days hike around swamps, between hills, and through forests is the quantum calculationa circuitous route exploring the solution set with a zig northeast, a zag due west, and so on. Eventually, however, each qubit falls into a binary state, and we arrive at our destination.

During a calculation, a qubit pointing in the east direction isnt simply weighted 50 percent north, 50 percent southit will specifically remember that it was an eastern direction. This preservation of the direction is called coherence, and it is the most important property for quantum computers.

Coherence is the property of a qubit to experience the full range of values and for qubits to share these values with each other. Four coherent qubits could possess values such as east, northwest, southeast, west, whereas incoherent qubits would possess only values north, north, south, north. Further, each of their values influences the values of their fellow coherent qubits.

Since qubits sharing mixed states speeds up computationthis is how they perform multiple calculations at onceit is absolutely essential the qubit maintain coherence during the calculation. Otherwise, we are just using a simple, slow digital computer only performing one calculation at a time.

A coherent quantum computer thus considers both 0 and 1 simultaneously, performing a calculation for the north as well as the south, but weighting the answer in a way that preserves the direction of the compass. Mathematically, this can be done using imaginary numbers, meaning we dont need to consider east as a direction unique from north or south but only as a strange combination of them.

Increasing coherence time has been a major obstacle in making commercially-viable quantum computers. Calculations require at least about 100 nanoseconds, and we have now achieved about 175 nanoseconds. As noted in my last article, this should improve as software improvesthe more you can do with a quantum computer, the more resources will pour into the field.

The upshot of all this? Quantum computers offer a massive increase in computing power. A single qubit may concurrently perform two calculations, two qubits may perform four, three qubits eight, and so forth, producing exponentially increasing speed. Just thirty qubits can simultaneously perform more than one billion calculations.

Aimed at the right problems and with the right software, the rise of quantum computers may mark a very significant moment in the history of computation.

Image Credit: Shutterstock

Read the original:

This Is What Makes Quantum Computers Powerful Problem Solvers - Singularity Hub