AI 热点

Palo Alto Networks 同意以 $4.5 billion 收购 Applied Compute

SiliconANGLE更新于 11月19日 07:06

Palo Alto Networks bets $4.5B on AI startup Applied Compute as earnings top forecasts - SiliconANGLE

AI 导读

Palo Alto Networks 同意以全股票交易方式收购 AI 模型训练与推理初创公司 Applied Compute,交易估值为 $4.5 billion,尚待监管批准。交易预计于其 fiscal third quarter 完成,Applied Compute 联合创始人兼 CEO Yash Patil 将出任 Palo Alto Networks 联席 AI 官,研究团队也将加入。Applied Compute 提供开放权重模型的后训练与推理平台,包含自动扩缩容推理引擎和名为 Shift-1 的模型路由器。

推荐理由

AI 评分 73

交易呈现了网络安全厂商向模型后训练与推理基础设施扩张的路径,以及其对企业混合使用开放和闭源模型的判断。

主题

标签

正文

Cybersecurity company Palo Alto Networks Inc. today announced that it has agreed to acquire artificial intelligence model training and inference startup Applied Compute Inc. in an all-stock deal valued at $4.5 billion.

The news came alongside fiscal first-quarter earnings and revenue that were well ahead of Wall Street expectations. Shares of Palo Alto Networks rose more than 8% in late trading, adding to a gain of nearly 10% during the regular session.

The acquisition, subject to regulatory clearance, is expected to close in Palo Alto Networks’ fiscal third quarter. Once the deal closes, Applied Compute co-founder and Chief Executive Yash Patil will become co-chief AI officer at Palo Alto Networks, with the startup’s research team also joining the company.

Founded in 2025 by former OpenAI Group PBC researchers Patil, Rhythm Garg and Linden Li, Applied Compute offers a platform for post-training and inference of open-weight models. The platform is designed to make it easier for companies to train, serve and improve their own models, with an autoscaling inference engine and a model router called Shift-1 also included.

Applied Compute raised $350 million at a $3.25 billion valuation earlier this month in a round led by Elad Gil, with Nvidia Corp., Shopify Ventures and HubSpot Ventures also participating. At the time, the company said that its annualized recurring revenue had topped $100 million. Notable customers include Microsoft Corp. and DoorDash Inc. Coming into its acquisition, Applied Compute had raised more than $500 million in total.

Palo Alto Networks is pitching the deal as a bet that enterprise AI will be hybrid, as a large share of companies run open-weight models tuned to their own data. According to the company, no large provider is currently delivering those models securely. With Applied Compute, enterprises will be able to train and deploy open-weight models with security built in, either in the public cloud or in sovereign environments.

“Every enterprise is going to run a mix of open and closed models, and every one of those models needs to be secured,” Chairman and Chief Executive Nikesh Arora said. “Applied Compute has built the best post-training and inference platform we’ve seen, and its research team is exceptional.”

The deal came on the same day Palo Alto Networks reported what Arora called a record first quarter. For the quarter that ended on Oct. 31, the company reported adjusted earnings per share of $1.05, up from 93 cents in the same quarter of the previous year, on revenue of $3.44 billion, up 39% year-over-year. Analysts were expecting earnings of 97 cents per share on revenue of $3.31 billion.

Next-generation security annual recurring revenue grew 66% from a year earlier, to $9.73 billion. Remaining performance obligations rose 35%, to $21 billion. Helped by a tax benefit of $457 million, net income on an unadjusted basis came to $482 million, up from $334 million in the same quarter a year ago.

Chief Financial Officer Dipak Golechha said that given the current threat environment and the demand ahead, the company is “materially raising” its full-year guidance.

For its fiscal second quarter, Palo Alto Networks expects adjusted earnings per share of $1.01 to $1.03 on revenue of $3.53 billion to $3.55 billion. Full-year revenue is now forecast at $14.52 billion to $14.57 billion, up from the $14.10 billion to $14.20 billion guided in September, with adjusted earnings of $4.26 to $4.30 per share ahead of the $4.18 expected by analysts polled by FactSet.

Photo: Palo Alto Networks

来源

SiliconANGLE