Real-time Analytics News for the Week Ending September 16

In this week’s real-time analytics news: MLCommons released its latest machine learning benchmark testing results.

Keeping pace with news and developments in the real-time analytics 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, an open AI engineering consortium, announced results from two MLPerf benchmark suites: MLPerf Inference v3.1, which delivers industry-standard Machine Learning (ML) system performance benchmarking, and MLPerf Storage v0.5.

The MLPerf Inference benchmark suite measures how fast systems can run models in a variety of deployment scenarios. The MLPerf Inference v3.1 introduced two new benchmarks to the suite. The first is a large language model (LLM) using the GPT-J reference model to summarize CNN news articles, which garnered results from 15 different submitters. The second is an updated recommender, modified to be more representative of industry practices, using the DLRM-DCNv2 reference model and a much larger datasets, with nine submissions.

There were a record number of participants in the inference benchmarking. In total, there were over 13,500 performance results and up to 40% performance gains from 26 different submitters, and over 2000 power results. Submitters include: ASUSTeK, Azure, cTuning, Connect Tech, Dell, Fujitsu, Giga Computing, Google, H3C, HPE, IEI, Intel, Intel-Habana-Labs, Krai, Lenovo, Moffett, Neural Magic, NVIDIA, Nutanix, Oracle, Qualcomm, Quanta Cloud Technology, SiMA, Supermicro, TTA, and xFusion. In addition, MLCommons noted first-time MLPerf Inference submitters Connect Tech, Nutanix, Oracle, and TTA.

MLPerf storage benchmark

The MLPerf Storage Benchmark Suite is the first open-source AI/ML benchmark suite that measures the performance of storage for ML training workloads. The organization notes that the MLPerf Storage Benchmark will be an effective tool for purchasing, configuring, and optimizing storage for machine learning applications, as well as for designing next-generation systems and technologies.

The benchmark had over 28 performance results from five companies. The submitters included Argonne National Laboratory (ANL), DDN, Micron, Nutanix, and Weka.

Benchmark measurements

The results for MLPerf Inference v3.1, MLPerf Storage v0.5, and additional information about the benchmarks can be found here:

• https://mlcommons.org/en/storage-results-05
• https://mlcommons.org/en/inference-edge-31
• https://mlcommons.org/en/inference-datacenter-31

Real-time analytics news in brief

Salesforce had a number of announcements of its own and with partners at its Dreamforce conference. It announced the Einstein 1 Platform, which features major advancements for Salesforce Data Cloud and Einstein AI capabilities, all built on Salesforce’s underlying metadata framework. A trusted AI platform for customer companies, the Einstein 1 Platform gives companies the ability to safely connect any data to build AI-powered apps with low code and deliver entirely new CRM experiences.

In other Salesforce news, the company announced the following:

• Anexpanded partnership with Amazon Web Services (AWS) for new Bring Your Own Lake (BYOL) and Bring Your Own Large Language Model (BYO LLM) integrations between AWS and Salesforce Data Cloud. These integrations build on AWS and Salesforce’s existing generative AI partnership, allowing customers to unify their data across Data Cloud and AWS services like Amazon Redshift and Amazon EMR.
• An expanded partnership with Google to bring together Salesforce and Google Workspace to drive productivity with AI. This partnership will deliver new bidirectional integrations that allow customers to bring together context from Salesforce and Google Workspace, including Google Calendar, Docs, Meet, Gmail, and more, to power generative AI experiences across platforms.
• A collaboration with Deloitte Digital to integrate Einstein, Salesforce’s AI technology, into Deloitte’s DigitalMIX platform. The new integration will unlock a range of capabilities, including predictive analytics, generative AI, and visualization, to help B2B businesses make employees more productive, drive faster revenue growth, and deepen customer relationships.

Partners announced additional offerings. Some announcements were made at the conference, and others merely coincided with the event. They include:

• Metazoa announced the launch of Intelligent Assistant on the Salesforce AppExchange. The product includes Metadata Studio, a prompt engineering platform, and Intelligent Search, an administrative thinking partner.
• Copado launched the Copado 1 platform, a turnkey end-to-end DevOps solution for Salesforce. Copado 1 unifies the entire software development lifecycle on Salesforce, including AI-enabled testing, into a single platform. Packed with DevOps best practices, the ready-to-use Copado 1 platform enables customers to accelerate DevOps maturity and time-to-value for their Salesforce deployments.
• Apromore announced seamless integration with MuleSoft RPA Manager by Salesforce. Businesses seeking enhanced efficiency and productivity can now accelerate identification of RPA opportunities and better predict the impact of changes through data-driven simulation.

Additional real-time news items from the week

Anaconda unveiled its Anaconda Toolbox, a new suite of tools built to enhance the experience and capabilities of Python in Excel, now available to Python in Excel beta users. These new features allow users of all experience levels, even those without any Python experience, to use Python in Excel to quickly generate code and visualization. Anaconda Toolbox includes the Anaconda Assistant, curated open-source data catalogs, synchronized data management, and more.  

Anodot announced the launch of CostGPT, an AI-powered tool offering cloud users precise and personalized analysis of their cloud costs. Specifically, CostGPT helps cloud users address common cost challenges, including complex pricing models, hidden costs, and inadequate cost monitoring and reporting, with a simple search.

Astronomer introduced new capabilities to Astro, its Apache Airflow-powered platform. These capabilities include a new architecture and deployment model with competitive consumption-based pricing. Part of the news is the announcement of a new component of the Astro platform: The Astro Hypervisor.Unlike conventional architectures that focus on the bare minimum required to run open-source projects on cloud-based containers, the introduction of Astro Hypervisor allows greater visibility and control over the Airflow deployments that Astronomer runs for their customers.

Cisco announced the availability of Cisco Secure Application (previously Security Insights for Cloud Native Application Observability) on the Cisco Full-Stack Observability Platform, enabling organizations to bring together application and security teams to securely develop and deploy modern applications. The latest release of Cisco Secure Application helps customers to securely manage cloud-native applications in addition to hybrid applications.

Checkly announced the introduction of Heartbeat Checks, a new feature that significantly expands the platform’s monitoring capabilities. Heartbeat checks monitor an application passively by waiting for a client to report back. Customers performing periodic actions such as sending daily status emails or monthly activity reports can now get alerted when their scheduled application actions fail. Alongside this, Checkly has also rolled out Smart Retries, an enhancement aimed at intelligent alert management.

Deci announced the launch of new generative AI Foundation Models, DeciDiffusion 1.0 and DeciLM 6B, as well as its inference Software Development Kit (SDK) – Infery LLM. By using Deci’s open-source generative models and Infery LLM, AI teams can reduce their inference compute costs and use widely available and cost-friendly GPUs such as the NVIDIA A10 while also improving the quality of their offering.

DiffusionData announced the release of Diffusion 6.10. The latest developer-centric enhancements to the framework aim to free up resources, speed, and simplify development. Enhancements in Diffusion 6.10 include floating point support in topic views, conditional update and OR operator constraints, the ability to retrieve metrics using the API, update stream recovery, and more.

Dremio announced the launch of its next-generation Reflections technology for SQL query acceleration. Queries using Reflections often run 10 to 100 times faster than unaccelerated queries. The new launch introduces Dremio Reflection Recommender, which automatically evaluates an organization’s SQL queries and generates a recommended Reflection to accelerate them.

Esperanto Technologies announced a Generative AI Appliance based on RISC-V technology. Esperanto’s Generative AI Appliance is an integrated software/hardware solution that can be installed in private data centers or at the enterprise edge using an industry-standard server form factor. The solution delivers high levels of data privacy and lower total cost of ownership (TCO).

FeatureByte announced FeatureByte Copilot, an automated, intelligent feature ideation solution. Powered by Generative AI, FeatureByte Copilot saves data science teams significant time, effort, and resources while moving AI projects from ideation to implementation faster, at scale, and with greater accuracy.

Heap introduced new updates to its platform. Specifically, the company has upgraded its data analysis tools to handle large data sets more quickly. The new capabilities allow enterprises to understand actions users take across all platforms. They also allow companies to know why users take those actions. Heap’s platform can be leveraged by enterprises across the globe or start-ups of any size. Session replay, heatmaps, and mobile auto-capture have all been updated to handle increased scale, privacy, and analysis requirements for the modern digital enterprise.

Infineon Technologies AG, which acquired Imagimob earlier this year, released the first result of that acquisition – the integration of Infineon’s ModusToolbox Software with Imagimob’s ML development platform. This integration is designed to enable the quicker, easier, and more cost-effective launch of AI-based machine learning models, allowing companies to reduce time to market.

KX announced the general availability of KDB.AI Cloud, a vector database for real-time contextual AI. KDB.AI Cloud enables developers to bring temporal and semantic context and relevancy to their AI-powered applications. Built to handle high-speed, time-series data and multi-mode query data processing, it allows business users, for example, to query real-time financial market data using natural language search with semantic relevance.

Labelbox introduced the Labelbox platform, a solution to help enterprises fine-tune and evaluate LLMs to deliver LLM systems with confidence. Labelbox helps ML teams fine-tune LLMs by providing a suite of tools to perform techniques such as reinforcement learning with human feedback (RLHF), reinforcement learning from AI Feedback (RLAIF), evaluation, and red teaming. In addition, as part of an expanded partnership announced earlier in March, Labelbox is building on generative AI technology from Google Cloud to help enterprises build LLM solutions with Vertex AI.

Mirantis announced the release of Mirantis Kubernetes Engine (MKE) 3.7.0, which adds an automatic rollback feature. With a new zero-touch automated rollback feature, MKE 3.7.0 minimizes the risk associated with updates. If a problem occurs during an upgrade, by default, it will quickly revert to the prior MKE version so that production systems can remain operational. The new release also offers fine-grained observability of components.

ngrok announced the launch of ngrok Global Server Load Balancing (GSLB). The ngrok GSLB distributes traffic across connected servers in multiple geographies to enhance application performance and resiliency without adding any overhead to ITOps.

Pegasystems Inc. announced the general availability of Pega Infinity ’23. This enhanced version of Pega’s portfolio of products offers new capabilities that help organizations get closer to becoming autonomous enterprises by unifying AI and automation. Pega Infinity ’23 features deeper intelligence, low-code, and autonomy capabilities within its entire portfolio of products in a variety of markets, including the Pega Platform, Pega Customer Decision Hub, and more. This includes Pega GenAI, a new set of more than 20 capabilities available across products on Pega Cloud. 

Qlik today announced Qlik Staige, a holistic set of solutions to help customers confidently embrace the power of Artificial Intelligence (AI). With Qlik Staige, customers can innovate and move faster by making secure and governed AI part of everything they can do with Qlik. To that end, Qlik Staige helps organizations build a trusted data foundation for AI, leverage modern AI-enhanced analytics, and deploy AI for advanced use cases.

Rocket Software released its latest Rocket DevOps [10.2.2] solution for IBM i, allowing organizations to adopt modern technology like CI/CD best practices at their own pace—without disrupting business. Rocket DevOps enables the testing and integration of AI and popular open-source and third-party tools— like Git, JIRA, Jenkins, and VS Code— into applications, all while running on IBM i and with no additional platform support required.

SAS revealed expanded capabilities to its SAS Viya flagship data, AI, and analytics platform. The new products include SAS Viya Workbench, a lightweight development environment that quickly spins up and executes code in a cloud-native way. Other new products include an on-ramp to experience SAS Viya and SAS App Factory (a rapid application development environment for creating fit-for-purpose, AI-driven applications). SAS App Factory has planned general availability for 2024.

SiMa.ai launched Palette Edgematic, a free visual development environment designed for any organization to get started with and accelerate ML at the edge. Palette Edgematic enables a “drag and drop,” code-free approach where users can create, build, and deploy their own models and complete computer vision pipelines automatically in minutes versus months while evaluating the performance and power consumption needs of their edge ML application in real time.

Teradata announced ask.ai, a new generative AI capability for VantageCloud Lake. The natural language interface is designed to allow anyone with approved access to ask questions about their company’s data and receive instant responses from VantageCloud Lake, the company’s cloud analytics and data platform for AI.

Vianai Systems announced the release of veryLLM, an open-source toolkit that enables reliable, transparent, and transformative AI systems for enterprises. The veryLLM toolkit empowers developers and data scientists to build a universally needed transparency layer into Large Language Models (LLMs) to evaluate the accuracy and authenticity of AI-generated responses.

Yugabyte announced the general availability of YugabyteDB 2.19 with bimodal query execution and built-in cloud-native connection management. Together, these capabilities expand the reach of distributed PostgreSQL to applications at every scale while simplifying application architecture and migration.

Partnerships, collaborations, and more

Canonical, the publisher of Ubuntu, announced the listing of Ubuntu Pro 22.04 LTS with real-time kernel on the AWS Marketplace. With Real-time Ubuntu on the AWS Marketplace, Canonical is innovating at the edge for AWS workloads. Enterprises in sectors like automotive, industrial, and telecom can now unlock real-time compute and reduce development time by validating their code in a cloud environment.

NeuroBlade announced it will work with the Velox community to open source hardware (HW) acceleration to the Velox execution engine. This effort aims to help organizations process massive datasets efficiently. To that point, the integration of NeuroBlade’s cutting-edge SPU technology into Velox yields an acceleration of over ten times. This effort aims to help organizations process massive datasets efficiently.

Oracle and Microsoft announced Oracle Database@Azure, which gives customers direct access to Oracle database services running on Oracle Cloud Infrastructure (OCI) and deployed in Microsoft Azure datacenters. Oracle Database@Azure delivers all the performance, scale, and workload availability advantages of Oracle Database on OCI with the security, flexibility, and services of Microsoft Azure, including AI services like Azure OpenAI.

Pineconeannounced an integration with Amazon Bedrock, a fully managed service from Amazon Web Services (AWS) for building generative AI applications. The announcement means customers can now drastically reduce hallucinations and accelerate the go-to-market of Generative AI (GenAI) applications such as chatbots, assistants, and agents.

PingCAP announced it has achieved Amazon Web Services (AWS) Data and Analytics Competency status. This designation recognizes PingCAP’s demonstrated success in helping customers collect, store, govern, and analyze data at any scale.

Quasar announced that it is partnering with PTC on advanced data solutions for industrial companies. The partner activities will focus on Quasar’s data management capabilities and the industrial connectivity capabilities of PTC’s Kepware platform.

Redis announced the integration of Redis Enterprise Cloud’s vector database capabilities with Amazon Bedrock, a service to build generative AI applications with foundation models. The solution allows customers to streamline application development by capitalizing on developer efficiency and scalability of a fully managed, high-performance database while making it easy to use an array of leading foundation models (FMs) via API.

Tangoe announced it has joined the FinOps Foundation community as a Premier member. The FinOps Foundation, an integral part of The Linux Foundation’s non-profit technology consortium, is focused on advancing the people who manage the value of cloud.

WekaIO announced that it is developing a new solution in collaboration with Stability AI. The WEKA Data Platform Converged Mode for Cloud solution will enhance Stability AI’s ability to train multiple AI models in the cloud, including its popular Stable Diffusion model, and extend efficiency, cost, and sustainability benefits to its customers.

If your company has real-time analytics news, send your announcements to [email protected].

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

• Real-time Analytics News for the Week Ending September 9
• Real-time Analytics News for the Week Ending September 2
• Real-time Analytics News for the Week Ending August 26
• Real-time Analytics News for Week Ending August 19
• Real-time Analytics News for Week Ending August 12
• Real-time Analytics News for Week Ending August 5