A public cloud is a platform that uses the standard cloud computing model to make resources, such as virtual machines, applications, or storage, available to remote users.  Public cloud services may be free or offered through several subscriptions or on-demand pricing schemes, including a pay-per-usage model.

Public cloud is an alternative application development approach to traditional on-premises IT architectures. In the basic public cloud computing model, third party hosts scalable, on-demand IT resources and provides them to users over a network connection, either the public internet or a dedicated network.

A low-code development platform is an app that provides the Graphical User Interface for Programming and develops the code at a fast rate, and reduces the traditional programming efforts. Therefore, many public clouds and big tech companies use low-code and no-code platforms to build internal workflow applications and rapidly develop customer-facing experience. Low-code and no-code platforms provide an alternative option when the business requirements matched the platform’s capabilities.

Many low-code platforms have been available in the market for more than a decade. And some of these support tens of thousands of business applications. Over time such platforms have enhanced capabilities, developer experience, hosting options, enterprise security, application integrations, and other competencies that allow rapid development and easy maintenance of the functionally rich application.

Why Public Clouds and machine learning, Target Low-code Platforms

There are two reasons why public clouds and big tech companies target low-code platforms.

First, public clouds companies target development and engineering teams that want to code apps, automate Continuous Integration/Continuous Deployment pipelines, and instantiate infrastructure as code—developing products for citizen developers and others who wish to build them with low-code requires different experiences, tools, and functionality.

Second, stand-alone-low-code platforms have evolved through multiple computing paradigms. Some go back to client-server days. Newcomers would offer matching capabilities and the strategies and motivations to remain relevant.

In some cases, the low-code platforms are woefully behind stand-alone platforms. On the other hand, they demonstrate how low-code can enable machine learning, chatbots, voice interfaces, spatial search, and more.

Low-code Development Platforms

monday.com

monday.com offers a low code development platform that helps to digitize processes and workflows. It increases employees’ productivity and engagement, helping with fast building functionality as per requirements.

The platform helps to automate the workflow without coding. It has interactive boards and custom forms that provide business data in a quick and standardize way. monday.com can be seamlessly incorporated with existing data and tools. It has over 50 prebuilt adapters to integrate it with in-house built systems through an open API.

Visual LANSA

Visual LANSA’s low-code platform stimulates and simplifies enterprise apps while making the development team more productive. It develops apps faster, easier, and at a lower cost than traditional methods. It can write code inside the IDE. It is the only platform that runs on IBMi, Windows, and Web.

Quixy

Quixy is a no-code BPM and Application Development Platform. It can be used by the business of any industry to create complex enterprise-grade apps. Application development will be 10X faster without writing any code. Quixy has dozens of prebuilt solutions for various use cases such as CRM and Project & Task Management.

KiSSFLOW

KiSSFLOW, BPM, and workflow software enables to create custom apps and automate business processes. It delivers more than 45 pre-installed apps to develop business apps. It provides a cloud-based solution that can be used by businesses of any size and from any industry.

The post PUBLIC CLOUDS AND BIG TECH CONSIDER THE LOW-CODE PLATFORMS appeared first on Artificial Intelligence.

Underground cloud services may seem like an oxymoron, but they are quite real, and criminals are using them to speed up attacks and leave very little room for compromised businesses to react.

This is according to a new report from cybersecurity firm Trend Micro, which found terabytes of internal business data and logins – including for Google, Amazon and PayPal – for sale on the dark web.

The logins are sold through access to the cloud logs where they’re stored. As a result, Trend Micro argues, more accounts are monetized and the time from compromise to the account actually being used for nefarious purposes is cut from weeks to days or hours.

Just as businesses enjoy the speed and scalability of cloud services, so do criminals; more computing power and bandwidth allows them to optimize their operations.

Criminals that buy the logs of cloud-based stolen data usually use the data for the purposes of secondary infection, with ransomware being one of the more popular choices.

The report argues that this is a new trend that may gain even more popularity in the future, and even create a “new type of cybercriminal”: an expert in data mining that uses machine learning to enhance pre-processing and extraction of information to maximize usefulness to potential buyers.

Trend Micro believes criminals will focus on standardizing their services and pricing.

The post Hackers are leaning more heavily on cloud resources appeared first on Artificial Intelligence.

Deutsche Telekom unveils plans to radically simplify the complex Internet of Things. With a unique platform, it brings all players together, including developers, operators, partners and suppliers. This “hub” will be the new industry meeting place for the Internet of Things. The most important goal is to make access and operation of an increasingly heterogeneous IoT ecosystem as simple and manageable as possible. Transparency instead of complexity, with standards and open interfaces: This is Deutsche Telekom response to the most urgent demand in the world of IoT. 

Revolutionary approach

The new industry environment brings together the complete range of the most important elements that are essential for IoT solutions: connectivity, devices, cloud services and solutions for data analysis. Deutsche Telekom is not only relying on its own IoT offerings. The marketplace with its diverse partner portfolio offers customers the greatest possible flexibility. This considerably simplifies the setup and operation of individual IoT solutions. It also drastically reduces the development time from individual customer solutions to market readiness. 

With the new “hub” for IoT we are exploiting the full potential of the Internet of Things,“ says Rami Avidan, responsible for Deutsche Telekom’s IoT business. ”The amalgamation of all IoT solutions into one framework provides a business environment that is unique to date. With this, Deutsche Telekom is establishing the world’s first open IoT network and expanding its role from IoT player to IoT orchestrator.“

The launch is planned for the second half of 2020. The IoT hub will be continuously expanded. Deutsche Telekom invites partners and customers to test the new IoT environment.

IoT has never been easier

During operation, standardized interfaces ensure uncomplicated communication with all important clouds, protocols and technologies. In other words, different IoT ecosystems now play together seamlessly. With this revolutionary approach, data and applications from the devices are easily delivered to where they are needed: on an integrated dashboard or directly into the systems and user interfaces. Logging on to different systems is no longer necessary. All information about the IoT Hub can be found here. 

Foundation of Deutsche Telekom IoT GmbH on July 1, 2020

With the spin-off of the IoT business into an independent GmbH as of July 1, 2020, Deutsche Telekom aims to accelerate its market entry. Competitors in the IoT environment are particularly agile and have lean decision-making structures. The independent Deutsche Telekom IoT GmbH (DT IoT) is intended to strengthen Deutsche Telekom’s position in this growth market. 

”As an independent company, we are more agile and can better address the fast-moving IoT market”, says Rami Avidan, CEO of Deutsche Telekom IoT GmbH. The newly founded DT IoT will have end-to-end responsibility for the entire IoT business at Deutsche Telekom.

Digital X: IoT ready for medium-sized companies

At Digital X on June 23, 2020, Rami Avidan, responsible for IoT at Deutsche Telekom, gives a strategic outlook on the Internet of Things. He will also present a new IoT platform from Deutsche Telekom. In workshops, participants will discuss how the Internet of Things can make supply chains crisis-proof. Digital X is Europe’s largest cross-industry digitization initiative. In addition to Deutsche Telekom, more than 200 national and international partners are involved. It supports small and medium-sized companies in their digital transformation. Further details on the digitization initiative can be found at telekom.com/digitalx. 

The post Deutsche Telekom launches the world’s first open platform for the Internet of Things (IoT) appeared first on Artificial Intelligence.

Cloudera today announced that it plans to make an instance of its data management platform based on Hadoop generally available this summer on Red Hat OpenShift, which is based on Kubernetes.

Arun Murthy, chief product officer for Cloudera, says Cloudera Data Platform (CDP) Private Cloud is a complement to the instances of the platform already available on Amazon Web Services (AWS) and Microsoft.

The goal is to enable IT teams to deploy a data warehouse based on CDP in the cloud or on-premises IT environments and move data across a hybrid cloud computing environment, says Murthy.

Thanks to the rise of Kubernetes it’s now easier to move workloads between cloud computing environments. However, moving data between cloud platforms has been more problematic. CDP simplifies the movement of data between cloud platforms, enabling IT teams to preserve metadata as well as the relevant security and governance controls that should be maintained, notes Murthy.

That’s critical because in the wake of the economic downturn brought on by the COVID-19 pandemic, many IT organizations are looking to centralize the management of multiple clouds to reduce the total cost, he adds.

Murthy says Cloudera will add support for other distributions of Kubernetes based on demand, noting Red Hat OpenShift currently is the dominant distribution of Kubernetes being deployed in on-premises IT environments.

CDP is based on two distributions of Hadoop coming together as a result of the Cloudera-Hortonworks merger at the beginning of last year. Since then, Hadoop and Kubernetes have played key roles in driving development of artificial intelligence applications incorporating machine and deep learning algorithms. Hadoop provides a means to manage massive amounts of data, while the containers orchestrated by Kubernetes make it possible to employ microservices to build and deploy what would otherwise be an unwieldy monolithic AI application.

Of course, as the amount of data being aggregated reaches into the petabytes, the term “big data” has become somewhat passé. The issue is not so much the amount of data being stored and processes as much as it is making sure the right data is being made available to the right microservice at the right time. In effect, sets of data need to be managed as a logical entity that can be accessed by multiple microservices, notes Murthy.

Cloudera, in fact, already makes available separate data warehouse, machine learning and data management and analytics services on top of CDP to simplify the management of data within the context of specific use cases.

It may be a while before IT teams master all the nuances of data management in a hybrid cloud computing era enabled by Kubernetes. However, as organizations seek to derive more value from the data they collect, they will need more flexible approaches to managing massive amounts of data. With the rise of agile development methodologies and DevOps, it’s never been easier to create and deploy an application. By comparison, giving those applications access to the data they require remains positively glacial in far too many organizations.

The post Cloudera’s Data Management Platform Comes to OpenShift appeared first on Artificial Intelligence.

The ongoing global COVID-19 pandemic has upended businesses around the world, forcing companies to retool their organizations to weather one of the worst disruptions to the global economy in decades. In the face of unpredictability, both decentralization in the workplace and system architectures have taken center stage as methods to combat ever-changing circumstances. While each company faces its own set of challenges, decentralized organizations are inherently better equipped to take on today’s inclement business environment.

Decentralization via microservices

Binance’s system architecture is a good case study when it comes to decentralization. Over the past two years, our core development team has devised a decentralized solution to Binance’s software platform, shifting from a more monolithic sub-architecture to a microservices-based solution.

While a monolithic architecture has its benefits, different components of its software applications are tightly coupled and built within the same software framework, presenting a problem when any singular component needs to be altered or updated.

A microservices-based architecture, on the other hand, decouples these software components so that they can work independently with fewer opportunities for inadvertent cross-interference.

This approach has enabled Binance developers to work from anywhere in the world, as various teams can work on different parts of the software independently while communicating via an application program interface, or API. A geographically diverse development team also allows Binance team members to react quickly and autonomously to achieve the highest degree of system security, resilience and reliability. Solutions can be built and executed rapidly, and when urgent updates need to be applied, there is always someone awake who can respond quickly during local working hours.

Transitioning to a microservices-based solution allows teams to work remotely as well as work independently, which side-steps the frictional costs of collaborating over different time zones and working hours.

Cloud-based server solutions

Binance has always been among the first to embrace innovative solutions that put security and users first. On the server side, Binance utilizes a cloud-based server solution that allows the team to deploy services closer to local users, thereby reducing latency and allowing users to enjoy a more responsive user experience.

While other organizations may rely on solutions based on traditional physical data centers or hybrid clouds, Binance is able to maintain an entirely cloud-based solution that allows for rapid server deployment to users around the world, meeting local users’ needs without having to establish a costly physical presence.

Built from the ground up on cloud-based infrastructure and with zero legacy infrastructure to contend with, the Binance DevOps team is able to manage thousands of servers around the world from consolidated interfaces. This means Binance is able to uphold the highest standards for security across all servers, with the same security standard applied to every operating server.

When it comes to compliance, cloud-based solutions allow the Binance team to quickly deploy local servers that meet local requirements. These local servers often have unique software integrations that are designed to maintain security and compliance in local markets.

Security-first innovation

The Binance development team:

“Security is always top priority. Internally, from product design to architecture implementations, security is always the ground rule we need to work with to keep user funds safe.”

The Binance internal system follows a zero-trust model to ensure maximum security when it comes to identity verification. This is especially important, as Binance developers need to access critical internal systems remotely. Every Binance employee needs to go through an authentication process before they are able to successfully login. No users — internal or external — are trusted until their identity is verified. These measures are applied to every single system.

Once users are verified, permissions to login to certain internal systems are granted on a least-privilege access model, depending on the roles and permissions assigned to a particular employee.

System stability and performance

In addition to security, Binance developers are continually seeking ways to improve the system architecture to achieve a high degree of resiliency. Resilient systems are configured to maintain a seamless user experience despite unpredictable external factors such as high traffic or demand.

Data-driven integrations continually monitor for incoming traffic system performance and respond to an influx in demand by adjusting the system architecture automatically, using a method called autoscaling to offer greater uptimes during periods of volatility. As trading activity has intensified during these recent months, Binance developers have been working diligently to ensure that the system maintains fast response times, creating “zero user perception” of system pressure.

Organizational decentralization

Finally, a decentralized organization needs to be decentralized by design. Binance employees are expected to execute tasks autonomously and serve as lead subject matter experts wherever possible to cut down on decision-making bottlenecks. Most employees own and lead their own projects, set their own goals, and continually find new ways to innovate upon existing solutions. This independence allows team members to execute quickly and work efficiently while connecting on an ad hoc-basis with team members. Collaboration is, of course, key, but projects move more quickly with a nimble team at the helm.

In addition to a decentralized organizational design philosophy, Binance has employed a remote workforce for most of its existence, with 1200+ team members in 50+ countries and regions collaborating together via voice and video calls, messaging apps and occasional face-to-face meetups whenever possible.

In the face of the global pandemic, tech giants like Apple, Microsoft, Twitter and Facebook have all instituted mid-to-long-term remote work policies. Ultimately, a fully decentralized workplace provides opportunities for companies to weather challenging business environments while experiencing minimal operational disruptions. The current global pandemic will undoubtedly push companies to further decentralize where possible, serving as a bellwether for change in workplaces around the world.

The post Remote Work Isn’t Enough: Shifting Toward a Decentralized System Architecture appeared first on Artificial Intelligence.

The move follows strategic cloud services deals that the Indian firm has signed with both IBM and Google in recent times.

India-headquartered IT services player HCL Technologies has formed its dedicated HCL Microsoft Business Unit focused on supplying Microsoft cloud technologies, along with those around IoT and artificial intelligence.

The Microsoft unit will create intellectual property that extends the Microsoft platform for customer-specific scenarios and will also provide additional support to clients in the financial services, healthcare and life sciences, manufacturing, retail and travel industries.

The business unit, which follows technology deals that HCL has signed with both Google Cloud and IBM, brings together more than 5,500 professionals that serve over 2,000 customers globally.

Kalyan Kumar, corporate vice president and CTO for IT services at HCL Technologies, said: “Increasingly, customers are making bold strides, incorporating IoT solutions with machine learning for analytics, running these solutions in the public cloud and supported by CRM.

“This business unit combines HCL’s specialised services and global reach with Microsoft’s powerful cloud and business technologies, making a strong and unique offering for clients.”

Spearheading the HCL Microsoft Business Unit is Don Jones, who brings 20-plus years of Microsoft experience, “having created numerous successful solutions and go-to-market campaigns with Microsoft”, said Kumar.

Judson Althoff, executive vice president at Microsoft, said: “By establishing the Microsoft Business Unit, HCL is taking an important step forward in the long-standing partnership between our two companies.

“As a result, companies will benefit from unique products and services tailored to their digital journey while fostering modern work and collaboration. Together, we will be able to offer our joint customers a rich experience.”

The post HCL Forms Dedicated Microsoft Azure Cloud Services Unit appeared first on Artificial Intelligence.

The dawn of cloud computing has proven to be an essential event in this digital era. Having an on-site data centre is not a necessity now. It has led to substantial cost savings and improved agility for organizations.

A new model of business has emerged called the Infrastructure-as-a-service (IaaS) model. Here a third-party service provider takes care of providing hosting, maintaining core infrastructure which includes hardware, software, storage and servers for the customer.

Tech behemoths like Amazon, Microsoft and Google, have plunged into the market of IaaS and have upped the ante. These three bigwigs have addressed the data sovereignty and security concerns which thwarted the growth of cloud services in its initial days. The market size for IaaS is estimated to be around $32.4 billion in 2018, which was a growth of 31.3% from 2017.

Amazon has traditionally dominated the market for IaaS, but Microsoft and Google are quickly gaining ground. While the new CEO Satya Nadella has ushered Microsoft into a “Cloud first” era, Google is not much behind in the race with its GCP(Google cloud platform). Alphabet, Google’s parent company has spent $ 6 billion on R&D in the fourth quarter of 2018. Which is a 40% increase YOY, most of the spending was on future technologies like cloud and AI.

Microsoft has increased prices of its on-site only office 2019 packages by 10%, announcing a clear push towards the cloud strategy.

With competition rife amongst the “Big Three” of the cloud computing world, customers are bound to benefit. In this article, we shall compare the three cloud services provided by Microsoft, Amazon, and Google

Computing power

AWS

The leading computing service of the Amazon web services framework is the amazon elastic compute cloud. The database administrators can optimize for costs using the ECC with other Amazon web services which promote the right amount of flexibility and compatibility. The ECC platform could be scaled up or scaled down within minutes allowing the administrators to optimize their resources. The ECC allows the administrators to deploy thousands of server instances immediately.

Amazon gives you the power of machine learning using its AWS auto-scaling monitor. The monitor continually monitors your current requirements and adjusts its capacity accordingly, without increasing the price. Amazon guarantees 99.99% service availability as a part of their service level agreement(SLA).

Apart from this, Amazon offers Amazon Elastic Container service, which supports Docker containers. With this feature, you could manage the IP address of your website, access security groups, Cloudtrail logs, Cloudwatch events and Query the state of your application.

Azure

A network of virtual machines powers the Azure computing feature, which includes development, app deployment, datacenter extensions, and testing. The Microsoft Azure is compatible with Windows, SQL, SAP, Oracle and Linux. Azure offers a hybrid model consisting of an on-premise data centre and a public cloud.

A serverless container system called the Azure Kubernetes Service(AKS) allows containerized applications which can be deployed and managed faster. The AKS allows for continuous delivery and continuous integration experience. It will enable various teams working in a virtual office to work on a single platform.

AI and Machine learning tools

AWS Machine learning

Amazon is ahead in the race for integrating IoT and AI into the cloud. Amazon’s lex interface allows you to use the same technology which has is used in its groundbreaking voice assistant Alexa. Amazon even allows you to use the power of Sagemaker, and you can use it for deploying machine learning and for staff training. Amazon’s Lambda serverless environment is a boon for companies which who wish to completely untether themselves and deploy their apps from Amazon’s serverless infrastructure.

In 2015 Amazon launched its machine learning service, which helps developers in creating machine learning models. One year later, Amazon launched services like AWS Rekognition and Polly.

Microsoft has its own resource for machine learning called the Microsoft Azure Machine learning studio. The benefit of Azure Machine learning studio is that it allows the developers to use complex machine learning models through a simple graphical UI.

Storage

Storage is one of the critical functions of any cloud service. Both Azure and AWS have excellent storage capabilities, with both service providers giving necessary facilities like REST API and server-side data encryption. Blob storage is the name given by Azure to its storage mechanism while the storage mechanism of AWS is called S3(simple storage service). Automatic replication across various regions and high availability are the characteristics of the AWS storage solution. Both AWS and Azure use the block-storage function. In the block-storage service, the data is divided into small, equally sized pieces of data called blocks. This allows for faster access to the data. Amainfrastructurezon EBS(elastic block storage) is the block storage service of AWS, which acts as a primary storage device for Amazon EC2. While there are Azure virtual disks which connect to the Azure Virtual Machines using block storage.

Both AWS and Azure provide you with high availability through replication of your VM files to various different zones. In case your VM is damaged it is replicated quickly. AWS even gives you the option of taking snapshots of the VM to use them as backups at an extra fee.

Azure provides you with the option of launching your own operating system using a VHD file. You can upload the VHD file to a blob and launch it as a VM. The thing is that once you delete your VM in Azure, the uploaded VHD file also becomes unusable. This is not the case with AWS.

Hybrid

Hybrid cloud is a strategy in which companies choose to use a combination of different infrastructure environments like public cloud service providers and on-site servers. This approach is taken by companies who cannot afford to use 100% cloud infrastructure due to data residency concerns for e.g.:- banks.

Microsoft is well- established amongst enterprises as a good option for hybrid cloud infrastructure. Through Azure stack, businesses can easily use various Azure cloud service through their own data centre. The azure stack provides you with the freedom of deploying your applications either on Azure cloud or on your own datacentre without the hassle of rewriting the code.

Using Azure stack, your company can avail a host of services like virtual machines, networking, storage, load balancing, VPN gateway, containers, functions and active directories on your own datacenters. The hardware support is provided by a lot of vendors like Dell, Cisco, Lenovo, and Huawei. The pricing is flexible, starting at rates as low as $0.008 per virtual CPU per hour.

AWS launched its own hybrid infrastructure by the name of Outposts in 2018 at reinventing conference. An AWS Outpost is a fully managed infrastructure service wherein AWS provides a set of pre-configured hardware and software to the on-site location of the customer. These racks use the same equipment which powers AWS in all the regions that amazon services. These Outposts can be configured with a variety of EC2 instances and EBS volume storages. Customers can utilize AWS outposts to launch and manage a range of AWS services like ALB for load balancing, ECS, EKS for containers and EMR for big data along with RDS. Outposts allow the customers to use the same AWS management console, SDK(Software Development Kit) and CLI(Command-Line Interface) tools which AWS provides today.

Pricing

Pricing acts as a significant determinant for those who are considering a move to the cloud infrastructure. With competition rife between various cloud infrastructure providers, the pricing has seen a constant downward trend. If you take a look, the prices are roughly the same, but a detailed comparison is difficult as both offer slightly different pricing models and come up with many special offers and discounts to lure more users.

Both Azure and AWS offer free to try services which let you test the cloud waters, helping you in deciding whether the cloud is for you.

Customer base

It helps if there are proven use cases of companies using cloud infrastructure successfully. Being the oldest service provider in the IaaS sector, AWS is at a clear advantage here with names such as Netflix, Astrazeneca, Newscorp, Airbnb, Nike, Lonely Planet and Pfizer amongst an extensive list of customers who have chosen AWS as their preferred partner.

Azure has also got reputed customers in its kitty, which includes Ford, NBC News, Easyjet, Pearson, Wallmart, Twitter, Verizon, to name a few.

Conclusion

The choice of the ideal IaaS provider really depends on your needs, and there is no one-size-fits-all solution here. Both the solutions are offered by world-renowned companies who provide a high-level of security, free look-in services for you to try, excellent support and pay-as-you-use pricing.

AWS will suit your company if you want to go with a company which has the highest experience with cloud. AWS also has a more significant global reach than Azure. AWS offers better flexibility and a more comprehensive range of services than Azure. AWS is especially cos-effective and suitable if you are a large organization.

Azure will prove to be a good fit for companies who have most of their apps and platforms on Microsoft products like Windows and if you are a startup who is migrating to cloud for the first time.

The post The Cloud infrastructure race who is better: AWS or Azure? appeared first on Artificial Intelligence.

Issues surrounding data privacy are as legally unresolved today as they were two years ago, but the recent Equifax breach now puts a clear focus on them that strikes fear into the hearts of CIOs.

The Equifax data that was breached was not big data. However, big data is a major privacy concern for IT because so much of it is coming into enterprise data repositories from so many sources; and it comes in many shapes and sizes.

After Equifax, CIOs can rest assured that their CEOs and boards will be following their work in data privacy closely—and big data is one of the areas they’ll be most concerned about.

What operational steps can IT take to assure at a grass root level that sound data privacy practices are employed for their big data?

1. Continuously vet your big data cloud-based vendors for data privacy

Many cloud vendors can provide the levels of privacy and security that you want for your big data—but you have to demand and be willing to pay for it. Never assume that by default your cloud vendor will automatically apply best practices. Your staff should carefully evaluate the privacy protections that each of your big data cloud vendors offers and determine whether these data protection levels meet your own internal governance standards. If a cloud vendor’s data privacy practices don’t meet your own governance standards, pass on the vendor. Also ask your external IT auditors to review all cloud-based vendor data protection and security practices as part of the IT audits that the auditors perform for your company. Vendor data protection and security levels should minimally be checked on an annual basis.

2. Use private clouds

Most public cloud vendors offer private cloud services, too. Placing your data in a private cloud is more expensive than being a multi-tenant customer in a public cloud, but the private cloud deployment better separates your organization’s data from that of others. Cloud-wise, it is the next best thing to keeping your data on premises.

3. Anonymize data

You can the protect the data privacy of your customers and still perform critical trends analysis. One way that this anonymizationcan be accomplished is by encrypting data elements that personally identify someone. Another way is by identifying data from individuals with similar values (let’s say that the value you are are measuring is income) and then averaging them into a composite income value that gets pulled into a larger data analysis. Other methods are data redaction or masking.

4. Locate all the big data enclaves in your company and vet these for data privacy

As organizations distribute big data throughout departments and business units, there is always a risk that the data held within departments is changed so that data privacy levels are no longer met. The department responsible for big data governance and administration should regularly identify and track the big data marts that are distributed throughout the company. These localized big data marts should also be periodically audited by external IT auditors for data privacy compliance. If business units and other non-IT departments are using cloud-based services, the data privacy practices of their vendors should be verified for compliance to corporate standards. Cases of non-compliance should be immediately documented and mitigated.

5. Set your sights on GDPR

If you’re a North American company and you aren’t doing business internationally, you might not immediately have to concern yourself with the European Union’s General Data Protection Regulation (GDPR).

The GDPR, which aims for more stringent protections of individuals’ data, goes into effect in May 2018. According to a Gartner prediction, over 50% of companies affected by GDPR will not have met its requirements by 2018. The fines for non-compliance are hefty – up to 4% of annual revenue.

Keeping GDPR in sight matters because even if your company doesn’t do business in Europe today, it might in the future; and GDPR is where data privacy practices are headed in the future. If you comply with it now, you’re ahead of the game.

6. Perform social engineering audits

It’s the dark side of IT, but the reality is: employee sabotage of critical data happens, as does inadvertent and sometimes purposeful inappropriate data sharing between employees and with individuals outside of the organization. All are reasons to include a social engineering audit along with your annual IT audit when your external auditor arrives. A social engineering audit looks for phishing attacks, phone and physical entry attacks and other types of technical and social deception that can often be traced back to your own employees. You can uncover potential areas of vulnerability, and also use the audit as means of identifying the types of employee training that could be helpful.

The post 6 big data privacy practices every company should adopt in 2018 appeared first on Artificial Intelligence.

Despite the enormous speed at processing reams of data and providing valuable output, artificial intelligence applications have one key weakness: Their brains are located at thousands of miles away.

Most AI algorithms need huge amounts of data and computing power to accomplish tasks. For this reason, they rely on cloud servers to perform their computations, and aren’t capable of accomplishing much at the edge, the mobile phones, computers and other devices where the applications that use them run.

In contrast, we humans perform most of our computation and decision-making at the edge (in our brain) and only refer to other sources (internet, library, other people…) where our own processing power and memory won’t suffice.

This limitation makes current AI algorithms useless or inefficient in settings where connectivity is sparse or non-present, and where operations need to be performed in a time-critical fashion. However, scientists and tech companies are exploring concepts and technologies that will bring artificial intelligence closer to the edge.

Distributed computing on blockchain

A lot of the world’s computing power goes to waste as thousands and millions of devices remain idle for a considerable amount of time. ‌Being able to coordinate and combine these resources will enable us to make efficient use of computing power, cut down costs and create distributed servers that can process data and algorithms at the edge.

Distributed computing is not a new concept, but technologies like blockchain can take it to a new level. Blockchain and smart contracts enable multiple nodes to cooperate on tasks without the need for a centralized broker.

This is especially useful for Internet of Things (IoT), where latency, network congestion, signal collisions and geographical distances are some of the challenges we face when processing edge data in the cloud. Blockchain can help IoT devices share compute resources in real-time and execute algorithms without the need for a round-trip to the cloud.

Another benefit to using blockchain is the incentivization of resource sharing. Participating nodes can earn rewards for making their idle computing resources available to others.

A handful of companies have developed blockchain-based computing platforms. iEx.ec, a blockchain company that bills itself as the leader in decentralized high-performance computing (HPC), uses the Ethereum blockchain to create a market for computational resources, which can be used for various use cases, including distributed machine learning.

Golem is another platform that provides distributed computing on the blockchain, where applications (requestors) can rent compute cycles from providers. Among Golem’s use cases is training and executing machine learning algorithms. Golem also has a decentralized reputation system that allows nodes to rank their peers based on their performance on appointed tasks.

Portable AI coprocessors

From landing drones to running AR apps and navigating driverless cars, there are many settings where the need to run real-time deep learning at the edge is essential. The delay caused by the round-trip to the cloud can yield disastrous or even fatal results. And in case of a network disruption, a total halt of operations is imaginable.

AI coprocessors, chips that can execute machine learning algorithms, can help alleviate this shortage of intelligence at the edge in the form of board integration or plug-and-play deep learning devices. The market is still new, but the results look promising.

Movidius, a hardware company acquired by Intel in 2016, has been dabbling in edge neural networks for a while, including developing obstacle navigation for drones and smart thermal vision cameras. Movidius’ Myriad 2 vision processing unit (VPU) can be integrated into circuit boards to provide low-power computer vision and image signaling capabilities on the edge.

More recently, the company announced its deep learning compute stick, a USB-3 dongle that can add machine learning capabilities to computers, Raspberry PIs and other computing devices. The stick can be used individually or in groups to add more power. This is ideal to power a number of AI applications that are independent of the cloud, such as smart security cameras, gesture controlled drones and industrial machine vision equipment.

Both Google and Microsoft have announced their own specialized AI processing units. However, for the moment, they don’t plan to deploy them at the edge and are using them to power their cloud services. But as the market for edge AI grows and other players enter the space, you can expect them to make their hardware available to manufacturers.

Algorithms that rely on less data

Currently, AI algorithms that perform tasks such as recognizing images require millions of labeled samples for training. A human child accomplishes the same with a fraction of the data. One of the possible paths for bringing machine learning and deep learning algorithms closer to the edge is to lower their data and computation requirements. And some companies are working to make it possible.

Last year Geometric Intelligence, an AI company that was renamed to Uber AI Labsafter being acquired by the ride hailing company, introduced a machine learning software that is less data-hungry than the more prevalent AI algorithms. Though the company didn’t reveal the details, performance charts show that XProp, as the algorithm is named, requires much less samples to perform image recognition tasks.

Gamalon, an AI startup backed by the Defense Advanced Research Projects Agency (DARPA), uses a technique called “Bayesian Program Synthesis,” which employs probabilistic programming to reduce the amount of data required to train algorithms.

In contrast to deep learning, where you have to train the system by showing it numerous examples, BPS learns with few examples and continually updates its understanding with additional data. This is much closer to the way the human brain works.

BPS also requires extensively less computing power. Instead of arrays of expensive GPUs, Gamalon can train its models on the same processors contained in an iPad, which makes it more feasible for the edge.

Edge AI will not be a replacement for the cloud, but it will complement it and create possibilities that were inconceivable before. Though nothing short of general artificial intelligence will be able to rival the human brain, edge computing will enable AI applications to function in ways that are much closer to the way humans do.

The post How do you bring artificial intelligence from the cloud to the edge? appeared first on Artificial Intelligence.

The Raspberry Pi is one of the most exciting developments in hobbyist computing today. Across the world, people are using it to automate beer making, open up the world of robotics and revolutionise STEM education in a world overrun by film students. These are all laudable pursuits. Meanwhile, what is Microsoft doing with it? Creating squirrel-hunting water robots.

Over at the firm’s Machine Learning and Optimization group, a researcher saw squirrels stealing flower bulbs and seeds from his bird feeder. The research team trained a computer vision model to detect squirrels, and then put it onto a Raspberry Pi 3 board. Whenever an adventurous rodent happened by, it would turn on the sprinkler system.

Microsoft’s sciurine aversions aren’t the point of that story – its shoehorning of a convolutional neural network onto an ARM CPU is. It shows how organizations are pushing hardware further to support AI algorithms. As AI continues to make the headlines, researchers are pushing its capabilities to make it increasingly competent at basic tasks such as recognizing vision and speech.

As people expect more of the technology, cramming it into self-flying drones and self-driving cars, the hardware challenges are increasing. Companies are producing custom silicon and computing nodes capable of handling them.

Jeff Orr, research director at analyst firm ABI Research, divides advances in AI hardware into three broad areas: cloud services, on‑device, and hybrid. The first focuses on AI processing done online in hyperscale data centre environments like Microsoft’s, Amazon’s and Google’s.

At the other end of the spectrum, he sees more processing happening on devices in the field, where connectivity or latency prohibit sending data back to the cloud.

“It’s using maybe a voice input to allow for hands-free operation of a smartphone or a wearable product like smart glasses,” he says. “That will continue to grow. There’s just not a large number of real-world examples on‑device today.” He views augmented reality as a key driver here. Or there’s always this app, we suppose.

Finally, hybrid efforts marry both platforms to complete AI computations. This is where your phone recognizes what you’re asking it but asks cloud-based AI to answer it, for example.

The cloud: rAIning algorithms

The cloud’s importance stems from the way that AI learns. AI models are increasingly moving to deep learning, which uses complex neural networks with many layers to create more accurate AI routines.

There are two aspects to using neural networks. The first is training, where the network analyses lots of data to produce a statistical model. This is effectively the “learning” phase. The second is inference, where the neural network then interprets new data to generate accurate results. Training these networks chews up vast amounts of computing power, but the training load can be split into many tasks that run concurrently. This is why GPUs, with their double floating point precision and huge core counts, are so good at it.

Nevertheless, neural networks are getting bigger and the challenges are getting greater. Ian Buck, vice president of the Accelerate Computing Group at dominant GPU vendor Nvidia, says that they’re doubling in size each year. The company is creating more computationally intense GPU architectures to cope, but it is also changing the way it handles its maths.

“It can be done with some reduced precision,” he says. Originally, neural network training all happened in 32‑bit floating point, but it has optimized its newer Volta architecture, announced in May, for 16‑bit inputs with 32‑bit internal mathematics.

Reducing the precision of the calculation to 16 bits has two benefits, according to Buck.

“One is that you can take advantage of faster compute, because processors tend to have more throughput at lower resolution,” he says. Cutting the precision also increases the amount of available bandwidth, because you’re fetching smaller amounts of data for each computation.

“The question is, how low can you go?” asks Buck. “If you go too low, it won’t train. You’ll never achieve the accuracy you need for production, or it will become unstable.”

Beyond GPUs

While Nvidia refines its architecture, some cloud vendors have been creating their own chips using alternative architectures to GPUs. The first generation of Google’s Tensor Processing Unit (TPU) originally focused on 8‑bit integers for inference workloads. The newer generation, announced in May, offers floating point precision and can be used for training, too. These chips are application-specific integrated circuits (ASICs). Unlike CPUs and GPUs, they are designed for a specific purpose (you’ll often see them used for mining bitcoins these days) and cannot be reprogrammed. Their lack of extraneous logic makes them extremely high in performance and economic in their power usage – but very expensive.

Google’s scale is large enough that it can swallow the high non-recurring expenditures (NREs) associated with designing the ASIC in the first place because of the cost savings it achieves in AI‑based data centre operations. It uses them across many operations, ranging from recognizing Street View text to performing Rankbrain search queries, and every time a TPU does something instead of a GPU, Google saves power.

“It’s going to save them a lot of money,” said Karl Freund, senior analyst for high performance computing and deep learning at Moor Insights and Strategy.

He doesn’t think that’s entirely why Google did it, though. “I think they did it so they would have complete control of the hardware and software stack.” If Google is betting the farm on AI, then it makes sense to control it from endpoint applications such as self-driving cars through to software frameworks and the cloud.

FPGAs and more

When it isn’t drowning squirrels, Microsoft is rolling out field programmable gate arrays (FPGAs) in its own data centre revamp. These are similar to ASICs but reprogrammable so that their algorithms can be updated. They handle networking tasks within Azure, but Microsoft has also unleashed them on AI workloads such as machine translation. Intel wants a part of the AI industry, wherever it happens to be running, and that includes the cloud. To date, its Xeon Phi high-performance CPUs have tackled general purpose machine learning, and the latest version, codenamed Knight’s Mill, ships this year.

The company also has a trio of accelerators for more specific AI tasks, though. For training deep learning neural networks, Intel is pinning its hopes on Lake Crest, which comes from its Nervana acquisition. This is a co‑processor that the firm says overcomes data transfer performance ceilings using a type of memory called HBM2, which is around 12 times faster than DDR4.

While these big players jockey for position with systems built around GPUs, FPGAs and ASICs, others are attempting to rewrite AI architectures from the ground up.

Knuedge is reportedly prepping 256-core chips designed for cloud-based operations but isn’t saying much.

UK-based Graphcore, due to release its technology in 2017, has said a little more. It wants its Intelligence Processing Unit (IPU) to use graph-based processing rather than the vectors used by GPUs or the scalar processing in CPUs. The company hopes that this will enable it to fit the training and inference workloads onto a single processor. One interesting thing about its technology is that its graph-based processing is supposed to mitigate one of the biggest problems in AI processing – getting data from memory to the processing unit. Dell has been the firm’s perennial backer.

Wave Computing is also focusing on a different kind of processing, using what it calls its data flow architecture. It has a training appliance designed for operation in the data centre that it says can hit 2.9 PetaOPs/sec.

Edge-side AI

Whereas cloud-based systems can handle neural network training and inference, Client-side devices from phones to drones focus mainly on the latter. Their considerations are energy efficiency and low-latency computation.

“You can’t rely on the cloud for your car to drive itself,” says Nvidia’s Buck. A vehicle can’t wait for a crummy connection when making a split second decision on who to avoid, and long tunnels might also be a problem. So all of the computing has to happen in the vehicle. He touts the Nvidia P4 self-driving car platform for autonomous in-car smarts.

FPGAs are also making great strides on the device side. Intel has Arria, an FPGA co‑processor designed for low-energy inference tasks, while over at startup KRTKL, CEO Ryan Cousens and his team have bolted a low-energy dual-core ARM CPU to an FPGA that handles neural networking tasks. It is crowdsourcing its platform, called Snickerdoodle, for makers and researchers that want wireless I/O and computer vision capabilities. “You could run that on the ARM core and only send to the FPGA high-intensity mathematical operations,” he says.

AI is squeezing into even smaller devices like the phone in your pocket. Some processor vendors are making general purpose improvements to their architectures that also serve AI well. For example, ARM is shipping CPUs with increasingly capable GPU areas on the die that should be able to better handle machine learning tasks.

Qualcomm’s SnapDragon processors now feature a neural processing engine that decides which bits of tailored logic machine learning and neural inference tasks should run in (voice detection in a digital signal processor and image detection on a built‑in GPU, say). It supports the convolutional neural networks used in image recognition, too. Apple is reportedly planning its own neural processor, continuing its tradition of offloading phone processes onto dedicated silicon.

Smarter smartphones

This all makes sense to ABI’s Orr, who says that while most of the activity has been in cloud-based AI processors of late this will shift over the next few years as device capabilities balance them out. In addition to areas like AR, this may show up in more intelligent-seeming artificial assistants. Orr believes that they could do better at understanding what we mean.

“They can’t take action based on a really large dictionary of what possibly can be said,” he says. “Natural language processing can become more personalised and train the system rather than training the user.”

This can only happen using silicon that allows more processing at given times to infer context and intent. “By being able to unload and switch through these different dictionaries that allow for tuning and personalization for all the things that a specific individual might say.”

Research will continue in this space as teams focus on driving new efficiencies into inference architectures. Vivienne Sze, professor at MIT’s Energy-Efficient Multimedia Systems Group, says that in deep neural network inferencing, it isn’t the computing that slurps most of the power. “The dominant source of energy consumption is the act of moving the input data from the memory to the MAC [multiply and accumulate] hardware and then moving the data from the MAC hardware back to memory,” she says.

Prof Sze works on a project called Eyeriss that hopes to solve that problem. “In Eyeriss, we developed an optimized data flow (called row stationary), which reduces the amount of data movement, particularly from large memories,” she continues.

There are many more research projects and startups developing processor architectures for AI. While we don’t deny that marketing types like to sprinkle a little AI dust where it isn’t always warranted, there’s clearly enough of a belief in the technology that people are piling dollars into silicon.

As cloud-based hardware continues to evolve, expect hardware to support AI locally in drones, phones, and automobiles, as the industry develops.

The post What sort of silicon brain do you need for artificial intelligence? appeared first on Artificial Intelligence.