AI data centers

Farewell to Protocol Chaos! Huawei Proposes Unified Bus Architecture to Restructure Million-Chip AI Supernodes

by
Farewell to Protocol Chaos! Huawei Proposes Unified Bus Architecture to Restructure Million-Chip AI Supernodes

Translator’s Note In today’s world, where the demand for AI computing power is growing exponentially, the bottleneck of computing power has gradually shifted from the pure computing capability of chips to the communication efficiency between chips. NVIDIA has built a strong moat with its NVLink and InfiniBand technologies. Huawei’s UB-Mesh, showcased at Hot Chips, is … Read more

PCIe Bus: The High-Speed Data Engine in Communication and Networking

by
PCIe Bus: The High-Speed Data Engine in Communication and Networking

1. Core Advantages of PCIe Technology: Why is it the Cornerstone of Communication Networks? PCIe[Peripheral Component Interconnect Express] is a high-speed serial point-to-point dual-channel bus standard, whose technical characteristics perfectly meet the high bandwidth and low latency requirements of modern communication networks: 1. Full-duplex communication architecture allows each device to have independent transmit (TX) and … Read more

The Second Half of the AI Computing Power Battle: ASIC Chips Become the New Focus! These Listed Companies Are Fully Positioned

by
The Second Half of the AI Computing Power Battle: ASIC Chips Become the New Focus! These Listed Companies Are Fully Positioned

Oracle’s stock surged by36%, with a market value increase of1.78 trillion yuan. This astonishing performance not only shocked global investors but also revealed thatthe competition for AI computing power is shifting from general computing to dedicated chips. On Wednesday, Eastern Time, American tech giant Oracle signed a massive computing power procurement contract withOpenAI, becoming one … Read more

The Top 10 Most Authentic Companies in AI-Catalyzed Storage Chips

by
The Top 10 Most Authentic Companies in AI-Catalyzed Storage Chips

The storage chip industry driven by AI is expected to experience a “simultaneous increase in volume and price” by 2025. Major players like SanDisk and Micron have announced price hikes of 10%-30%. Samsung Electronics has ramped up DRAM production capacity to 60,000 wafers per month, with the explosive demand in the industry pointing directly to … Read more

Semiconductor Industry Shows Signs of Recovery: Key Signals from TI CEO

by
Semiconductor Industry Shows Signs of Recovery: Key Signals from TI CEO

Recently, at the Goldman Sachs Communication and Technology Conference, Texas Instruments (TI) CEO Haviv Ilan provided an in-depth analysis of the semiconductor industry’s development trends and corporate strategic layout. Through a systematic analysis of industry cycles, market segmentation characteristics, and corporate operational strategies, he presented a comprehensive view of industry development and corporate response plans. … Read more

FPGA Discussion: Where is the Massive Bandwidth of 100G Networks Used?

by
FPGA Discussion: Where is the Massive Bandwidth of 100G Networks Used?

Introduction When we talk about 100G networks, many people’s first reaction is: “Isn’t this bandwidth a bit excessive? Can ordinary server CPUs handle it?” Indeed, the gigabit ports on home computers often do not reach full capacity, let alone 100G. However, the reality is that in data centers, AI training clusters, distributed storage, and telecom … Read more

NVIDIA Announces Next-Generation Rubin Chip Has Taped Out

by
NVIDIA Announces Next-Generation Rubin Chip Has Taped Out

Following NVIDIA’s announcement at the end of August regarding the Rubin architecture chip set to enter mass production next year, NVIDIA discussed the progress of Rubin at the Goldman Sachs Technology Conference on September 8. NVIDIA’s Chief Financial Officer, Colette Kress, stated that the Rubin chip is already preparing for market entry,and the Rubin architecture … Read more

NVIDIA’s Next-Generation Rubin Chip Has Entered Tape-Out, Projected Data Center Capital Expenditure Exceeds $3 Trillion Over the Next 5 Years

by
NVIDIA's Next-Generation Rubin Chip Has Entered Tape-Out, Projected Data Center Capital Expenditure Exceeds $3 Trillion Over the Next 5 Years

Reprinted from Caixin, copyright belongs to the original author. Please contact for removal if there is any infringement. Following NVIDIA’s announcement at the end of August regarding the mass production of the Rubin architecture chips next year, NVIDIA discussed the progress of Rubin at the Goldman Sachs Technology Conference on September 8. NVIDIA CFO Colette … Read more

NVIDIA’s Next-Generation Rubin Chip Has Entered Tape-Out

by
NVIDIA's Next-Generation Rubin Chip Has Entered Tape-Out

Following NVIDIA’s announcement at the end of August regarding the mass production of Rubin architecture chips next year, NVIDIA discussed the progress of Rubin at the Goldman Sachs Technology Conference on September 8. NVIDIA CFO Colette Kress stated that the Rubin chips are preparing for market entry, and the Rubin architecture will consist of six … Read more

NVIDIA’s Next-Generation Rubin Chip Has Entered Tape-Out

by
NVIDIA's Next-Generation Rubin Chip Has Entered Tape-Out

Following NVIDIA’s announcement at the end of August regarding the mass production of the Rubin architecture chip planned for next year, NVIDIA discussed the progress of Rubin at the Goldman Sachs Technology Conference on September 8. NVIDIA CFO Colette Kress stated that the Rubin chip is already preparing for market entry, and the Rubin architecture … Read more