Nvidia CEO Says ‘Reasoning’ AI Will Depend on Cheaper Computing

Nvidia CEO Jensen Huang has stated that the future of Artificial Intelligence (AI) will involve services capable of “reasoning.” However, achieving this stage will require a significant reduction in computing costs.

Huang explained during a podcast with Arm Holdings Plc CEO Rene Haas that future AI tools will be able to process queries by taking hundreds or even thousands of steps, reflecting on their conclusions. This capability will distinguish them from current systems like OpenAI’s ChatGPT, which Huang uses daily.

Nvidia aims to drive these advancements by increasing its chip performance by two to three times annually while maintaining the same cost and energy consumption levels. This will revolutionize how AI systems handle inference, the process of recognizing patterns and making decisions based on data.

“We’re able to drive incredible cost reduction for intelligence,” Huang said. “We all realize the value of this. If we can significantly lower the cost, we can enable reasoning during inference.”

Nvidia, based in Santa Clara, California, currently dominates the market for accelerator chips—processors that enhance AI workloads. The company has also expanded into selling computers, software, AI models, and networking services to encourage more companies to adopt AI technology.

Despite its stronghold, Nvidia is facing competition. Data center operators like Amazon’s AWS and Microsoft are working on in-house alternatives, while Advanced Micro Devices (AMD), a long-time competitor in gaming chips, is emerging as a contender in the AI space. AMD is expected to provide updates on its AI products at an event on Thursday.

LPG Users Alert: Complete eKYC by August 16 Magic Photo Editor drops ₹12,579 before Pixel 11 launch! Jio’s Best Offer: Free WiFi for 1 Month, 100Mbps Internet!