NVIDIA– tag –
-
Qualcomm Unveils AI Inference-Optimized Data Center Products “Dragonfly,” Boosting Power Efficiency 18x with AI250 Rack and HBC Gen 1
Qualcomm USA Overview Qualcomm has launched "Qualcomm Dragonfly," a data center product line optimized for AI inference workloads. This line includes the Qualcomm Dragonfly AI250 rack and Qualcomm High Bandwidth Compute (HBC) Gen 1, whic... -
AMD’s Supercomputing Gains and Instinct GPUs Accelerate AI Data Center Growth with Meta’s 6-Gigawatt Deployment
Yahoo Finance USA Overview AMD's growth in supercomputing and the expansion of AI data centers with Instinct GPUs are fueling its next phase of AI-driven growth. AMD EPYC CPUs and Instinct GPUs power four of the world's top ten fastest s... -
Tesla Plans to Sell Modular AI Data Center Hardware ‘Megapod,’ Entering Nvidia-Dominated AI Market
Electrek USA Overview Tesla has revealed plans to sell modular AI data center hardware, dubbed 'Megapod,' through a new trademark application. Megapod is a self-contained computing system tailored for AI workloads, comprising computer se... -
2026 AI Chip Market Sees 44.6% Surge in Cloud Provider Custom ASIC Shipments, Challenging NVIDIA’s GPU Dominance
AIMultiple International Overview In the 2026 AI chip market, while NVIDIA's GPUs remain central, custom ASIC shipments from cloud providers are surging 44.6% year-over-year, reshaping the market structure. AIMultiple's analysis shows AS... -
OpenAI and Broadcom Unveil LLM-Optimized AI Chip “Jalapeño,” Significantly Outperforming State-of-the-Art in Performance per Watt
OpenAI / Broadcom USA Overview OpenAI and Broadcom have announced "Jalapeño," a custom AI chip specifically designed for Large Language Model (LLM) inference. This chip significantly surpasses current state-of-the-art products in perform... -
SK hynix Begins Shipping Next-Gen HBM4E AI Memory Chips with 12-Layer Stack, 48GB Capacity, 16Gbps per Pin, and Over 20% Power Efficiency Boost
Bisinfotech South Korea Overview SK hynix has commenced shipments of its next-generation HBM4E AI memory chips, featuring a 12-layer stack with 48GB capacity, ultra-fast data transfer rates of up to 16Gbps per pin, and over 20% improved ... -
Samsung Considers New HBM-Focused Advanced Semiconductor Packaging Plant in South Korea’s Honam Region to Meet AI Server Demand
SamMobile South Korea Overview Samsung Electronics is exploring the construction of a new advanced semiconductor packaging facility in South Korea's Honam region, with Gwangju emerging as a prime candidate. This project aims to bolster S... -
Micron’s HBM Capacity Sold Out Through 2026, Securing Position in NVIDIA’s Next-Gen AI Platform “Vera Rubin”
Investing.com USA Overview Micron Technology has announced that its high-bandwidth memory (HBM) production capacity is completely sold out through 2026, with long-term contracts secured, reflecting overwhelming demand for AI-driven HBM. ... -
SK Hynix Begins Shipping 12-Layer Next-Gen ‘HBM4E’ Samples for AI, Delivering Significant Performance and Power Efficiency Gains Amidst NVIDIA CEO’s Endorsement
SK hynix Inc. South Korea Overview SK Hynix has commenced shipping samples of its 12-layer 'HBM4E,' a next-generation DRAM for AI, to key customers. This HBM4E features data processing speeds of up to 16Gbps per pin and over 20% improved... -
Samsung’s Advanced Packaging Lag Clouds AI Chip Comeback, Falling Behind TSMC and Intel
digitimes Taiwan Overview Despite efforts in HBM and foundry services, Samsung Electronics' lagging advanced packaging capabilities remain a significant weakness in the AI chip supply chain. Industry sources and Korean media reports indi...
