GPU– tag –
-
Novel Photonic AI Accelerator Achieves 70% Power Reduction Over GPUs for Edge AI Breakthrough
YouTube Global Overview A new photonic AI accelerator has achieved a significant breakthrough, enabling complex simulations on edge devices while reducing power consumption by approximately 70% compared to GPUs. This technology fuses lig... -
Broadcom Unveils Industry’s First 400G/Lane Optical DSP ‘Taurus’ at OFC 2026, Paving Way for 200T AI Era with Cost-Effective 1.6T Transceivers
Broadcom Inc. USA Overview Broadcom introduced the industry's first 400G/lane optical DSP, 'Taurus,' at OFC 2026, showcasing cutting-edge solutions for scaling AI infrastructure. This groundbreaking DSP, when combined with Broadcom's hig... -
Marvell Technology Enters S&P 500, Stock Triples in 2026 as Data Center Business Drives 76% of Revenue, Led by AI Optical Modules and 51.2T Ethernet Switches
Investing.com AU Australia Overview Marvell Technology has been added to the S&P 500 index, with its stock nearly tripling in 2026, demonstrating rapid growth fueled by AI infrastructure investments. The company reported that 76% of ... -
CoreWeave to Construct Three AI Data Centers in Indonesia by 2028, Expanding into APAC Market and Driving AI Storage Demand
Converge Digest USA Overview CoreWeave announced plans to build three AI data center facilities in Indonesia by 2028, marking its first expansion into the Asia-Pacific region. This initiative will add a total of 360MW of IT load capacity... -
NTT Achieves Record Q1 2026 Revenue, Accelerates IOWN Global Expansion with APN in Chile, PEC-2 Switches in India, and GPU over APN Testbed in Japan
Investing.com Japan Overview NTT reported record revenues in Q1 2026, propelling accelerated global deployment of its IOWN technology, particularly the All-Photonics Network (APN) in Chile for remote mining operations and Photonic-Electr... -
Nvidia and Groq AI Accelerators Boost AI Application Performance by up to 1000x, Expediting Next-Gen AI Model Development
Grooper USA Overview AI accelerators are high-performance hardware or software systems designed to dramatically speed up AI application processing. Leading chipmakers like Nvidia and Groq are manufacturing specialized hardware capable of... -
AI Chip HBM Integration to Drive 29% ABF Substrate Shortage by 2028; $19.1B Expansion Unlikely to Offer Short-Term Relief
Unknown Taiwan Overview The rapid integration of High Bandwidth Memory (HBM) into AI chips is set to drive a 39% Compound Annual Growth Rate (CAGR) in ABF substrate demand between 2025 and 2028. This surge is projected to create a supply... -
Cisco-Backed Eliyan Raises $145M, Achieves $1B Valuation to Solve AI Interconnect Bottlenecks with Electro-Optical Solutions
Tech Funding News UK Overview Silicon Valley startup Eliyan has raised $145 million, including strategic investments from Cisco and Lumentum, pushing its valuation to $1 billion. This funding will accelerate its expansion into electro-op... -
AI Data Centers Rapidly Transition to Photonic Interconnects Beyond Copper Limits, Driven by 800G/1.6T Devices and OCI MSA Standardization
Tom's Hardware USA Overview AI data centers are rapidly migrating to photonic interconnects due to copper's physical limitations, with accelerating shipments of 800G pluggable devices and the introduction of 1.6T modules. The Optical Com... -
Cisco’s Bill Gartner on AI Data Center Optics Economics: 1.6T Devices Key to Unlocking Scalability, Not a Bottleneck
Light Reading USA Overview Cisco's Bill Gartner discussed the economic significance of 400G+ optical transceivers in AI data centers, outlining three tiers of AI infrastructure: scale-up (intra-rack), scale-out (inter-rack with 400G/800G...
