Cerebras (CBRS) Stock Plummets 9% on OpenAI Nvidia Partnership Claims

TLDR

• Shares of Cerebras declined 7-9% following allegations that OpenAI employs Nvidia processors rather than Cerebras technology for its “Ultrafast” AI service tier.
• According to SemiAnalysis, OpenAI’s GPT-6.1 Sol Ultrafast operates on Nvidia GPUs with low batch processing, not Cerebras’ Wafer-Scale Engine technology.
• Cerebras counts OpenAI as its biggest revenue source in the backlog pipeline, with both companies having publicly declared Cerebras would support Ultrafast mode last August.
• Both Cerebras and OpenAI have remained silent on the allegations.
• CBRS shares have been volatile since debuting at $350 during its May public offering.

Shares of Cerebras Systems experienced a significant selloff this week following a report that cast doubt on the chip maker’s involvement in running OpenAI’s high-speed AI model. Wednesday’s trading session saw the stock tumble as much as 8.87%.

Cerebras Systems Inc., CBRS

The selloff was sparked by semiconductor analysis firm SemiAnalysis, which claimed via social media that OpenAI’s latest GPT-6.1 Sol Ultrafast model operates on conventional Nvidia GPUs rather than Cerebras’ proprietary processors.

This assertion carries significant weight. When OpenAI unveiled its “Ultrafast” service tier, it promoted performance improvements up to eight times faster than standard offerings, delivering approximately 300 tokens per second. Cerebras had been publicly identified as the technology partner enabling these speed gains.

Cerebras has centered its entire value proposition on addressing precisely this type of challenge. The company’s Wafer-Scale Engine represents an enormous chip housing billions of processing cores and extensive memory capacity on a single silicon wafer. This architecture eliminates latency issues associated with transferring data across hundreds of discrete GPUs.

Breaking Down the SemiAnalysis Claims

A critical element in the report centers on one specific phrase: “low batch size.” In AI inference operations, batching refers to processing multiple requests simultaneously to maximize GPU utilization.

Nvidia processors typically require large batch sizes to achieve optimal performance levels. Low batch processing means handling fewer simultaneous requests, prioritizing response speed over computational efficiency. This represents the precise market segment Cerebras has targeted.

Should OpenAI have successfully configured a premium model on Nvidia infrastructure at low batch sizes while maintaining cost effectiveness, it undermines a core advantage Cerebras has promoted. This explains the swift market reaction.

Cerebras shares settled around $180.22 on Wednesday following the news circulation. This marks a substantial decline from the company’s $350 IPO valuation established in May.

The public offering generated considerable buzz. Market participants showed enthusiasm for a semiconductor company positioning itself to surpass Nvidia on inference speed for specific computing tasks.

That initial optimism has waned considerably. Financial results released following the IPO have not sustained the early investor excitement.

The Critical Importance of the OpenAI Relationship

OpenAI represents far more than just another customer for Cerebras. According to Barron’s, it stands as the company’s most significant client measured by revenue backlog.

Any diminished role with OpenAI, regardless of duration, creates legitimate investor concerns. Cerebras has relied substantially on its OpenAI partnership to support its market valuation.

Neither company has issued public statements addressing the SemiAnalysis claims. This lack of communication has left market analysts speculating without official confirmation.

The situation may not be permanent. Infrastructure capacity limitations or technical optimization challenges could explain the current Nvidia chip deployment.

Cerebras might eventually power the Ultrafast mode after resolving any technical challenges. Nothing from either organization has definitively eliminated that possibility.

Currently, the market response reflects uncertainty rather than definitively negative information. Investors are accounting for the risk that Nvidia’s CUDA software ecosystem and hardware advancements make it challenging for competitors to secure exclusive partnerships with leading AI research organizations.

Cerebras has not released any communication confirming or refuting its present involvement in OpenAI’s infrastructure strategy. As of Wednesday’s market close, the stock remained down approximately 8% for the trading day.

The post Cerebras (CBRS) Stock Plummets 9% on OpenAI Nvidia Partnership Claims appeared first on Blockonomi.