InfoQ Homepage News
-
AWS Unveils Gemini, a Distributed Training System for Swift Failure Recovery in Large Model Training
AWS and Rice University have introduced Gemini, a new distributed training system to redefine failure recovery in large-scale deep learning models. According to the research paper, Gemini adopts a daring strategy by utilizing CPU memory to ensure previously unheard-of speeds in failure recovery, overcoming obstacles related to high recovery costs and constrained checkpoint storage capacity.
-
Azure Synapse Link for Cosmos DB: New Analytics Capabilities
In recent news from the Azure team, Azure Synapse Link introduces a set of new capabilities, including its compatibility with existing MongoDB collections, integration with continuous backup, and custom partitioning, empowering users to leverage analytics workloads with ease and efficiency on top of Azure Cosmos DB data.
-
How Good Companies Can Leverage Agile to Fight Civilizational Debt
Growth, profit, and shareholder value are the cornerstones of today’s economic system, which according to Piotr Trojanowski have proven outdated, reductionistic and not sustainable. He proposes taking the cost of growth into account by using the concept of civilizational debt in agile transformations, and applying agile to realizing humankind's mission through our work.
-
Cohere Unveils Advanced Embedding Model Embed v3
Cohere has unveiled Embed v3, their most advanced embedding model designed to transform semantic search and generative AI.
-
Microsoft Introduces New Tiers for Its API Management Service in Azure
Microsoft recently announced the public preview of two new Azure API Management tiers: Basic v2 and Standard v2.
-
Mojo Language SDK Available: Mojo Driver, VS Code extension, and Jupyter Kernel
Mojo SDK is available for developers. It contains the mojo driver, the Visual Studio Code extension and the Jupyter kernel. For now, SDK is available for MacOS and Linux.
-
WildFly 30 Application Server Gets Ready for JDK 21
WildFly, an application framework written in Java and developed and maintained by Red Hat, released its latest version 30 in October 2023. The key focus of this release was to prepare for the latest LTS Java version 21, one feature request and several dependency upgrades and bug fixes.
-
AI Researchers Improve LLM-Based Reasoning by Mimicking Learning from Mistakes
Researchers from Microsoft, Peking University, and Xi’an Jiaotong University claim to have developed a technique to improve large language models' (LLMs) ability to solve math problems by replicating how humans learn from their own mistakes.
-
AWS Restructures and Consolidates Its Well-Architected Framework
AWS published a new set of updates to its Well-Architected Framework, with changes across all six pillars of the framework. The performance efficiency and operational excellence pillars have been restructured and consolidated to reduce the number of best practices. Other pillars received improved implementation guidance, including recommendations and steps on reusable architecture patterns.
-
OpenAI Announces New Models and APIs at First Developer Day Conference
OpenAI announced additions and price reductions across its platform at its first Developer Day. The updates include the introduction of a new GPT-4 Turbo model, an Assistants API, and multimodal capabilities, among others.
-
Microsoft Releases DeepSpeed-FastGen for High-Throughput Text Generation
Microsoft has announced the alpha release of DeepSpeed-FastGen, a system designed to improve the deployment and serving of large language models (LLMs). DeepSpeed-FastGen is the synergistic composition of DeepSpeed-MII and DeepSpeed-Inference . DeepSpeed-FastGen is based on the Dynamic SplitFuse technique. The system currently supports several model architectures.
-
Jina AI's Open-Source Embedding Model Outperforms OpenAI's Ada
Multimodal AI company Jina AI recently released jina-embeddings-v2, a sentence embedding model. The model supports context lengths up to 8192 tokens and outperforms OpenAI's text-embedding-ada-002 on several embedding benchmarks.
-
Seven Essential Tracks at QCon London 2024: GenAI, FinTech, Platform Engineering & More!
InfoQ’s international software development conference, QCon London, returns on April 8-10, 2024. The conference will feature 15 carefully curated tracks and 60 technical talks over 3 days.
-
Polly v8 .NET Resilience Library: Resilience Pipelines, Built-in Telemetry, and More
Polly v8 is officially released. This version brings enhancements such as resilience pipelines, built-in telemetry support, and some changes within the configuration for individual resilience strategies.
-
eBPF Kubernetes Security Tool Tetragon Improves Performance and Stability
Isovalent has announced the 1.0 release of Cilium Tetragon, their eBPF-based Kubernetes security observability and runtime enforcement tool. Policies and filters can be applied directly via eBPF to monitor process execution, privilege escalations, and file and network activity.