Real-time Analytics News for the Week Ending November 21

In this week's real-time analytics news: MLCommons announced new results for the MLPerf Training v6.1 benchmark suite.

Nov 22, 2026
9 minute read
Real-time weekly news

Keeping pace with news and developments in the real-time analytics and AI market can be a daunting task. Fortunately, we have you covered with a summary of the items our staff comes across each week. And if you prefer it in your inbox, sign up here!

MLCommons announced new results for the MLPerf Training v6.1 benchmark suite. The two new benchmarks added in this round, and the submissions received, demonstrate the impact of recent architectural advances in AI training systems and the increasing diversity of the hardware used to train them.

The MLPerf Training benchmark suite comprises full system tests that stress models, software, and hardware for a range of machine learning (ML) applications. The open-source and peer-reviewed benchmark suite provides a level playing field for competition, driving innovation, performance, and energy efficiency across the industry. The suite’s benchmark collection is curated by a panel of experts from the AI community.

Advertisement

Version 6.1 adds two new benchmarks: DLRMv4-HSTU, a recommender; and Qwen3.5-397B, a large-scale LLM post-training workload.

DLRMv4-HSTU is a recommendation benchmark, updating and replacing the DLRM-dcnv2 benchmark included in earlier versions of the MLPerf benchmark. The Qwen3.5-397B benchmark is a first-of-its-kind workload for MLPerf Training: large-scale post-training using Group Relative Policy Optimization (GRPO). Much of the recent innovations in AI, such as agentic functionality, rely on post-training activities.

Trends Identified in the Benchmarks

MLPerf Training v6.1 further increased the number of submitters, submitted systems, results, and the scale of typical systems, beyond some of the records set six months ago in v6.0. Participants in this round submitted 153 systems, yielding 250 test results. 57 of those systems, or 37% of the total, were cloud-based, an increase from 29% in v6.0. Twelve systems used 1,000 or more accelerators, with the largest system using 10,240 accelerators. Systems utilized NVIDIA, AMD, and Google accelerator hardware, including, for the first time, NVIDIA Rubin-generation GPUs and the Google TPUv7 (Ironwood).

Advertisement

Additionally, the MLPerf Training v6.1 round includes performance results from 24 submitting organizations: AMD, ASUSTeK, Microsoft Azure, Cisco, CoreWeave, Crusoe, Dell, Firmus, Fujitsu, Giga Computing, Google, HPE, Inventec, KRAI, Lambda, MiTAC, Nebius, NVIDIA, Oracle, Quanta Cloud Technology, Supermicro, tiny corp, TTA, University of Florida. 

View the full results for MLPerf Training v6.1 and find additional information about the benchmarks.

Real-time analytics news in brief

Microsoft announced the AI Transparency Initiative, a collaborative effort with customers and partners to establish a common, open approach to security and governance across models and providers. The initiative reflects a simple principle: Greater AI autonomy requires greater transparency.

The initiative has three parts:

·       First, Microsoft will expand its participation in the National Institute of Standards and Technology’s AI Agent Standards Initiative and draw on its research to develop open, industrywide standards for AI transparency, security and governance.

·       Second, Microsoft is collaborating with more than 40 organizations, including Accenture, Adobe, KPMG, Mastercard and Volvo Cars, to help develop open standards for AI systems that are accountable, independently verifiable and safe by design, regardless of the model or provider.

Advertisement

·       Third, Microsoft will incorporate this work into Microsoft AI, Microsoft Foundry and Microsoft Copilot, enabling customers to use the resulting evidence with their security and compliance tools, whether from Microsoft or other providers.

Alteryx introduced the VURA Governance Experience, a governance experience designed to help businesses validate AI-generated outputs and turn them into decisions they can trust. Built on Alteryx’s VURA (Visible, Understandable, Repeatable, Auditable) framework, the release brings trusted data and business logic into conversational AI tools like Claude, OpenAI, and Gemini.

ControlUp announced AI Visibility & Usage Insights, a new capability that gives IT a unified view of AI tools adoption, usage, and experience across the organization. With it, IT teams can see how employees are using AI applications, including Copilot, ChatGPT Enterprise, Gemini, Claude Enterprise, Cursor, Notion AI, and others, and get more value from their AI investments.

DataSwitch announced Mathi, the AI-powered judgment engine behind its Expert-to-Pipeline (E2P) platform. Mathi predicts mappings, derives business rules from existing code, applies domain ontologies, and documents legacy logic. Each recommendation includes a confidence score and evidence. Experts approve decisions in plain language; every approval improves subsequent recommendations, and approved decisions generate pipelines, validation tests, and lineage.

Advertisement

Digital.ai announced Digital.ai Release 26.3, enhanced with AI agent orchestration, enabling enterprises to bring AI agents into their existing governed delivery processes as first-class citizens, alongside people and automation. The solution seamlessly orchestrates customers’ agents just as it does people and automation, with the same policies, approvals, and audit trail as any other change, with human approval required wherever policy calls for it.

DDN announced a new software-defined deployment option for EXAScaler, enabled by the upcoming EXAScaler 7 release. This option allows customers to run DDN’s high-performance data infrastructure software on qualified industry-standard servers, starting with DDN AI2200, with qualified third-party appliances to follow. A limited beta is planned for early 2027.

Hammerspace announced a major expansion of its technology portfolio with the Hammerspace Data System, extending its proven data unification, orchestration and acceleration capabilities with integrated data intelligence and high-performance storage. The result is one complete, integrated Data System for AI and HPC that spans existing and new infrastructure across data centers and clouds, delivering greater performance efficiency, simpler operations and more value from existing storage investments.

HYPR launched HYPR AgentPass to solve the critical identity blind spot created by autonomous AI agents. The platform binds each agent to a named human supervisor, evaluates requests against identity and policy before actions run, routes high-risk actions to the supervisor for passkey-signed approval, and lets authorized users stop an agent on its next request from the console or a mobile device.

Advertisement

IBM announced IBM Cognos Analytics 12.2.0, which offers new agentic AI capabilities, enhanced dashboards, and improvements that simplify everyday reporting and administration. A major advancement in Cognos Analytics 12.2.0 is the general availability of Query Agent, previously available in preview. Query Agent helps business users explore enterprise data through natural-language conversations, reducing the technical effort needed to investigate performance and identify trends.

Itential introduced Infrastructure Agent Skills, an open-source library of reusable operational procedures and vendor-specific guidance for AI agents working across network and infrastructure environments. The skills work with Itential’s agent harness, FlowAI, and other harnesses that support the Agent Skills open standard, giving teams a shared foundation for turning operational intent into repeatable automation.

NVIDIA announced NVIDIA Rubin HPC, a new GPU in the NVIDIA Vera Rubin platform designed to accelerate simulation and data processing for science, engineering, and analytics workflows. The Rubin HPC GPU features a record number of CUDA cores (more than 50,000 on a single GPU), as well as a sparse data acceleration capability that keeps those cores fed with data and accelerates GPU-to-GPU connectivity with seventh-generation NVIDIA NVLink interconnect technology.

OneTrust announced its Fall ’26 Release, introducing new capabilities that embed governance directly into the workflows where business decisions are made. These latest innovations help OneTrust customers expand visibility into AI systems and agents, automate privacy management, simplify consent and risk operations, and bring AI-powered intelligence into third-party risk assessments.

Advertisement

The OpenSearch Software Foundation, a project of The Linux Foundation organized to support the OpenSearch open-source project, launched OpenSearch 3.10, giving engineering teams new ways to simplify hybrid search, pinpoint failures in production AI agents, and cut infrastructure overhead by ingesting cloud data at rest. A few updates include enhanced APM service maps, Prometheus metrics in Vega and Vega-Lite, agentic search templates, and more.

pgEdge announced the pgEdge AI DBA Workbench now supports decision models such as Jev, a decision model from TypeSafe AI, yielding significantly faster and more economical agentic monitoring and administration for Postgres databases. In pgEdge testing, Jev classified database monitoring alerts 15 times faster and at nearly one-fiftieth the cost of Claude Haiku 4.5. This is among the first practical applications of Jev and other decision models for use with the open-source Postgres database.

Phison Electronics announced the launch of the Phison AI Data Platform (AIDP) with generative business intelligence (GenBI), a turnkey, sovereign AI system. Powered by Pascari aiDAPTIV Cache Memory technology, Phison AIDP combines locally hosted AI models, integrated inference services, governed analytics and persistent enterprise storage to help organizations unlock business intelligence while maintaining control over their data, infrastructure and AI operations.

Advertisement

Pulumi announced a platform expansion into software delivery, letting AI agents take on software operations at the pace of code with the same governance applied to engineers. Pulumi Delivery and Pulumi Agents accelerate the work that follows code, including provisioning, configuration, secrets, policy, and software delivery. Pulumi Delivery lets teams define delivery pipelines as code in TypeScript, Python, Go, and other languages of their choice.

Qlik announced the availability of new advanced analytics and custom skills in Qlik Answers, bringing more of an organization’s analytical expertise into AI-assisted business decisions. The enhancements combine AI reasoning with the Qlik Analytics Engine and company-specific business context to help teams investigate complex questions and evaluate what to do next.

Rocket Software launched ContentEdge Constellation, a new platform for enterprises to use AI across regulated content where it already resides, while maintaining the governance and controls required around that information. The solution spans cloud, on-premises, hybrid, and mainframe environments, bringing documents, images, audio, video, and structured data into AI workflows without requiring organizations to centralize or migrate decades of enterprise content.

SnapLogic announced new advancements to its Agentic Integration and Applied AI Platform, including the launch of Discovery Hub, a centralized registry for discovering and reusing Model Context Protocol (MCP) servers and tools. The company also introduced expanded agent-building capabilities in SnapCode, enhanced pipeline governance and persistent conversations in SnapGPT, and plans for new Agent Observability capabilities.

Advertisement

Spur Intelligence announced its AI Data Feed to help organizations identify AI services, investigate suspicious activity, and make smarter security and access decisions. Covering approximately 150,000 IP addresses across a growing number of operators, the feed gives security and fraud teams information about the services accessing their systems. Teams can use that intelligence to investigate suspicious activity and set access policies that preserve useful AI traffic while reducing abuse.

Starburst announced Context Studio. Starburst has always federated data: Context Studio now federates context. It gathers the definitions an organization already keeps, without a rebuild, and applies them alongside the data when an agent queries it. Context Studio extends Starburst Data Products into the Enterprise Context Layer, built for agents and people. 

ThoughtSpot announced ThoughtSpot Customer Hosted, a new deployment option that enables organizations to run ThoughtSpot's AI-powered analytics platform within their own cloud infrastructure. Designed for enterprises with strict security, governance, and compliance requirements, the solution helps organizations bring agentic analytics to their teams while maintaining control over their infrastructure and sensitive business data.

VAST Data announced Job-Level Quality of Service (QoS), a new capability coming to the VAST AI Operating System that enables schedulers and administrators to dynamically control how individual computing jobs consume shared I/O resources, without stopping or restarting them.

Advertisement

Wherobots announced Waldo, an AI agent that turns ideas about the physical world into production spatial data analysis and pipelines. Data engineers and analysts describe what they want in natural language. Waldo finds the data, builds and runs the analysis, maps the result, and schedules the work to run again as new data arrives.

Partnerships, collaborations, and more

Microsoft and Salesforce announced Copilotforce, a new partnership that helps employees get more done in the tools they already use. The solution brings Salesforce customer context, business processes, and actions into Microsoft Copilot and other AI-powered productivity experiences.

If your company has real-time analytics news, send your announcements to ssalamone@rtinsights.com.

In case you missed it, here are our most recent weekly real-time analytics news roundups:

·       Real-time Analytics News for the Week Ending November 14

·       Real-time Analytics News for the Week Ending November 7

·       Real-time Analytics News for the Week Ending October 31

·       Real-time Analytics News for the Week Ending October 24

·       Real-time Analytics News for the Week Ending October 17

·       Real-time Analytics News for the Week Ending October 10

·       Real-time Analytics News for the Week Ending October 3

Advertisement

·       Real-time Analytics News for the Week Ending September 26

·       Real-time Analytics News for the Week Ending September 19

·       Real-time Analytics News for the Week Ending September 12

·       Real-time Analytics News for the Week Ending September 5

·       Real-time Analytics News for the Week Ending August 29

·       Real-time Analytics News for the Week Ending August 22

·       Real-time Analytics News for the Week Ending August 15

Salvatore Salamone

Salvatore Salamone is a physicist by training who has been writing about science and information technology for more than 30 years. During that time, he has been a senior or executive editor at many industry-leading publications including High Technology, Network World, Byte Magazine, Data Communications, LAN Times, InternetWeek, Bio-IT World, and Lightwave, The Journal of Fiber Optics. He also is the author of three business technology books.

Featured Resources from Cloud Data Insights

Real-time Analytics News for the Week Ending November 21
The Rise of the AI-Forward Engineer for the Modern Enterprise
Douglas Vargo
Nov 20, 2026
Real-time Analytics News for the Week Ending November 14
Real-time Analytics News for the Week Ending November 7
RT Insights Logo

Analysis and market insights on real-time analytics including Big Data, the IoT, and cognitive computing. Business use cases and technologies are discussed.

Property of TechnologyAdvice. © 2026 TechnologyAdvice. All Rights Reserved

Advertiser Disclosure: Some of the products that appear on this site are from companies from which TechnologyAdvice receives compensation. This compensation may impact how and where products appear on this site including, for example, the order in which they appear. TechnologyAdvice does not include all companies or all types of products available in the marketplace.