Cerebras
August 27, 2024
Cerebras News - August 27, 2024
Cerebras Systems has introduced a new AI inference tool aimed at competing with Nvidia's market dominance. The company also launched what it claims to be the world's fastest AI inference service. Learn about the technology behind these innovations and their potential impact on the AI industry.
() Introduction
() Cerebras Unveils New AI Inference Tool to Challenge Nvidia
() Cerebras Systems Launches 'World's Fastest' AI Inference Service
Cerebras Systems has launched a new AI inference tool designed to challenge Nvidia's dominance in the AI market. The tool leverages Cerebras' Wafer Scale Engines, offering a more cost-effective and efficient alternative to Nvidia GPUs for AI applications. The company plans to provide the tool through a developer key, its cloud service, and direct sales to customers managing their own data centers. Cerebras aims to charge as little as 10 cents per million tokens, significantly undercutting Nvidia's pricing. Additionally, Cerebras has filed a confidential prospectus with the SEC, indicating plans to go public. The new service, termed 'high-speed inference,' promises 1,000-tokens-per-second speeds and is powered by the WSE-3 processor, which boasts 1.4 trillion transistors and 44 gigabytes of onboard memory. This service is claimed to be up to 20 times faster than Nvidia's comparable offerings and is available in three tiers: free, Developer, and Enterprise. Early-access customers include GlaxoSmithKline, Perplexity AI, and Meter Inc., with discussions ongoing with multiple hyperscalers.
