- Nvidia partnered with SK hynix to develop advanced AI chips.
- SK Hynix plans to accelerate the development of HBM chips.
What happened
Nvidia‘s CEO emphasized the importance of its relationship with SK Hynix to build more advanced AI accelerators, stating that “aggressive” development of enhanced high bandwidth memory (HBM) processors is required. Huang described SK hynix’s HBM development goal as “super aggressive” but “super necessary” at the same time, as AI is growing from a text-generating model to system models that retrieve a broader collection of working data, including images and videos.
SK Hynix was the first chipmaker in the world to offer Nvidia with fifth-generation HBM3E chips, which are used in graphic processing units for AI computing. “The roadmap of high bandwidth memory is excellent, but frankly, we wish we got more bandwidth with lower energy,” Nvidia CEO Jensen Huang stated this in video footage during the SK Group’s annual technology conference, SK AI Summit 2024, this year.
Also read:Startup Lightmatter raises $400m advancing revolutionary photonic AI chips
Also read:Intel’s AI-chip aspirations dim amidst competitive storm
Why it’s important
Nvidia’s partnership with SK Hynix to develop advanced AI chips marks a significant step forward in high-performance computing capabilities. Nvidia aims to leverage SK Hynix’s expertise in memory solutions to create AI chips with enhanced processing power and efficiency. By collaborating on AI-focused hardware, both companies can meet the rapidly growing demand for faster and more powerful chips in AI applications.
SK Hynix plans to accelerate development of high-bandwidth memory (HBM) chips, crucial for handling large data loads in AI operations. HBM chips allow faster data transfer and processing speeds, making them essential for demanding AI workloads like machine learning and data analysis. SK Hynix’s HBM development aligns with Nvidia’s AI goals, supporting their joint push toward innovation in AI hardware solutions.
This partnership could set new standards in AI chip performance, helping Nvidia and SK Hynix lead the market. By advancing AI hardware, they empower companies worldwide to implement AI-driven technologies more effectively and efficiently.