Tag: AI
All the articles with the tag "AI".
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.
Trump Slams Nvidia's US-Bound Chip as "Weak" and "Insufficient"
Published: at 08:16 PMThe satirical article imagines a future where Trump criticizes Nvidia's chip designed for the US market as weak, highlighting the tension between US-China tech competition and Nvidia's business interests.
Nvidia Faces $5.5 Billion Hit From AI GPU Export Restrictions to China
Published: at 06:22 PMNew US export restrictions on advanced AI GPUs to China could cost Nvidia $5.5 billion in revenue. The move highlights the escalating tech tensions between the two nations and their impact on the semiconductor industry.
US Government Restrictions Lead to $800 Million Revenue Loss for AMD
Published: at 05:31 PMUS government restrictions on AI chip exports to China are costing AMD $800 million in revenue, disrupting their plans to serve the Chinese market with modified GPUs and forcing them to seek licenses.
CoreWeave Launches NVIDIA Grace Blackwell GB200 Systems, Becoming First Cloud Provider with General Availability
Published: at 12:06 PMCoreWeave becomes one of the first cloud providers offering NVIDIA's new GB200 Grace Blackwell systems at scale, providing major AI firms like Cohere, IBM, and Mistral AI with powerful infrastructure for next-gen AI model development and deployment.
CoreWeave Adopts NVIDIA GB200 NVL72 for Next-Gen AI Infrastructure
Published: at 10:42 AMCoreWeave is set to deploy NVIDIA's GB200 NVL72, a powerful, liquid-cooled system integrating Grace Hopper and Blackwell superchips, to significantly boost its infrastructure for demanding AI workloads like large language models.
The Radically Different NVIDIA HGX B300 NVL16
Published: at 10:02 AMNVIDIA's HGX B300 NVL16 platform introduces a major design shift by integrating ConnectX-8 NICs, streamlining the architecture but also mandating NVIDIA's networking for internal GPU communication, impacting ecosystem flexibility.
NVIDIA Announces Blackwell-Powered GeForce RTX 5060 Series Starting at $299
Published: at 03:05 PMNVIDIA officially announces the GeForce RTX 5060 series, powered by the Blackwell architecture, starting at $299 for the 8GB RTX 5060, with RTX 5060 Ti models (8GB at $379, 16GB at $429) emphasizing performance gains and enhanced AI features like DLSS and new generative AI capabilities for mainstream gamers.