Tag: AI
All the articles with the tag "AI".
System Thinking for Scaling AI Networks to One Million GPUs
Published: at 12:05 AMThe article highlights the need for system-level thinking when designing networks for million-GPU AI training, urging co-design and application-awareness to overcome limitations of traditional architectures.
Massive Chip Shipment Surge to Malaysia: A Response to AI GPU Smuggling Crackdowns and Tariffs?
Published: at 04:21 PMA 366% surge in chip shipments to Malaysia signals potential rerouting to bypass Nvidia AI GPU smuggling crackdowns and anticipated tariffs, revealing vulnerabilities in global tech supply chains.
Telus Bolsters AI Infrastructure with 500 Nvidia Hopper GPUs in Rimouski Data Center
Published: at 03:33 PMTelus is significantly boosting its AI capabilities by deploying 500 Nvidia Hopper GPUs in its Rimouski data center. This strategic investment aims to meet increasing demand for AI compute resources and solidify Telus' position in the AI market.
AMD Radeon GPUs Get Significant Stable Diffusion Performance Boost with Optimized Models
Published: at 03:20 PMAMD's new, Radeon-optimized Stable Diffusion models deliver significant performance boosts, potentially increasing the appeal of AMD GPUs for AI image generation, and challenging Nvidia's dominance in the AI space.
AMI Data Center Manager Version 6.0 Enhances AI and GPU Management
Published: at 02:38 PMAMI Data Center Manager 6.0 enhances AI and GPU management through improved monitoring, power management, and workload orchestration. This update caters to data centers facing increased AI/HPC demands for improved efficiency.
Understanding TPUs, CPUs, and GPUs: A Comparative Analysis
Published: at 01:04 PMThe article differentiates between CPUs, GPUs, and TPUs, highlighting their unique architectures and optimal use cases. CPUs are general purpose, GPUs excel at parallel processing, and TPUs are specialized for deep learning, enhancing performance for specific AI tasks.
CoreWeave Pioneers with Early Access to NVIDIA Grace Blackwell GPUs
Published: at 12:35 PMCoreWeave is the first cloud provider to offer access to NVIDIA Grace Blackwell GPUs, catering to advanced AI applications. This move positions CoreWeave as an innovator and accelerates AI development.
KDDI, Sakura Internet, and Hi-Res to Establish Joint GPU System in Japan to Boost AI Development
Published: at 11:00 AMKDDI, Sakura Internet, and Hi-Res are joining forces to build a large GPU computing system in Japan, aimed at supporting AI development and reducing reliance on foreign cloud services.
CoreWeave Leverages NVIDIA GB200 NVL72 for Accelerated AI Infrastructure
Published: at 01:36 AMCoreWeave is set to be among the first to offer NVIDIA's GB200 NVL72, a rack-scale AI computing platform, accelerating AI model training and deployment with superior performance and efficiency.
Nvidia Faces $5.5 Billion GPU Write-Off Due to US Export Restrictions to China
Published: at 08:32 PMNvidia is writing off $5.5 billion due to US export restrictions preventing the sale of H20 GPUs to China. This impacts components and purchase commitments, highlighting the risks of geopolitical tensions.