InfoQ Homepage News
-
OpenAI Releases ChatGPT Search Feature
OpenAI recently released ChatGPT Search which allows ChatGPT to search the web when answering user questions. Instead of being limited to knowledge available at the time of training, ChatGPT can now incorporate current information from the web and include links to its sources.
-
Meta MobileLLM Advances LLM Design for On-Device Use Cases
With MobileLLM, Meta researchers aim to show that, for smaller models, quality is not a direct product of how many billions parameters they have; rather, it is the result of carefully designing their architecture. To prove their point, they coupled deep and thin architectures with embedding sharing and grouped-query attention mechanisms to improve accuracy over prior state-of-the-art models.
-
Java Evolves to Tackle Pinning with JEP 491
JEP 491, "Synchronize Virtual Threads without Pinning," is now Proposed to Target for JDK 24. This initiative aims to resolve Java's "pinning" issue, a limitation of the synchronized keyword, which restricts scalability with virtual threads. JEP 491 eliminates this constraint, unlocking high-performance concurrency without extensive refactoring.
-
Apache Tomcat 11.0 Delivers Support for Virtual Threads and Jakarta EE 11
Apache Tomcat 11 represents a pivotal advancement in web server technology, supporting Jakarta EE 11 and featuring virtual threads for efficient concurrency via Project Loom. Enhanced WebSocket performance, improved asynchronous processing, and robust security measures make it ideal for modern applications. Transitioning requires namespace adjustments, but offers tools for a seamless migration.
-
Microsoft Introduces Vector Data Abstractions Library for .NET
On October 29th 2024, Microsoft released Microsoft.Extensions.VectorData.Abstractions library for .NET in preview. It makes it easier to integrate .NET solutions with the AI Semantic Kernel SDK, using abstractions over concrete AI implementations and models.
-
Meta AI Introduces Thought Preference Optimization Enabling AI Models to Think Before Responding
Researchers from Meta FAIR, the University of California, Berkeley, and New York University have introduced Thought Preference Optimization (TPO), a new method aimed at improving the response quality of instruction-fine tuned LLMs.
-
Java News Roundup: Jakarta EE 11, GlassFish 8.0-M8, JReleaser 1.15, JHipster 8.7.3, Quarkus 3.16
This week's Java roundup for October 28th, 2024 features news highlighting: an update to the upcoming release of Jakarta EE; the eighth milestone release of GlassFish 8.0; and point releases of JReleaser 1.15.0, JHipster 8.7.3 and Quarkus 3.16.0.
-
OpenJDK News Roundup: Compact Object Headers, Synch Virtual Threads, Disable Security Manager
There was a flurry of activity in the OpenJDK ecosystem in which: four JEPs, having successfully completed their respective reviews, are now Targeted for JDK 24; and five new JEPs have been Proposed to Target for JDK 24 and will be under review during the week of November 4, 2024.
-
Cloudflare Introduces Short-Lived SSH Access, Eliminating the Need for SSH Credentials
Cloudflare recently announced Access for Infrastructure SSH, a feature that replaces traditional SSH keys with short-lived certificates. The new option leverages BastionZero’s integration into Cloudflare One and reduces the complexity of managing SSH keys while enhancing security by substituting long-term SSH keys with temporary, ephemeral certificates.
-
Software Architecture Tracks at QCon San Francisco 2024 – Navigating Current Challenges and Trends
At QCon San Francisco 2024, software architecture is front and center, with two tracks dedicated to exploring some of the largest and most complex architectures today. Join senior software practitioners as they provide inspiration and practical lessons for architects seeking to tackle issues at a massive scale.
-
Cloudflare Introduces Workflows for Building Scalable Resilient Multi-Step Applications
Cloudflare's "Workflows" is a revolutionary execution engine in open beta, enabling developers to build scalable, multi-step applications that autonomously manage errors and state across failures. With seamless retries and modular components, it streamlines development and enhances resource efficiency. Harness the power of Cloudflare's ecosystem for robust app performance.
-
PostgreSQL 17 Released with Improved Vacuum Process and Performance Gains
The PostgreSQL Global Development Group recently announced the general availability of PostgreSQL 17, the latest version of the popular open-source database. This release focuses on performance improvements, including a new memory management implementation for vacuum, storage access optimizations, and enhancements for high-concurrency workloads.
-
Staying Innovative on a Journey from Start-Up to Scale-Up
As ClearBank grew, it faced the challenge of maintaining its innovative culture while integrating more structured processes to manage its expanding operations and ensure regulatory compliance. Within boundaries of accountability and responsibility, teams were given space to evolve their own areas, innovate a little, experiment, and continuously improve, to remain innovative.
-
Meta Spirit LM Integrates Speech and Text in New Multimodal GenAI Model
Presented in a recent paper, Spirit LM enables the creation of pipelines that mixes spoken and written text to integrate speech and text in the same multimodal model. According to Meta, their novel approach, based on interleaving text and speech tokens, makes it possible to circumvent the inherent limitations of prior solutions that use distinct pipelines for speech and text.
-
Slack Evolves Chef Architecture for Safety and Stability
In a recent blog post, Slack Engineering has detailed significant improvements to its Chef infrastructure. This manages tens of thousands of EC2 instances running its services, databases, and applications, and it recently moved from a single Chef stack to a more resilient, sharded infrastructure.