The NVIDIA GPU Technology Conference (GTC) 2024 will take place on March 18, kicking off with CEO Jensen Huang’s opening keynote address. The event promises to unveil groundbreaking developments in artificial intelligence (AI) and high-performance computing (HPC), particularly in relation to Blackwell AI/HPC graphics processing units (GPUs). The focus will be on innovations and technologies designed to push the boundaries of graphics processing units and shape the future of computing power. Attendees can look forward to interesting announcements and insights into the latest advancements in GPU technology.
NVIDIA’s keynote on August 2, 2023 will focus on artificial intelligence (AI). The company has spent the past few months announcing its new “Blackwell” GPUs for the HPC and AI industries, which will mark the transition to a new generation of AI computing chips. The Blackwell GPUs will be based on TSMC’s 3nm process and feature a chiplet design. This means that the GPUs will consist of several smaller chips that are connected together. A chiplet design offers a number of advantages, including higher performance, better scalability and lower power consumption. The Blackwell GPUs will replace NVIDIA’s Hopper GPUs, which are currently the most powerful AI chips on the market. Production of the Blackwell GPUs is expected to begin in the second half of 2024.
It is unlikely that the company will announce new consumer GPUs during the conference. NVIDIA’s gaming-focused “Ada Lovelace-Next” GPUs are expected to hit the market sometime in 2025. However, recent GTCs have been heavily focused on AI and HPC, with few announcements of consumer software. Still, GTC 2024 will generate some excitement. NVIDIA is, after all, a leading provider of AI and HPC hardware, and the announcements at the conference will likely shape the future of the industry.
Source: NVIDIA
Kommentieren