45k user reviews & 1 million votes, over 29k web hosting brands & 85k plans since 2004!
X
Menu
WHTop → 📋 News → Vertiv Adds ThermoKey After April Liquid Cooling Acquisition
Vertiv Adds ThermoKey After April Liquid Cooling Acquisition
📅 - Vertiv has completed its acquisition of ThermoKey, the Italian heat-rejection specialist, adding manufacturing depth in Europe after buying Strategic Thermal Labs in April for liquid-cooling engineering. The sequence gives Vertiv more control over the AI data center cooling chain, from chip-level heat capture to facility-scale rejection, as density, permitting, and power constraints worsen globally across key markets.
Financial terms were not disclosed. ThermoKey, founded in 1991, makes heat exchangers, dry coolers, and related heat-rejection systems. Its operations in Rivarotta, Italy, will continue as a manufacturing, engineering, and support hub, according to Vertiv. Giuseppe [...][... Check source for end of article ...]
📅 - Take Control of Your Live Streams: Set Up a Free RTMP Server - Relying on third-party platforms for live broadcasting often means dealing with strict limitations, unexpected costs, and a lack of control over your data. By setting up your own Real-Time Messaging Protocol (RTMP) server on a dedicated server, you bypass these restrictions entirely. Our latest guide at iDatam walks you through how to build a free RTMP server using NGINX in just seven straightforward steps.
Why Build Your Own RTMP Server?
Total Ownership: You own the stream and the data. No unexpected platform bans, and absolutely no arbitrary bitrate caps to throttle your visual quality.
Cost-Effective: By utilizing the open-source NGINX RTMP module, you completely avoid expensive monthl
📅 - Comprehensive Guide: Diagnosing and Fixing Dedicated Server Performance Issues - A dedicated server is the backbone of your digital operations. However, even the most robust servers can experience performance degradation over time. If your applications are lagging and load times are increasing, it is time to investigate. Our comprehensive guide on iDatam breaks down the root causes of server slowdowns and provides tips to restore peak performance.
1. Software Inefficiencies & Bottlenecks Often, the hardware is fine, but the software is poorly optimized.
Unoptimized Databases: Heavy, unindexed SQL queries cripple server speed. Use slow query logs to identify the culprits.
Memory Leaks: Poorly written code can consume all available RAM, forcing the server to use ext
📅 - Zscaler Report Finds AI Is Rewriting Phishing Economics - Zscaler says phishing volumes fell for a second straight year, but its 2026 ThreatLabz report argues attackers are becoming more efficient, using AI-generated sites, encrypted delivery and session hijacking kits to improve initial access campaigns while probing enterprise identity and collaboration systems before compromise. The finding complicates how security teams read declining attack counts today, globally, materially.
A 20% year-over-year drop in phishing volume would normally sound like progress. It may still be progress in some places. But Zscaler's reading is less comfortable: attackers are sending fewer crude lures and spending more effort on attacks that look cleaner, move [...]
📅 - US Data Center Tax Breaks At Risk as AI Strains Grids, Budgets - US State tax breaks that helped turn data centers into one of America's hottest infrastructure categories are coming under pressure as lawmakers question whether subsidies built for early cloud projects still make sense in an AI market consuming more power, land, water, and political goodwill than many communities expected when those incentives were written into law during cloud cycles.
The fight is no longer limited to one angry county board meeting or one proposed transmission corridor. It has moved into state budgets.
Illinois Governor J.B. Pritzker has moved to halt data center tax breaks. Minnesota ended incentives last year. Washington State is set to let its program expire on July [...]
📅 - QumulusAI Lands $124M for Blackwell AI Inference Cloud Deals - QumulusAI says it has signed more than $124 million in three-year GPU-as-a-service subscriptions, a midmarket AI infrastructure win tied to 1,280 NVIDIA Blackwell GPUs. The contracts, including nearly $21.9 million upfront, put fresh numbers around a less glamorous but increasingly decisive market: production inference, where cost per request matters as much as raw accelerator access globally now.
The company named Hyperbolic as one of the customers. The second was described only as another leading AI inference platform. That is not unusual in this part of the infrastructure market. It is also inconvenient, because customer concentration and contract quality matter when the assets [...]
📅 - Nokia Brings Agentic AI to IP Network Operations Platform - Nokia is adding an agentic AI framework to its Network Services Platform, giving telecom operators a controlled way to test AI-driven operations in live IP networks. The framework, due commercially by the end of 2026, starts with troubleshooting and is built around trusted network data, operator policies, and explainable actions rather than free-running automation claims.
The announcement lands in a market that wants autonomous networks but does not quite trust them yet. Not at scale. Not when a bad decision can break enterprise services, mobile backhaul, cloud connectivity, or wholesale IP routes across several domains.
Telecom operators already use automation. They also already [...]