The process of building and training machine learning models is always in the spotlight. There is a lot of talk about different Neural Network architectures, or new frameworks, facilitating the idea-to-implementation transition. Moreover, many developers are putting a lot of effort into developing tools that take care of the peripherals: data management and validation, resource management, service infrastructure, the list goes on. Despite the AI craze, most projects never make it to production. In 2015, Google published a seminal paper called the Hidden Technical Debt in Machine Learning Systems.

Carbon Relay announced this week that its Red Sky platform for configuring and optimizing container applications using machine learning algorithms now also makes it possible to scale Kubernetes clusters more efficiently. Company CTO Ofer Idan says Carbon Relay has extended the machine learning algorithms it developed for its IT operations platform based on artificial intelligence (AIOps) to now include support for the Kubernetes Horizontal Pod Autoscaler (HPA). That capability can be employed to ensure application performance is maintained consistently as applications scale or prevent the overprovisioning of infrastructure resources, he says. The Red Sky platform is available in both open source and enterprise editions. The enterprise edition includes deep reinforcement learning capabilities to continually train the artificial intelligence (AI) agent, automatic Kubernetes application configuration, data sharing and advanced automation and scheduling capabilities.

D2iQ has added a curated distribution of Kubeflow, open source software that makes it easier to deploy workflows that incorporate machine learning algorithms on a Kubernetes cluster, as an extension to its existing portfolio of automation tools. Jie Yu, chief architect for D2iQ, says KUDO for Kubeflow will make it easier for IT teams to deploy workloads that include frameworks such as Spark and Horovod on Kubernetes clusters. At the core of KUDO for Kubeflow is Kommander, a role-based tool that provides centralized management, governance and visibility into disparate Kubernetes regardless of where they are running. IT organizations that are building and deploying artificial intelligence (AI) applications based on machine learning algorithms have embraced containers to simplify building and managing all the elements of what otherwise would be a massive monolithic application that would be too unwieldy to build, update and deploy. Kubernetes, meanwhile, has become the de facto default standard for orchestrating containers.

Opsani, a provider of a cloud service that leverages artificial intelligence (AI) to help organizations reduce cloud costs, announced it will make Opsani AI available free for the next three months to organizations that have deployed Kubernetes in the cloud as long as they are spending more than $100,000 a month on cloud services. Company CEO Ross Schibler says this initiative, dubbed Project Vital, represents an effort to help organizations preserve headcount at a time when the economic fallout from the COVID-19 pandemic is requiring organizations to dramatically reduce headcounts. By enabling organizations to more efficiently scale down Kubernetes infrastructure resources, many of those organizations may be able to reduce their number of layoffs. At the same time, Schibler notes there are some organizations in, for example, the entertainment and education services segment that are being asked to maximize existing cloud resources to support increased use of applications. In either scenario, IT teams are now being asked to cope with an unprecedented amount of uncertainty, he says.

Run:AI this week announced the general availability of a namesake platform based on Kubernetes that enables IT teams to virtualize graphical processor unit (GPU) resources. Company CEO Omri Geller says the goal is to enable IT teams to maximize investments in expensive GPUs by leveraging a single line of code to plug in its platform on top of Kubernetes. That would enable IT teams to take advantage of container orchestration to schedule artificial intelligence (AI) workloads across multiple GPUs, and allows certain AI workloads to be prioritized over others, he says. Geller notes that GPUs don't lend themselves well to traditional virtual machines. Kubernetes provides an alternative approach to virtualizing bare-metal GPU resources, which are among the most expensive IT infrastructure resource any IT organization can invoke in the cloud or deploy in on-premises IT environments.

Domino Data Lab is making the case for a multi-cloud approach to building and deploying applications infused with machine learning algorithms now that its platform runs on Kubernetes. Company CEO Nick Elprin says that as organizations move to employ machine learning algorithms to build various types of applications, many of them don't appreciate the extent to which relying on proprietary services is locking them into a "walled garden" that only runs on a specific cloud computing platform. Many of those same organizations may even wake up one morning to discover they are suddenly now competing with Amazon, Google or Microsoft, all of which are rapidly expanding the type of services they provide based on machine learning algorithms, he notes. By opting to build machine learning models on a platform provided by Domino Data Lab, organizations can deploy those models on any public cloud or on-premises IT environment as they best see fit, Elprin says. Longer-term, Domino Data Labs is betting most applications employing machine learning algorithms also will be likely to span multiple clouds, he adds.

Red Hat and NVIDIA this week announced a joint effort under which AI applications can be built using containers, which then are deployed on instances of Red Hat OpenShift that are running on supercomputers utilizing NVIDIA graphical processor units (GPUs). The Red Hat OpenShift platform, which is based on an instance of Kubernetes, then would be able to host AI applications that are much more manageable using containers. Instead of trying to maintain a massive monolithic AI application that is unwieldy to maintain and update, organizations will be able to update components of the AI application as they see fit. The announcement was made at the NVIDIA GPU Technology Conference. Ron Pacheco, director of product management for Red Hat Enterprise Linux (RHEL), says the first step toward achieving this goal is deploying RHEL on NVIDIA DGX-1 hardware systems. After that, Red Hat and NVIDIA have pledged to make NVIDIA GPU Cloud (NGC) containers available on Red Hat OpenShift.

Turbonomic this week extended the reach of its workload automation platform for Kubernetes into the public cloud. Asena Hertz, senior product marketing manager for Turbonomic, says the company has added support for Amazon Elastic Container Service for Kubernetes (EKS), Microsoft Azure Kubernetes Service (AKS) and Google Kubernetes Engine (GKE). In doing so, Turbonomic ensures that any container-as-a-service (CaaS) environment has the resources required to both meet specific application performance levels as well as comply with any number of regulations. In addition to support for public clouds, the fall edition of the company's namesake platform also adds support for Pivotal Container Service (PKS) alongside existing support for other on-premises editions of Kubernetes. Hertz says the goal is to enable IT organizations to create self-service instances of Kubernetes that leverage a Turbonomic decision engine based on machine learning algorithms to make sure Kubernetes pods have the exact resources needed at any given point in time.

Anaconda Inc. announced today it is leveraging Docker containers and Kubernetes clusters to accelerate the development of AI applications built and deployed using graphical processor units (GPUs) from NVIDIA. Previously, Anaconda added support for Docker and Kubernetes to version 5.0 of Anaconda Enterprise, a commercially supported instance of an open source platform for developing, governing and automating data science and AI pipelines on Intel processors. A version 5.2 of Anaconda Enterprise extends that platform to add support for GPUs. Matthew Lodge, senior vice president of products and marketing at Anaconda, says that training AI applications has been proven to be significantly faster on GPUs. But over time, developers of AI applications will be employing a broad range of algorithms across Intel processors, GPUs, field programmable gate arrays and new classes of processors such as the TPU processors developed by Google, which are designed specifically for AI applications.

PipelineAI is taking advantage of the community edition of Docker to help organizations develop artificial intelligence (AI) applications faster at less cost. Company CEO Chris Fregly says PipelineAI Community Edition is free, public-hosted edition of the PipelineAI Enterprise Edition, with which developers can employ Apache Kafka streaming software to drive data in real time into an AI models built using Spark ML, Scikit-Learn, Xgboost, R, TensorFlow, Keras or PyTorch frameworks. The PipelineAI platform makes use of graphical processor units (GPUs) and traditional x86 processors to host an instance of Docker Community Edition that makes available various AI frameworks that need to access data in real time, says Fregly. Over time, most AI applications will need to access multiple AI models to automate a process. PipelineAI aims to reduce the cost of creating those AI models by making it less expensive for developers to determine which AI framework will work best during the life cycle of their AI application.