Run:AI | April 22, 2020
A virtualized AI infrastructure for DL would run a single workload on multiple shared physical resources.
Article by Google’s Francois Chollet discusses the skill acquisition-based approach to gathering intelligence.
The conventional computing stack – from processor to firmware to virtualization, abstraction, orchestration and operating layers through to end-user software.
Is the recent progress in deep learning true artificial intelligence? A widely-discussed article by Google’s Francois Chollet discusses the skill acquisition-based approach to gathering intelligence – the one currently in use in modern DL. He argues that with huge data sets available for training models, AI is mastering skill-acquisition but not necessarily the “scope, generalization difficulty, priors, and experience” that true AI should incorporate. Even with our progress in AI, and specifically DL, we are nowhere near the limits of what DL can achieve with bigger, better-trained more accurate models, those that take into account not only skill but experience, and generalization of that experience.
Understandably, this has put intense focus on computing power, particularly the hardware that enables data scientists to run complex training experiments.
Nvidia increasing sees DL as a key market for its GPUs and bought Mellanox to speed communication inside a GPU cluster. With its recent acquisition of Habana, Intel is likely betting that custom AI accelerator hardware is a better match.
Other AI-first hardware includes Cerebras’s massive chip in a custom box that’s designed for the specific types of intensive, long-running workloads that training DL models require. In the cloud, Google’s Tensor Processing Units offer another bespoke option.
For companies running their own DL workloads, more compute is generally better. Whether exotic AI accelerators or tried-and-tested GPUs, quicker model training means more iterations, faster innovation and reduced time-to-market. It may even mean we can achieve “strong” AI (i.e., AI than goes beyond “narrow AI,” which is the capability of doing a single, discrete task) quicker.
In 2020, continuing the trend of recent years, companies will invest in ever-more AI hardware, in an effort to satisfy data scientists’ demands for compute to run bigger models to solve more complex business problems. But hardware isn’t the whole picture.
The conventional computing stack – from processor to firmware to virtualization, abstraction, orchestration and operating layers through to end-user software – was designed for traditional workloads, prioritizing high-availability, short-duration operations.Training a DL model, though, is the opposite of this sort of workload. While running a model, an experiment may need 100 percent of all the computing power of one or multiple processors for hours or even days at a time.
Read more: Virtualization can transform your company’s IT infrastructure
VIRTUAL DESKTOP STRATEGIES
Windstream | June 02, 2021
Windstream Enterprise (WE), a provider of managed communications services, has announced new feature enhancements and hardware for its VMware-powered SD-WAN.
VMware's three new SD-WAN edge devices represent the company's next-generation technology, named a Leader in the 2020 Gartner Magic Quadrant for WAN Edge Infrastructure. These new edge devices provide improved performance with enhanced processors that offer additional functionalities and simplified design and deployment with less complexity. Windstream Enterprise will now provide customers with industry-leading SD-WAN technology due to the enhancements.
By integrating SD-WAN with security, the new SD-WAN edges assist in the evolution of the SASE (Secure Access Service Edge). MNS is deployed as a Virtual Network Function (VNF) on the VMware edge, providing unified threat management security for the SD-WAN network.
Customers will be able to access the MNS VNF using Windstream Enterprise's award-winning WE Connect portal to remotely self-service, view, and configure their VNF firewalls to enable the following benefits:
• View firewall details and policies to ensure proper network protection configuration.
• Manage web filtering policies, such as the ability to block specific URLs or categories depending on content.
• Policies for application control that protect employees from inappropriate, harmful, and unprotected apps
• Reports are generated automatically to review and audit online activity and potentially risky behavior.
Windstream Holdings is a communications and software company that is privately held. Windstream provides managed communications services, such as SD-WAN and UCaaS, and high-capacity bandwidth and transport services to businesses across the United States. In addition, with an enhanced fiber network and 5G fixed wireless service, the company provides premium Internet, entertainment, and security services to consumers and small and midsize businesses primarily in rural areas in 18 states.
VMware | October 28, 2021
To bring innovative apps and services to customers more quickly, service providers are betting heavily on the cloud. And as apps and data continue to distribute across clouds and locations, service providers require flexibility to deploy and consistently manage workloads anywhere. That is why VMware is unveiling enhancements to its Telco Cloud Platform designed to help service providers speed their multi-cloud transformation.
“With network modernization efforts well underway, service providers are now focused on how to monetize 5G services. And how to do it fast. Thanks to cloud-native architectures, time to market for launching new services has moved from months and years to days and weeks. Over 150 CSPs around the world rely on VMware Telco Cloud Platform to accelerate time to revenue while providing automation and consistent operations across networks for 4G, 5G and beyond. The VMware Telco Cloud Platform provides multi-cloud capabilities from the core to the edge without locking the service provider into a single cloud architecture or deployment. It thereby delivers the agility and speed our customers need for their massive 5G deployments.”
Sanjay Uppal, senior vice president and general manager, Service Provider and Edge, VMware
The enhancements to VMware Telco Cloud Platform include:
VMware Telco Cloud Platform – Public Cloud will let service providers seamlessly incorporate public clouds into their overall 5G architecture.
VMware Telco Cloud Platform – Edge will provide the foundation for delivery of next-generation communications and compute services at the service provider and enterprise edge.
New VMware Telco Cloud Automation capabilities will enable consistent, multi-cloud operations across the entire network. In addition, support for network slicing will be added through Project Lattice, now in tech preview.
VMware’s RAN Intelligent Controller (RIC) platform will unlock app innovation and velocity across radio access networks (RAN).
Introducing VMware Telco Cloud Platform – Public Cloud
Service providers have started to embrace public clouds as a part of their network architecture. As network functions are deployed across on-premises telco clouds and public clouds, service providers require a platform that enables consistent multi-cloud operations and automation.
Introduced VMware Telco Cloud Platform – Public Cloud, powered by VMware Cloud, offers service providers choice and flexibility to run their workloads in multi-cloud environments with ease. An extension of VMware Telco Cloud Platform, VMware Telco Cloud Platform – Public Cloud is comprised of the same field-proven virtualization solution, carrier-grade container-as-a-service (CaaS), and multi-layer automation, consistent across core, edge, RAN, and now with public cloud.
VMware Telco Cloud Platform – Public Cloud will provide service providers with:
The ability to migrate telco and IT workloads to public cloud infrastructure-as-a-service (IaaS), specifically VMware Cloud on AWS, quickly and cost efficiently while having the flexibility to bring them back to the on-premises telco clouds, if and when necessary.
A choice to deploy leading network functions, through VMware’s partner ecosystem, on the best clouds to serve their functional purposes based on cost, availability, and performance.
Consistent operations with multi-layer automation across any telco cloud – from core, edge, RAN, to public cloud.
VMware Telco Cloud Platform – Public Cloud is expected to be generally available in VMware’s Q4 FY22, initially supporting service providers to deploy their network functions on VMware Cloud on AWS environments. Learn more about VMware Telco Cloud Platform – Public Cloud here.
Announcing VMware Telco Cloud Platform – Edge
At VMworld 2021, VMware introduced VMware Edge, a product portfolio that will enable organizations to run, manage and better secure edge-native apps. Service providers will play a key role in VMware’s unified edge platform strategy. That’s why today VMware is announcing VMware Telco Cloud Platform – Edge, a variation of Telco Cloud Platform adapted to host and manage multi-vendor network edge solutions.
An open and multi-cloud platform, VMware Telco Cloud Platform – Edge will provide:
A foundation for deploying next-generation communication services, including Multi-Access Edge Computing (MEC), Private Mobile Networking (PMN) and Content Delivery Networks (CDNs), among others.
Ability to deploy edge-native apps at the service provider edge or the enterprise edge.
Automation and operational consistency across edges and other network sites.
A bridge to VMware’s extensive partner ecosystem spanning hyperscalers, network equipment providers, ISVs, edge-native app developers, and more.
An example of a VMware Edge ecosystem partner is Lenovo. Its comprehensive edge product portfolio coupled with VMware Telco Cloud Platform – Edge will enable automated edge services that will accelerate time to value. The combined solution will provide a strong foundation for edge deployments across different market segments.
New Telco Cloud Automation Capabilities Will Enable Multi-Cloud Operations
Spanning the entire network, Telco Cloud Automation is the orchestration and automation engine for Telco Cloud Platform. It enables consistent operations and cloud-first approach to management and orchestration by abstracting cloud complexities and exposing the new multi-cloud ecosystem through standard telco interfaces and unified resources.
VMware announced new capabilities for Telco Cloud Automation, including:
Consistent Multi-Cloud Operations:Telco Cloud Automation customers will be able to leverage Amazon Web Services’ (AWS) public cloud resources through either VMware Cloud on AWS or Amazon Elastic Kubernetes Service (EKS).
Network Slicing:VMware unveiled Project Lattice, a tech preview, that will extend Telco Cloud Automation capabilities with 3GPP standard-compliant network slicing management. This will enable service providers to plan, design, and instantiate end-to-end network slices across the RAN, edge, core and transport network domains. An add-on module to Telco Cloud Automation, the new network slicing capability will help service providers unify these domains and close the gap between the delivered services and required network resources from physical or cloud infrastructure.
Accelerate onboarding and instantiation of network functions: Telco Cloud Automation will offer new designer capabilities to support more lifecycle automation options and accelerate onboarding and instantiation of network functions through automated CaaS and infrastructure customizations.
Network operability improvements: Telco Cloud Automation will enable new automated Day 2 operations for CaaS operations, BIOS and firmware upgrades that reduce maintenance time while allowing greater management flexibility across data centers.
Augmented CI/CD automation: Updates will make it easier to integrate Telco Cloud Automation into existing CI/CD pipelines with new SDK and enable a service provider to decrease new app deployment and update time—reducing integration costs and OpEx.
Infrastructure Automation Enhancements: Telco Cloud Automation will also offer Mount Bryce driver support, and Precision Time Protocol (PTP) network function automation. These integrations will help automate away manual infrastructure-level operations that prove essential for running vRAN and other high-performance apps.
New Telco Cloud Automation capabilities will be generally available in VMware’s Q4 FY22 apart from Project Lattice, which is in tech preview. Learn more about VMware Telco Cloud Automation for multi-cloud operations here and Project Lattice for network slicing here.
Introducing VMware RIC
The RAN Intelligent Controller (RIC) is a new function introduced by the O-RAN Alliance that enables service providers to deploy cloud-native control and management apps in the RAN. VMware is introducing VMware RIC, a platform that abstracts the underlying RAN infrastructure and provides developer APIs for xApps and rApps to program the RAN.
VMware RIC platform will consist of two RICs that can run independent of each other, or together:
VMware Centralized RIC, an implementation of the non-real-time RAN Intelligent Controller (non-RT RIC) in the O-RAN Alliance reference architecture. It will manage and host rApps that have control functions with response timing greater than one second.
VMware Distributed RIC, an implementation of the near-real-time RAN Intelligent Controller (near-RT RIC) in the O-RAN Alliance reference architecture. It will manage and host xApps that have control functions with response timing less than one second, typically on the order of 10’s of milliseconds.
VMware is a leading provider of multi-cloud services for all apps, enabling digital innovation with enterprise control. As a trusted foundation to accelerate innovation, VMware software gives businesses the flexibility and choice they need to build the future. Headquartered in Palo Alto, California, VMware is committed to building a better future through the company’s 2030 Agenda.