Intel Chip Chat

Bradley, a national law firm with business clients around the world, faced an all too common IT challenge – replacing an aging infrastructure. With 541 lawyers across multiple sites, new offices continuing to open, and ever-increasing demands on data, flexibility and scalability were top priorities for the project. The firm embraced edge computing to get greater performance and lower TCO from its infrastructure. It outfitted a new primary data center with DataOn HCI and Intel® OptaneTM technologies. Existing remote sites required small redundant storage and compute solutions to provide local access to data. Ellen Kirby, Senior Manager of Technical Operations and Senior Enterprise Architect, describes the value and performance of the edge implementation and lessons learned along the way. For information visit http://www.bradley.com. Intel, the Intel logo, and Optane are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries. Other names and brands may be claimed as the prope

Good maps rely on good data. The Red Cross’ Missing Maps Project leverages AI to provide governments and aid workers with the tools they need to navigate disasters like hurricanes and floods. Using a wealth of satellite imagery and machine learning, the Missing Maps Project is working to make villages, roads, and bridges more accessible in the wake of devastation. In this episode of Chip Chat, we talked to Intel deep learning algorithms engineer and researcher Alexei Bastidas about the technology behind making maps from satellite images. Bastidas has direct experience with under-mapped areas. Prior to working with Intel he served in the United States Marine Corps in Afghanistan, and today he draws on that experience when working to provide better information for humanitarian aid workers via the power of AI. Learn more about the Missing Maps Project at https://www.missingmaps.org Intel and the Intel logo trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries. Other names

Microsoft SQL Server is one of the most deployed databases on the planet, with millions of customers worldwide. So it’s important to understand how customers will be affected by emerging technologies, particularly in this era of digital transformation. Ken LeTourneau, Enterprise Solutions Architect in the Data Center Group at Intel, explains what impact data center technologies like Intel® OptaneTM DC Persistent Memory will have for customers deploying SQL Server 2019. What worked, even a few years ago, in the data center won’t support the demands of today. As Ken says, "It's no longer about hindsight, it's about insight.” SQL Server 2019 provides industry-leading performance and security, with intelligence over all of an enterprise’s data -- structured or unstructured. SQL Server 2019 with new big data features provides an easier path to acting on data, and with support for Intel Optane DC Persistent Memory, it can deliver larger memory-optimized databases for scale-up and scale-out system landscapes. As K

Deep learning workloads have evolved considerably over the last few years. Today’s models are larger, deeper, and more complex than neural networks from even a few years ago, with an explosion in size in the number of parameters per model. The Intel Nervana Neural Network Processor for Training (NNP-T) is a purpose-built deep learning accelerator to speed up the training and deployment of distributed learning algorithms. Carey Kloss is the VP and General Manager of the AI Training Products Group at Intel. In this interview, Kloss outlines the architecture and potential of the Intel Nervana NNP-T. He gets into major issues like memory and how the architecture was designed to avoid problems like becoming memory-locked, how the accelerator supports existing software frameworks like PaddlePaddle and TensorFlow, and what the NNP-T means for customers who want to keep on eye on power usage and lower TCO. To learn more about the Intel Nervana Neural Network Processor for Training go to: https://www.intel.ai/nervan

Academic research and theoretical work on deep learning have predicted an exciting future for deep learning, and now real-world technology is catching up with some of the most exciting potential of AI. Inference is a particularly fascinating part of AI, as it’s what powers the ability of neural networks to “predict” what certain data looks or sounds like. The Intel Nervana Neural Network Processor for Inference (NNP-I) is purpose-built for intensive inference workloads, and accelerating this crucial part of artificial intelligence. Gadi Singer is the VP and General Manager of the Artificial Intelligence Products Group at Intel, and a 29-year veteran of the company. In this interview, Gadi holds forth on both the high-level design philosophy informing the NNP-I’s structure, and the finer details of its design, such as power efficiency, optimizing data movement, and software support. He also talks about which industries and areas can potentially be transformed by better inference, such as image analysis, autom

In this episode, Rick Cnossen talks with Allyson about a joint solution created by Intel and VMware to make the experience of operationalizing enterprise analytics and AI simpler and more cost effective. Rick is Senior Director of Global AI Sector Solutions with the Enterprise and Government team at Intel. He explains that operationalizing analytics and AI is a goal for many companies, but there are still barriers. It's complicated and can be expensive and must be done in the context of hybrid cloud transformation with continued support for legacy workloads. Intel and VMware have created an "easy button" of sorts with the new Hybrid Cloud Data Analytics Solution. It’s an end-to-end solution that provides comprehensive analytics; it’s also easy to manage and can run in the cloud. For more information, visit intel.com/vmware and download the solution brief. Intel technologies' features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance v

In this episode of Intel Chip Chat, Jim Gordon, GM of Ecosystem Business Development, Strategy and Communication for Intel, joins us to talk about Intel’s next generation Intel® Xeon E processors and Intel® Software Guard Extensions (Intel® SGX). Jim talks about how Intel delivers the building blocks for security to CSPs, ISVs and other ecosystem partners that enable them to improve the security capabilities they offer to their customers, both in hardware and software. Intel now has general availability of the new Intel® Xeon E-2200 processors for server with double the Intel SGX enclave size – now 256MB. Take a listen to learn how this is leading to even more possibilities for new security use cases and developer innovation in HW based security. To learn more about developing solutions with Intel SGX, visit https://software.intel.com/sgx.

Natural Language Processing is a complex and important field of AI. It’s what allows AI assistants to understand human voices, chatbots to understand typed text, and AI to read scanned text from analogue media. Understanding language and producing language are at once simple and complex problems for AI. While an AI may know the meaning of individual words perfectly well, syntax, sentiment, and context can add meaning that is more difficult to detect. For this episode we’re joined by Peter Izsak, a senior deep learning data scientist and lead developer of NLP Architect at Intel. NLP Architect is an open source library built to facilitate research and collaboration within the NLP community and create a platform for NLP collaboration. Now on its fifth version, NLP Architect’s newest features allow for greater understanding of linguistic context that AI often finds so challenging. For more about NLP Architect go to: https://www.intel.ai/research-projects or http://nlp_architect.nervanasys.com/ Intel and the

Keeping track of data is a challenge that any enterprise needs to deal with. This isn’t just because of the amount of data generated by users, applications, and devices. It’s also because of how data is distributed. Organizations today often grapple with an IT footprint that spreads across the globe. Data and applications often cross borders, time zones, and oceans, and getting a clear picture of data use and activity is often beyond the scope of a single IT professional, or even a traditional IT department. Geoff Tudor is the Vice President of Vizion.ai at Panzura. Vizion.ai provides a single, consistent view of an organization’s data wherever it is on Earth, as well as a unified view of their application activity. By integrating machine learning to detect anomalous activity and potential breaches, Vizion.ai improves security and ease of use. All of this requires immense computational power with high throughput. Panzura manages this with a container-based infrastructure and a highly efficient storage layer.

Alex Quach, General Manager of Intel’s Wireline and Core Network Division, joins Chip Chat to talk about his team’s efforts to drive new levels of network performance to support advanced edge use cases and the continued evolution of 5G services. With early 5G deployments supported by existing LTE networks, many communication service providers (CommSPs) have been focused on transformation of the radio access networks to roll out new devices and enhanced mobile broadband services. In order to deliver on 5G’s promise of ultra-reliable low latency communications (URLLC), CommSPs are extending user plane virtualization from the core networks to the near-edge, far-edge and on-premise edge. A next generation central office (NGCO) moves virtualization solutions one step towards the customer from the core. CommSPs are retooling thousands of COs to support 5G user plane function, wireline access and support services, such as content delivery networks, cloud gaming and augmented reality (AR), with lots of headroom fo

The Texas Advanced Computing Center (TACC) has created some of the most powerful scientific supercomputers in the world, and currently supports over 3,500 different projects. TACC’s system’ capabilities are built with large-scale scientific data in mind, and it has provided millions computational hours for partners like CERN, LIGO Lab, and individual researchers tackling data-intensive problems. In this interview, we talked to Dr. Dan Stanzione, the Associate Vice President for Research at the University of Texas at Austin and the Executive Director of TACC. He discusses the architecture and capabilities of Frontera, TACC’s newest HPC cluster. In June 2019, Top500 ranked Frontera as the fifth most powerful supercomputer in the world making it the most powerful HPC system at any academic institution, and Stanzione theorizes that it may be the most capable general-purpose HPC system in the world for some applications. Frontera’s architecture includes 8,000 servers, each powered by 2nd Generation Intel® Xeon®

The insurance industry is no stranger to leveraging data in decision making. Underwriters have been checking ledgers and consulting information for over a hundred years, and today the industry has access to more data than ever before. The amount of data that goes into a quote is immense, often more than a single professional can easily manage. DataCubes is giving underwriters the ability to use AI to derive value from that data and asses risk efficiently. Phil Alampi is the Vice President of Customer Engagement at DataCubes, and his goal is to allow insurance professionals to do more in the way of decision-making and less in the way of administrative work that can be easily automated. Using AI powered by 2nd Generation Intel® Xeon® Scalable processors, DataCubes transforms unstructured data like PDFs into pre-populated applications, saving underwriters hundreds of hours. Armed with machine learning, Alampi and his team are helping bring data-driven decision-making to a centuries-old industry. For more abou

Brian Kelly, CEO of CloudGenera, joins us to discuss what it will take for enterprises to realize success in multi-cloud adoption. A good multi-cloud strategy includes a right mix of on- and off premises investments. As organizations address and reduce their technical debt, improve the economics and cost transparency in the data center, and establish best practices for making workload placement decisions, they will uncover their right mix. The smartest organizations will operate in a multi-cloud environment leveraging the best that private and public cloud can afford. CloudGenera provides vendor-agnostic IT analysis for a clear view of private, hybrid and public cloud solutions to help enterprises speed time to digital transformation. CloudGenera’s algorithms automate decisions for workload placement as well as simulate business cases for workload transformation to achieve better value for spend, lower risk, and improve performance. CloudGenera teamed with Intel to write “The Blueprint for Cloud Success,”

In this episode of Chip Chat, Technical Program Manager at Google Cloud, Paul Mason, discusses how Google is working to solve customer challenges migrating to multi-cloud environments. Anthos, a new solution announced in April, leverages Kubernetes to allow customers an easier path to a cloud native format. Anthos allows customers to leverage Google's management style outside of the Google Cloud to encompass on-premise, private cloud and even other public cloud deployments. Intel has worked closely with Google to help optimize architecture for Kubernetes and this continued partnership has also supported optimizations for Anthos. Learn more about how to get started leveraging Anthos for a multi-cloud environment at https://cloud.google.com/anthos. To learn more about Intel's Select Solution leveraging Anthos check out the hybrid cloud section on http://www.intel.com/selectsolutions. Explore Cloud Service Providers resources at http://www.intel.com/csp Intel, the Intel logo, Xeon, and Optane are tradem

Artificial intelligence is a powerful, disruptive technology, and Anna Bethke wants to use that power for good. Bethke is the head of AI for Social Good at Intel Artificial Intelligence. She and her team connect social good organizations with the expertise and technology they need to accomplish their mission. Bethke, often working with the Intel® Software Innovator Program, has supported a variety of organizations and innovators. Her partners have used AI to help locate missing children, build better wheelchairs, improve agriculture, and test water purity. AI for Social Good also seeks to support more ethical and transparent AI. Artificial intelligence can amplify human biases depending on what data is fed into algorithms, and training AI with appropriate data can help mitigate the effects of bias. Bethke acknowledges that everyone has a different idea of what it means to be ethical, but contends that with conversation and education, we can build AI that solves problems and promotes positive change. For mor

In this episode we’re joined by Carlos Morales, the General Manager of the AI Software Group at Intel, to talk about Nauta, an open-source deep learning training platform. Nauta (the Latin word for “sailor”) runs on Kubernetes* (the Greek word for “helmsman”) and is intended to give data scientists and their teams better control as they navigate the seas of data. It’s multi-node, and multi-user integrated, combining years of best practices into a single easy-to-install software package. Scaling deep-learning processes over multiple nodes can be challenging. Nauta is built for easy resource sharing over large teams. This allows for better access to valuable deep-learning resources. Nauta is fine-tuned to run over multiple 2nd Generation Intel® Xeon® Scalable processor based servers with multiple users, granting more flexibility and better access to valuable deep-learning resources. For more on Nauta-enabled solutions, check out this episode of Intel on AI: https://soundcloud.com/intelonai/dell-emc-re

Dr. Rene Meyer, Vice President of Technology at AMAX joins Chip Chat to talk about how enterprises are preparing for emerging workloads being fueled by the convergence of artificial intelligence (AI) and high performance computing (HPC). In addition to leading the AI & High Performance Technology group at AMAX, Dr. Meyer is a former research fellow and adjunct professor at Stanford University with multiple patents and published work in the space of emerging non-volatile memory technologies. In this interview, Dr. Meyer discusses AMAX’s focus on appliances for storage, cloud, and hyper-scale integration. Dr. Meyer goes into more depth about why training in AI is now moving to deploying models at the edge, and why 2nd Generation Intel® Xeon® Scalable processors can be a good fit for such tasks given their advancements in machine learning and security technologies. He also explores how the oil and gas industry is one of the driving forces behind AI and HPC convergence, and how Intel® Optane™ DC persistent memory

Not all advanced computing takes place in data centers. Maxx Garrison, product manager with Logic Supply, talked with Chip Chat about how the Internet of Things (IoT) is evolving and advantages that come with moving compute power to the edge with help from Intel technology. Logic Supply’s sealed systems turn up in demanding settings where contamination is a concern – farms, factories, hospitals, emergency vehicles – close to users and environmental data. As applications get smarter, data is coming from a variety of sources, and the amount is increasing. Sending it to the cloud can be expensive, and transmission can be a liability in decision-critical situations that demand low latency. The ability to gather, analyze, and deliver information quickly makes computing at the edge vital, and also opens complex use cases, like sensor fusion, where visual, thermal, and vibration data from local and remote sensors are combined and run through models. Sensor fusion with edge AI could expand what systems can do

As AI expands into the mainstream, how will it succeed at the edge? Matt Jacobs, Senior VP Commercial Systems, Penguin Computing, Inc. talks about the growth of AI, its move from the data center, and easing customers into the edge space. An established leader in data center and HPC solutions, Penguin takes a system-level view of the move to AI, a trend they’ve seen developing over the past two years. In the next 18 months, symbiotic technologies will be converging to create “real growth opportunity,” in Jacobs’ words. Deploying AI at the edge comes with its own needs, and they differ from those of traditional data centers. Easily maintained low-power environments well-tuned for workloads are key, and the spread of processing power from the data center to the near-edge to the far-edge calls for software that can ensure workload portability. New compute platforms, like 2nd Generation Intel® Xeon® Scalable processors with built-in AI acceleration, suit the unique requirements of workloads at the edge. Upcomin

In this episode of Intel Chip Chat, Pat Conte, Senior VP of Global Business Development at HyTrust talks about the ever-evolving security landscape and the work they are doing to deliver a more secure cloud infrastructure. HyTrust is a cybersecurity company focused on locking down virtual infrastructure that support all cloud environments, and protecting the critical data within the cloud workloads. Security is a top priority for IT managers, and HyTrust partners with Intel to help secure the underlying infrastructure. Pat chats about BoundaryControl, a HyTrust solution that integrates Intel® Trusted Execution Technology (Intel TXT) and Intel® Trusted Platform Modules (Intel TPM) technologies to essentially put a boundary around a designated server or servers to help ensure that workload is only getting executed on the server that you designate it to. New capabilities and evolving landscape gives Intel and HyTrust the opportunity to enhance encryption and geo-fencing and adopt new Intel® Security Technologies