As the use cases for artificial intelligence continue to grow, advanced AI performance is becoming increasingly essential. High-performance AI systems are critical for solving complex problems, supporting human decision-making and adapting to dynamic real-world situations efficiently and effectively.

Recognizing this need, Cerebras Systems Inc. recently unveiled Qwen3-32B, a cutting-edge, open-weight large language model that delivers sophisticated reasoning and real-time responsiveness, offering unmatched speed and performance, according to Andrew Feldman (pictured), founder and chief executive officer of Cerebras.

Cerebras’ Andrew Feldman talks to theCUBE about the need for high AI performance.

“We’re putting out a new model, Qwen 32B, and as with all our models, it is the fastest in the industry,” Feldman said. “Somewhere between 20 and 30 times faster than the fastest GPU offering. Qwen’s a really interesting model. It’s a reasoning model, and it’s one of the smartest models. It behaves in line with the leading closed source models, while it’s an open-source model. So, a fast model, a smart model. I think it’ll give some open AI models a real run for their money.”

Feldman spoke with theCUBE’s John Furrier at the Cerebras Supernova event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed why high AI performance matters — and how Cerebras is delivering it with Qwen3-32B. (* Disclosure below.)

The relationship between reasoning models and AI performance

Reasoning models are crucial to AI performance because they allow systems to go beyond simple pattern recognition, enabling logical, context-aware decision-making. This enhances the model’s cognitive capabilities, leading to more intelligent behavior, as showcased by Qwen 3-32B, Feldman pointed out. 

“Reasoning models behave much more like humans,” he said. “If you set out to write a paper, you don’t just disgorge the paper. You don’t just think for a second and then write the whole paper. Likely you’d write a draft, you’d edit it, you’d think about it again, you’d review it and you would iterate. Reasoning models perform that workflow. They develop an answer, they review it, they improve it, and this requires more work.”

By offering specialized AI hardware solutions that deliver faster performance and greater efficiency, Cerebras continues to offer tailormade solutions. This is why the company has partnered with IBM Corp. and Meta Platforms Inc. to enable faster AI inference, according to Feldman. 

“Meta announced that they would be providing an API space service,” he said. “You could go to their API and it would serve Llama 4, and behind their API was Cerebras. This was the first hyperscaler win for Cerebras. Immediately following that, IBM announced that they would be doing approximately the same thing. We would be providing fast inference to their clients. We have expanded our reach from the developer all the way to the largest financial institutions in the world through IBM.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the Cerebras Supernova event: 

Photo: SiliconANGLE