Computing Archives - Business Intelligence Info

Tag Archives: Computing

Alphabet is repurposing Google TPUs for quantum computing simulations

March 11, 2021 Big Data

Join Transform 2021 for the most important themes in enterprise AI & Data. Learn more.

Sandbox at Alphabet, Google parent company Alphabet’s second, secretive software development team, plans to launch a set of APIs called Floq that will allow developers to use tensor processing units (TPUs) to simulate quantum computing workloads. The announcement, which was made during a February livestream that garnered little mainstream attention, hints at the potential for hardware originally designed for AI applications to extend into the quantum realm.

Experts believe that quantum computing, which at a high level entails the use of quantum-mechanical phenomena like superposition and entanglement to perform computation, could one day accelerate AI workloads compared with classical computers. Scientific discoveries arising from the field could transform energy storage, chemical engineering, drug discovery, financial portfolio optimization, machine learning, and more, leading to new business applications. Emergen Research anticipates that the global quantum computing market for the enterprise will reach $ 3.9 billion by 2027.

According to Sandbox at Alphabet research scientist Guillaume Verdon, Floq, which will initially be made available in alpha to 50 teams in the QHack Open Hackathon, will offer a simulator API that leverages the “bleeding edge” of AI compute for experimentation. The Sandbox at Alphabet team repurposed TPUs, chips developed by Google specifically for AI training and inference, to accelerate simulations in the cloud so that developers can use frontends like TensorFlow Quantum and PennyLane to create quantum models and run them remotely on Floq.

Google’s TPUs are liquid-cooled and designed to slot into server racks; deliver up to 100 petaflops of compute; and power Google products like Google Search, Google Photos, Google Translate, Google Assistant, Gmail, and Google Cloud AI APIs. Google announced the third generation in 2018 at its annual I/O developer conference and this morning took the wraps off the successor, which is in the research stages.

Verdon says that Floq simulators, dubbed Floq Units, run 10 to 100 times faster than state-of-the-art simulations accelerated by GPUs when compared in terms of runtime. Moreover, by quantum volume — the maximum size of square quantum circuits that can be implemented successfully by a machine — Floq currently exceeds 4 billion and in the future will reach into the trillions. In simplest terms, quantum circuits are a sequence of matrix math operations performed on qubits, the quantum version of a bit.

“The team has been experimenting with how to use Floq for physics, machine learning — all kinds of cool applications,” Verdon said. “And we’ve developed our own open source library for tensor networks that runs on TPUs … It’s [surprising] how good the chips are for quantum simulation. It’s almost like they were designed for this task.”

When it launches more broadly, Floq will complement Cirq, Google’s service that gives developers access to its quantum computing hardware. And it’ll compete with a number of other quantum simulators already available on the market, including a service in IBM’s Quantum Experience suite and simulators from Intel, Amazon, and Microsoft.

But Floq’s reliance on TPUs looks to set it apart from the pack — at least on the performance end of the equation.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

up-to-date information on the subjects of interest to you
our newsletters
gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
networking features, and more

Become a member

Let’s block ads! (Why?)

Big Data – VentureBeat

R3 launches confidential computing platform Conclave for enterprises

February 11, 2021 Big Data

The rise of confidential computing has primarily focused on delivering technology for data centers which can then turnaround and offer this new security architecture to enterprise customers. The goal is to get companies in such sectors as finance and health care to put more sensitive data in the cloud.

But U.K.-based R3, which was originally founded as a blockchain company, wants to extend this security from those enterprises to their partners and customers. To do that, R3 has launched Conclave, a confidential computing platform designed to allow data to remain encrypted as it passes between enterprises for processing.

In doing so, R3 CTO Richard Brown said the platform will allow companies to share more critical data sets to solve mutual problems without having to reveal the underlying data. The hope is that this creates additional trust that enables better security by enhancing the ability to detect fraud while also unlocking new types of analytics.

“If you’re sharing data, you’ve got to worry about what happens with it when it leaves your premises,” Brown said. “How do you make sure that it doesn’t get into the wrong hands? The idea with Conclave is to use the underlying confidential computing capabilities to allow firms to build new applications and services that can convince their customers that their data will be processed securely when they send it to you.”

The confidential computing movement had been evolving for several years around a set of technologies that allow data to remain encrypted while it was being processed by applications. Traditionally, such data had to be decrypted first, then processed, then re-encrypted, a process that made users of highly sensitive information nervous about putting it in the cloud.

About 18 months ago, a group of high-profile tech companies created the Confidential Computing Foundation, an open-source group to develop standards around the emerging field. Since then, companies like IBM, Intel, Microsoft, and Google have become far more vocal about the potential for confidential computing to unleash a new generation of growth and innovation for cloud computing.

In those cases, the technology is for data center operators. R3 wants to compliment that with Conclave.

“The key idea with confidential computing is how can you take an existing workload and run it securely in the cloud,” Brown said. “Conclave is looking the other way and saying if you’re a software firm providing live data processing services to your customers, ‘How can you give your customers assurance about what will happen to their information when they send it to you?’”

In the first case, a bank sending data to a cloud provider can be sure the cloud provider can’t access the underlying data. In the Conclave example, a customer sending sensitive data to a bank to be analyzed can be sure that employees of the bank can’t see or access the data.

“This technology allows you to extend that trust one step further and allow the application provider or the service provider to convince their customers that their data is protected,” Brown said.

R3 continues to offer its enterprise blockchain platform Corda, which helps digitize processes for transactions. Conclave, which is now being released in its first public commercial version after extensive beta testing, can be bought separately or used in tandem with Corda.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

up-to-date information on the subjects of interest to you
our newsletters
gated thought-leader content and discounted access to our prized events, such as Transform
networking features, and more

Become a member

Let’s block ads! (Why?)

Big Data – VentureBeat

Move Your High-performance Computing Environment to the Cloud with TIBCO and Microsoft

February 5, 2021 TIBCO Spotfire

Reading Time: 2 minutes

We’ve got a New Years’ resolution for you: Begin modernizing your grid applications. Running on Microsoft Azure, the TIBCO GridServer^® solution provides performance, speed, and analytics benefits without the time, effort, and cost associated with maintenance, software upgrades, facility overhead, and personnel.

Key Benefits of TIBCO GridServer on Microsoft

TIBCO GridServer is a market-leading infrastructure platform for grid and elastic computing—and the backbone of businesses operating in the world’s most demanding markets. By combining Microsoft Azure and GridServer software, your organization can burst workloads to the cloud when the demand arises and reap the following benefits:

Accelerated computation-intensive applications: GridServer on Microsoft enables mission-critical risk management, complex pricing, and other computation-intensive applications to quickly execute. With this pairing, computational-intensive applications are processed in minutes or seconds rather than hours.
Elastic autoscaling, monitoring, and security: On Microsoft Azure, TIBCO GridServer infrastructure provides powerful, elastic grid computing and scalability. It can run millions of tasks using tens of millions of data points in parallel, allowing for better performance, quicker results, and more accurate analyses.
Improves market risk management systems: GridServer infrastructure also provides the power and speed required for risk calculations and risk-related tasks. Potential use cases include asset pricing, high-frequency trading, trader limit management, and risk reporting.

TIBCO GridServer and Microsoft Azure in Action

One European financial institution with hundreds of software engineers uses GridServer infrastructure extensively when making risk and trading decisions. The engineering team deploys 200,000 virtual cores in multiple grids with many versions of Windows and Linux, providing massive compute power for real-time quantitative analysis and recommendations. Using TIBCO GridServer infrastructure on Microsoft Azure, the bank runs millions of calculations in half the time at 60 percent of the cost.

Using TIBCO GridServer infrastructure on Microsoft Azure, the bank runs millions of calculations in half the time at 60 percent of the cost. Click To Tweet

Register Now to Learn More

If modernizing your high-performance computing (HPC) environment on the cloud is in one of your 2021 resolutions, then joining TIBCO and Microsoft for an upcoming session on March 4th is a great way to start off the year. Technology experts will demonstrate how banks and insurance companies can quickly execute mission-critical services such as risk management, complex pricing, and other computation-intensive applications to decrease time-to-results and increase performance levels. Register now!

Let’s block ads! (Why?)

The TIBCO Blog

What Is Edge Computing (& Why Should You Care)?

February 4, 2021 BI News and Info

Benefits (and a few challenges) of edge computing

Modern edge computing architecture allows data from IoT devices to be processed at the edge of the network, either in lieu of being sent to a data center or cloud, or before being set. This arrangement presents several clear benefits:

Speed: Because it effectively eliminates lag times that come with distance, edge computing can facilitate real-time data processing without latency. This is critical for devices and applications that depend on short and predictable response times, like autonomous vehicle operating systems (as milliseconds can really matter on a busy roadway) and digital-enabled factories (where smart devices perpetually monitor the manufacturing process).
Connection: In addition to being able to improve response times, edge computing can offer complex processing (like predictions from a model) in situations where cloud connectivity is unavailable or has limited capacity, like ships at sea or rural utility substations.
Security: Cloud computing architecture is inherently centralized, which means a cloud-reliant network is especially vulnerable to a single attack or power outage. Edge computing effectively disperses this kind of risk, as data is largely processed on local devices and storage can be distributed between local servers and data centers as needed.
Cost: Edge computing helps you categorize and manage data, which can save you a lot of money associated with bandwidth and storage costs. A simple example of this is a security camera in a warehouse. Most of the time such a camera is going to be recording empty space or common activity. Sending that footage to the cloud would be a waste of both bandwidth and the cost of storage those hours of footage. If you had an AI-powered security camera, though, it could analyze imagery, detect anything out of the ordinary, and only prioritize relevant footage for cloud storage. That’s effectively how edge computing works, but on a larger scale.

As with any technological development, there are also associated challenges, namely:

Security: Yes, security is listed here as both a benefit and a challenge. That’s because while the cloud presents a centralized vulnerability, as described above, edge devices represent a distributed vulnerability, in that each device can be compromised or otherwise operate incorrectly. In the case of cloud servers, you might well be using something like AWS that has the most advanced security available, but if you’re running your own systems, you’ll have to set up security on your own.
Storage requirements: Because less data is being sent to the cloud, edge computing requires higher storage requirements on the edge devices themselves. And even though storage is less expensive than ever before, it still requires power costs, which need to be factored into any investment. As mentioned above, though, doing processing on-site can help optimize storage needs so you’re only saving essential data instead of everything.
Maintenance requirements: Maintenance issues are similar to security ones, as the distributed architecture means there are more network combinations and discrete computing nodes that have to be maintained.

The issues raised above are indeed challenges, but they’re problems that can easily addressed. Every network architecture will have its shortcomings, but when you integrate edge computing into your network you’re providing yourself with increased flexibility to customize as necessary.

This is especially true in circumstances where remote servers can not be immediately accessed, such as on ships at sea. Right now, for instance, the U.S. Navy is preparing to install edge computing infrastructure aboard its fleet of aircraft carriers so that these distant ships can run critical (and data-rich) applications independently of a central network.

This same problem impacts the industrial transportation industry as well, which is why cargo ships, freight trains, and even semi-trailers are implementing edge architectures to overcome communication limitations, process sensor data, analyze fuel consumption, and even assist with autonomous navigation.

The future of edge computing

For the moment, edge computing is not a replacement for the cloud, but rather a complementary piece of network technology. Cloud-based data centers are still uniquely positioned to handle large flows of data (including massive files) and run applications that are not critically time-sensitive.

But the volume of data generated from the Internet of Things will continue to grow, and edge devices and servers are going to be necessary to process it at the source. That’s why research firm Gartner has predicted that by 2025, a staggering 75 percent of enterprise data will be generated and processed at “the edge.”

Edge computing is already being leveraged in industrial and manufacturing sectors, and the use cases described ahead will spur further applications in the near future:

Increased data collection: Edge computing allows sensors to gather critical data in remote locations where cloud connections are not stable or cost-effective, such as on cargo ships (as discussed above) or at oil fields or underground mining sites.
Increased interoperability: IoT service providers offer different (and proprietary) devices, APIs, and data formats, which has led to interoperability issues. Edge computing architecture, though, provides a standard platform for IoT applications and converts communication protocols for older “legacy” machines.
Increased cost savings: Edge computing immediately reduces costs associated with bandwidth and data storage. And by facilitating real-time processing of sensor data, it can reduce energy consumption by automatically adjusting lighting, cooling, and other environmental controls and ensure machines are operating consistently and productively.

As these examples make clear, it’s not strictly edge computing itself that is so valuable, but rather how it provides a cost-effective way to scale up IoT adoption. Increased use of IoT-enabled sensors and machinery, in turn, provides greater opportunities for machine learning and AI solutions.

Wrapping up

With ever-increasing amounts of data been generated every day, we’re sure to see the demand for—and the impact of—edge computing solutions grow in the coming years. If you think an edge computing solution would be helpful for your organization, schedule a free AI assessment and we’ll show you how running AI and ML models at the edge can optimize and streamline your business.

Let’s block ads! (Why?)

RapidMiner

Intel inks agreement with Sandia National Laboratories to explore neuromorphic computing

October 2, 2020 Big Data

As a part of the U.S. Department of Energy’s Advanced Scientific Computing Research program, Intel today inked a three-year agreement with Sandia National Laboratories to explore the value of neuromorphic computing for scaled-up AI problems. Sandia will kick off its work using the 50-million-neuron Loihi-based system recently delivered to its facility in Albuquerque, New Mexico. As the collaboration progresses, Intel says the labs will receive systems built on the company’s next-generation neuromorphic architecture.

Along with Intel, researchers at IBM, HP, MIT, Purdue, and Stanford hope to leverage neuromorphic computing — circuits that mimic the nervous system’s biology — to develop supercomputers 1,000 times more powerful than any today. Chips like Loihi excel at constraint satisfaction problems, which require evaluating a large number of potential solutions to identify the one or few that satisfy specific constraints. They’ve also been shown to rapidly identify the shortest paths in graphs and perform approximate image searches, as well as mathematically optimizing specific objectives over time in real-world optimization problems.

Intel’s 14-nanometer Loihi chip contains over 2 billion transistors, 130,000 artificial neurons, and 130 million synapses. Uniquely, the chip features a programmable microcode engine for on-die training of asynchronous spiking neural networks (SNNs), or AI models that incorporate time into their operating model such that the components of the model don’t process input data simultaneously. Loihi processes information up to 1,000 times faster and 10,000 more efficiently than traditional processors, and it can solve certain types of optimization problems with gains in speed and energy efficiency greater than three orders of magnitude, according to Intel. Moreover, Loihi maintains real-time performance results and uses only 30% more power when scaled up 50 times, whereas traditional hardware uses 500% more power to do the same.

Intel and Sandia hope to apply neuromorphic computing to workloads in scientific computing, counterproliferation, counterterrorism, energy, and national security. Using neuromorphic research systems in-house, Sandia plans to evaluate the scaling of a range of spiking neural network workloads, including physics modeling, graph analytics, and large-scale deep networks. The labs will run tasks on the 50-million-neuron Loihi-based system and evaluate the initial results. This will lay the groundwork for later-phase collaboration expected to include the delivery of Intel’s largest neuromorphic research system to date, which the company claims could exceed more than 1 billion neurons in computational capacity.

Earlier this year, Intel announced the general readiness of Pohoiki Springs, a powerful self-contained neuromorphic system that’s about the size of five standard servers. The company made the system available to members of the Intel Neuromorphic Research Community via the cloud using Intel’s Nx SDK and community-contributed software components, providing a tool to scale up research and explore ways to accelerate workloads that run slowly on today’s conventional architectures.

Intel claims Pohoiki Springs, which was announced in July 2019, is similar in neural capacity to the brain of a small mammal, with 768 Loihi chips and 100 million neurons spread across 24 Arria10 FPGA Nahuku expansion boards (containing 32 chips each) that operate at under 500 watts. This is ostensibly a step on the path to supporting larger and more sophisticated neuromorphic workloads. Intel recently demonstrated that the chips can be used to “teach” an AI model to distinguish between 10 different scents, control a robotic assistive arm for wheelchairs, and power touch-sensing robotic “skin.”

In somewhat related news, Intel today announced it has entered into an agreement with the U.S. Department of Energy to develop novel semiconductor technologies and manufacturing processes. In collaboration with Argonne National Laboratory, the company will focus on the development and design of next-generation microelectronics technologies such as exascale, neuromorphic, and quantum computing.

Let’s block ads! (Why?)

Big Data – VentureBeat

D-Wave’s 5,000-qubit quantum computing platform handles 1 million variables

September 29, 2020 Big Data

D-Wave today launched its next-generation quantum computing platform available via its Leap quantum cloud service. The company calls Advantage “the first quantum computer built for business.” In that vein, D-Wave today also debuted Launch, a jump-start program for businesses that want to begin building hybrid quantum applications.

“The Advantage quantum computer is the first quantum computer designed and developed from the ground up to support business applications,” D-Wave CEO Alan Baratz told VentureBeat. “We engineered it to be able to deal with large, complex commercial applications and to be able to support the running of those applications in production environments. There is no other quantum computer anywhere in the world that can solve problems at the scale and complexity that this quantum computer can solve problems. It really is the only one that you can run real business applications on. The other quantum computers are primarily prototypes. You can do experimentation, run small proofs of concept, but none of them can support applications at the scale that we can.”

Quantum computing leverages qubits (unlike bits that can only be in a state of 0 or 1, qubits can also be in a superposition of the two) to perform computations that would be much more difficult, or simply not feasible, for a classical computer. Based in Burnaby, Canada, D-Wave was the first company to sell commercial quantum computers, which are built to use quantum annealing. But D-Wave doesn’t sell quantum computers anymore. Advantage and its over 5,000 qubits (up from 2,000 in the company’s 2000Q system) are only available via the cloud. (That means through Leap or a partner like Amazon Braket.)

5,000+ qubits, 15-way qubit connectivity

If you’re confused by the “over 5,000 qubits” part, you’re not alone. More qubits typically means more potential for building commercial quantum applications. But D-Wave isn’t giving a specific qubit count for Advantage because the exact number varies between systems.

“Essentially, D-Wave is guaranteeing the availability of 5,000 qubits to Leap users using Advantage,” a D-Wave spokesperson told VentureBeat. “The actual specific number of qubits varies from chip to chip in each Advantage system. Some of the chips have significantly more than 5,000 qubits, and others are a bit closer to 5,000. But bottom line — anyone using Leap will have full access to at least 5,000 qubits.”

Advantage also promises 15-way qubit connectivity, thanks to a new chip topology, Pegasus, which D-Wave detailed back in February 2019. (Pegasus’ predecessor, Chimera, offered six connected qubits.) Having each qubit connected to 15 other qubits instead of six translates to 2.5 times more connectivity, which in turn enables the embedding of larger and more complex problems with fewer physical qubits.

“The combination of the number of qubits and the connectivity between those qubits determines how large a problem you can solve natively on the quantum computer,” Baratz said. “With the 2,000-qubit processor, we could natively solve problems within 100- to 200-variable range. With the Advantage quantum computer, having twice as many qubits and twice as much connectivity, we can solve problems more in the 600- to 800-variable range. As we’ve looked at different types of problems, and done some rough calculations, it comes out to generally we can solve problems about 2.6 times as large on the Advantage system as what we could have solved on the 2000-qubit processor. But that should not be mistaken with the size problem you can solve using the hybrid solver backed up by the Advantage quantum computer.”

1 million variables, same problem types

D-Wave today also announced its expanded hybrid solver service will be able to handle problems with up to 1 million variables (up from 10,000 variables). It will be generally available in Leap on October 8. The discrete quadratic model (DQM) solver is supposed to let businesses and developers apply hybrid quantum computing to new problem classes. Instead of accepting problems with only binary variables (0 or 1), the DQM solver uses other variable sets (integers from 1 to 500, colors, etc.), expanding the types of problems that can run on Advantage. D-Wave asserts that Advantage and DQM together will let businesses “run performant, real-time, hybrid quantum applications for the first time.”

Put another way, 1 million variables means tackling large-scale, business-critical problems. “Now, with the Advantage system and the enhancements to the hybrid solver service, we’ll be able to solve problems with up to 1 million variables,” Baratz said. “That means truly able to solve production-scale commercial applications.”

Depending on the technology they are built on, different quantum computers tend to be better at solving different problems. D-Wave has long said its quantum computers are good at solving optimization problems, “and most business problems are optimization problems,” Baratz argues.

Advantage isn’t going to be able to solve different types of problems, compared to its 2000Q predecessor. But coupled with DQM and the sheer number of variables, it may still be significantly more useful to businesses.

“The architecture is the same,” Baratz confirmed. “Both of these quantum computers are annealing quantum computers. And so the class of problems, the types of problems they can solve, are the same. It’s just at a different scale and complexity. The 2000-qubit processor just couldn’t solve these problems at the scale that our customers need to solve them in order for them to impact their business operations.”

D-Wave Launch

In March, D-Wave made its quantum computers available for free to coronavirus researchers and developers. “Through that process what we learned was that while we have really good software, really good tools, really good training, developers and businesses still need help,” Baratz told VentureBeat. “Help understanding what are the best problems that they can benefit from the quantum computer and how to best formulate those problems to get the most out of the quantum computer.”

D-Wave Launch will thus make the company’s application experts and a set of handpicked partner companies available to its customers. Launch aims to help anyone understand how to best leverage D-Wave’s quantum systems to support their business. Fill out a form on D-Wave’s website and you will be triaged to determine who might be best able to offer guidance.

“In order to actually do anything with the quantum processor, you do need to become a Leap customer,” Baratz said. “But you don’t have to first become a Leap customer. We’re perfectly happy to engage with you to help you understand the benefits of the quantum computer and how to use it.”

D-Wave will make available “about 10” of its own employees as part of Launch, plus partners.

Let’s block ads! (Why?)

Big Data – VentureBeat

Baidu offers quantum computing from the cloud

September 24, 2020 Big Data

Following its developer conference last week, Baidu today detailed Quantum Leaf, a new cloud quantum computing platform designed for programming, simulating, and executing quantum workloads. It’s aimed at providing a programming environment for quantum-infrastructure-as-a-service setups, Baidu says, and it complements the Paddle Quantum development toolkit the company released earlier this year.

Baidu says a key component of Quantum Leaf is QCompute, a Python-based open source development kit with a hybrid programming language and a high-performance simulator. Users can leverage prebuilt objects and modules in the quantum programming environment, passing parameters to build and execute quantum circuits on the simulator or cloud simulators and hardware. Essentially, QCompute provides services for creating and analyzing circuits and calling the backend.

Quantum Leaf dovetails with Quanlse, which Baidu also detailed today. The company describes Quanlse as a “cloud-based quantum pulse computing service” that bridges the gap between software and hardware by providing a service to design and implement pulse sequences as part of quantum tasks. (Pulse sequences are a means of reducing quantum error, which results from decoherence and other quantum noise.) Quanlse works with both superconducting circuits and nuclear magnetic resonance platforms and will extend to new form factors in the future, Baidu says.

The unveiling of Quantum Leaf and Quanlse follows the release of Amazon Braket and Google’s TensorFlow Quantum, a machine learning framework that can construct quantum data sets, prototype hybrid quantum and classic machine learning models, support quantum circuit simulators, and train discriminative and generative quantum models. Facebook’s PyTorch relies on Xanadu’s multi-contributor project for quantum computing PennyLane, a third-party library for quantum machine learning, automatic differentiation, and optimization of hybrid quantum-classical computations. And Microsoft offers several kits and libraries for quantum machine learning applications.

Let’s block ads! (Why?)

Big Data – VentureBeat

Baidu open-sources Paddle Quantum toolkit for AI quantum computing research

May 27, 2020 Big Data

Baidu today announced Paddle Quantum, an open source machine learning toolkit designed to help data scientists train and develop AI within quantum computing applications. It’s built atop Baidu’s PaddlePaddle deep learning platform, and Baidu claims it’s “more flexible” compared with other quantum computing suites, reducing the complexity of one popular algorithm — quantum approximate optimization algorithm (QAOA) — by a claimed 50%.

Experts believe that quantum computing, which at a high level entails the use of quantum-mechanical phenomena like superposition and entanglement to perform computation, could one day accelerate AI workloads compared with classical computers. Moreover, AI continues to play a role in cutting-edge quantum computing research. Baidu is aiming to target researchers on both sides of the equation with Paddle Quantum — toolkits that include quantum development resources, optimizers, and quantum chemistry libraries.

Paddle Quantum supports three quantum applications — quantum machine learning, quantum chemical simulation, and quantum combinatorial optimization — and developers can use it to build quantum models from scratch or by following step-by-step instructions. It includes resources addressing challenges like combinatorial optimization problems and quantum chemistry simulations, as well as complex variable definitions and matrix multiplications enabling quantum circuit models and general quantum computing. It also features an implementation of QAOA that translates into a quantum neural network by identifying a model through classical simulation or running directly on a quantum computer.

“Since Baidu announced the establishment of [the] Institute for Quantum Computing in March 2018, one of our primary goals [has been] to build bridges between quantum computing and AI,” Baidu said in a statement. “[Paddle Quantum] … can help scientists and developers quickly build and train quantum neural network models and provide advanced quantum computing applications.”

VB Transform 2020 Online – July 15-17. Join leading AI executives: Register for the free livestream.

Today Baidu also unveiled the latest version of its PaddlePaddle machine learning framework, which over the past few months has gained 39 new algorithms for a total of 146 and more than 200 pretrained models. Among them are Paddle.js, a deep learning JavaScript library that allows developers to embed AI within web browsers or programs in apps like Baidu App and WeChat; Parakeet, a text-to-speech toolkit with cutting-edge algorithms like Baidu’s latest proposed WaveNet model; Paddle Large Scale Classification Tools (PLSC), which enables image classification model training across graphics cards; and EasyData, a new drag-and-drop data service for data collection, labeling, cleaning, and enhancement.

Over 1.9 million developers now use PaddlePaddle, according to Baidu, and 84,000 enterprises have created more than 230,000 models with the framework since its debut — up from 65,000 enterprises and 169,000 models as of last November. (PaddlePaddle, which was originally developed by Baidu scientists for the purpose of applying AI to products internally, was open-sourced in September 2016.) The company anticipates growth will accelerate in light of the recently relaunched PaddlePaddle hardware ecosystem initiative, which will see manufacturers such as Intel, Nvidia, Arm China, Huawei, MediaTek, Cambricon, Inspur, and Graphcore contribute expertise and promote AI app development.

The unveiling of Paddle Quantum follows the release earlier this year of Google’s TensorFlow Quantum, a machine learning framework that can construct quantum data sets, prototype hybrid quantum and classic machine learning models, support quantum circuit simulators, and train discriminative and generative quantum models. Facebook’s PyTorch has its own project for quantum computing in PennyLane, a library for quantum machine learning, automatic differentiation, and optimization of hybrid quantum-classical computations.

Let’s block ads! (Why?)

Big Data – VentureBeat

IonQ CEO Peter Chapman on how quantum computing will change the future of AI

May 10, 2020 Big Data

Businesses eager to embrace cutting-edge technology are exploring quantum computing, which depends on qubits to perform computations that would be much more difficult, or simply not feasible, on classical computers. The ultimate goals are quantum advantage, the inflection point when quantum computers begin to solve useful problems, and quantum supremacy, when a quantum computer can solve a problem that classical computers practically cannot. While those are a long way off (if they can even be achieved), the potential is massive. Applications include everything from cryptography and optimization to machine learning and materials science.

As quantum computing startup IonQ has described it, quantum computing is a marathon, not a sprint. We had the pleasure of interviewing IonQ CEO Peter Chapman last month to discuss a variety of topics. Among other questions, we asked Chapman about quantum computing’s future impact on AI and ML.

Strong AI

The conversation quickly turned to Strong AI, or Artificial General Intelligence (AGI), which does not yet exist. Strong AI is the idea that a machine could one day understand or learn any intellectual task that a human being can.

“AI in the Strong AI sense, that I have more of an opinion just because I have more experience in that personally,” Chapman told VentureBeat. “And there was a really interesting paper that just recently came out talking about how to use a quantum computer to infer the meaning of words in NLP. And I do think that those kinds of things for Strong AI look quite promising. It’s actually one of the reasons I joined IonQ. It’s because I think that does have some sort of application.”

VB Transform 2020 Online – July 15-17: Join leading AI executives at the AI event of the year. Register today and save 30% off digital access passes.

In a follow-up email, Chapman expanded on his thoughts. “For decades it was believed that the brain’s computational capacity lay in the neuron as a minimal unit,” he wrote. “Early efforts by many tried to find a solution using artificial neurons linked together in artificial neural networks with very limited success. This approach was fueled by the thought that the brain is an electrical computer, similar to a classical computer.”

“However, since then, I believe we now know, the brain is not an electrical computer, but an electrochemical one,” he added. “Sadly, today’s computers do not have the processing power to be able to simulate the chemical interactions across discrete parts of the neuron, such as the dendrites, the axon, and the synapse. And even with Moore’s law, they won’t next year or even after a million years.”

Chapman then quoted Richard Feynman, who famously said “Nature isn’t classical, dammit, and if you want to make a simulation of nature, you’d better make it quantum mechanical, and by golly it’s a wonderful problem, because it doesn’t look so easy.”

“Similarly, it’s likely Strong AI isn’t classical, it’s quantum mechanical as well,” Chapman said.

Machine learning

One of IonQ’s competitors, D-Wave, argues that quantum computing and machine learning are “extremely well matched.” Chapman is still on the fence.

“I haven’t spent enough time to really understand it,” he admitted. “There clearly is a lot of people who think that ML and quantum have an overlap. Certainly, if you think of 85% of all ML produces a decision tree. And the depth of that decision tree could easily be optimized with a quantum computer. Clearly there’s lots of people that think that generation of the decision tree could be optimized with a quantum computer. Honestly, I don’t know if that’s the case or not. I think it’s still a little early for machine learning, but there clearly is so many people that are working on it. It’s hard to imagine it doesn’t have application.”

Again, in an email later, Chapman followed up. “ML has intimate ties to optimization: many learning problems are formulated as minimization of some loss function on a training set of examples. Generally, Universal Quantum Computers excel at these kinds of problems.”

Chapman listed three improvements in ML that quantum computing will likely allow:

The level of optimization achieved will be much higher with a QC as compared to today’s classical computers.
The training time might be substantially reduced because a QC can work on the problem in parallel, where classical computers perform the same calculation serially.
The amount of permutations that can be considered will likely be much larger because of the speed improvements that QCs bring.

AI is not a focus for IonQ

Strong AI or ML, IonQ isn’t particularly interested either. The company leaves that part to its customers and future partners.

“There’s so much to be to be done in a quantum,” Champan said. “From education at one end all the way to the quantum computer itself. I think some of our competitors have taken on lots of the entire problem set. We at IonQ are just focused on producing the world’s best quantum computer for them. We think that’s a large enough task for a little company like us to handle.”

“So, for the moment we’re kind of happy to let everyone else work on different problems,” he added. “We just think, producing the world’s best quantum computer is a large enough task. We just don’t have extra bandwidth or resources to put into working on machine learning algorithms. And luckily, there’s lots of other companies that think that there’s applications there. We’ll partner with them in the sense that we’ll provide the hardware that their algorithms will run on. But we’re not in the ML business per se.”

Let’s block ads! (Why?)

Big Data – VentureBeat

Folding@home crowdsourced computing project passes 1 million downloads amid coronavirus research

March 31, 2020 Big Data

Folding@home software for donating compute for medical research passed 1 million downloads, director Greg Bowman said in a tweet today. The Folding@home Consortium is made up of 11 laboratories around the world studying the molecular structure of diseases like cancer, ALS, and influenza. Research into COVID-19 started earlier this month. The crowdsourced effort is now powered by hundreds of thousands of Nvidia GPUs and tens of thousands of AMD GPUs.

There are now over 1M devices running @foldingathome ! This includes over 356K @nvidia GPUs, over 79K @AMD GPUs, and over 593K CPUs! Thanks to all our volunteers! We’re planning more blog posts on our #COVID19 work/results this week, please stay tuned.

— Greg Bowman (@drGregBowman) March 30, 2020

According to an Nvidia blog post about the milestone, nearly 400,000 gamers donated GPUs to the effort in recent days.

Last week, Folding@home said it crossed the one exaflop in compute power milestone, or collectively larger than any supercomputer ever assembled. By comparison, the Summit supercomputer at Oak Ridge National Laboratory in Tennessee has repeatedly ranked first in Top500 supercomputing ranks and is able to muster 148 petaflops of compute power.

The compute is being used to simulate “potentially druggable protein targets” and understand how the virus that causes COVID-19 interacts with the ACE2 receptor, according to the Folding@home website.

“While we will rapidly release the simulation datasets for others to use or analyze, we aim to look for alternative conformations and hidden pockets within the most promising drug targets, which can only be seen in simulation and not in static X-ray structures,” organizer John Chodera said in a March 10 blog post to launch a series of protein folding projects into production.

In another recent initiative at the intersection of gaming and medical research, earlier this month the University of Washington introduced FoldIt, a puzzle to solve protein folding challenges. In work at the intersection of protein folding and AI, Google’s DeepMind released predictions of understudied proteins associated with SARS-CoV-2 generated by the latest version of AlphaFold.

A number of open source projects are underway to accelerate progress toward a cure. Cloud computing providers in China like Alibaba, Baidu, and Tencent, as well as AWS, Google Cloud, and Azure in the U.S. are also lending compute to researchers. The CORD-19 data set is made up of tens of thousands of scholarly works and was made available last week for both medical and NLP researchers by a group including Microsoft Research, the Allen AI Institute, and the White House.

Let’s block ads! (Why?)

Big Data – VentureBeat

Tag Archives: Computing

Alphabet is repurposing Google TPUs for quantum computing simulations

VentureBeat

R3 launches confidential computing platform Conclave for enterprises

VentureBeat

What Is Edge Computing (& Why Should You Care)?

Benefits (and a few challenges) of edge computing

The future of edge computing

Wrapping up

Intel inks agreement with Sandia National Laboratories to explore neuromorphic computing

D-Wave’s 5,000-qubit quantum computing platform handles 1 million variables

5,000+ qubits, 15-way qubit connectivity

1 million variables, same problem types

D-Wave Launch

Baidu open-sources Paddle Quantum toolkit for AI quantum computing research

IonQ CEO Peter Chapman on how quantum computing will change the future of AI

Strong AI

Machine learning

AI is not a focus for IonQ

Folding@home crowdsourced computing project passes 1 million downloads amid coronavirus research

Recent Posts

Categories

Archives