Follow
Subscribe via Email!

Enter your email address to subscribe to this platform and receive notifications of new posts by email.

Palo Alto Networks to Acquire Applied Compute in $4.5B Deal

Palo Alto Networks agreed to buy AI startup Applied Compute in an all-stock deal valued at $4.5 billion, expanding into open-weight model training and inference while reporting record quarterly revenue.
People gather at a Palo Alto Networks exhibition booth during an industry trade show event.

Palo Alto Networks agreed to acquire artificial intelligence startup Applied Compute in an all-stock deal valued at $4.5 billion on November 18, 2026. The transaction targets open-weight model software for enterprises. The announcement arrived as Palo Alto Networks beat Wall Street quarterly forecasts with $3.44 billion in revenue and raised full-year guidance [1].

Terms of the Palo Alto Networks Acquisition

The all-stock transaction values Applied Compute at $4.5 billion and remains subject to regulatory clearance before closing in its third fiscal quarter, according to terms announced by both companies. Leadership changes take effect upon closing. Applied Compute co-founder and Chief Executive Yash Patil will join the executive group at Palo Alto Networks as co-chief AI officer. The startup’s research team will also move into the company to guide model development. Investors responded fast to the transaction, sending shares of Palo Alto Networks up nearly 10% during regular trading before the stock added more than 8% in late market sessions [1].

Patil set up Applied Compute in 2025 with fellow researchers Rhythm Garg and Linden Li after they left OpenAI Group PBC to build software for open-weight neural network systems. Their work caught early attention from large corporate clients, drawing Microsoft Corp. and DoorDash Inc. as customers. Annualized recurring revenue topped $100 million before the acquisition [1].

Venture backing totaled over $500 million. Applied Compute raised $350 million at a $3.25 billion valuation earlier in November in a funding round led by Elad Gil, with Nvidia Corp., Shopify Ventures, and HubSpot Ventures participating. That rapid valuation increase reflected strong commercial uptake, as the company had pushed annualized recurring revenue past $100 million with major enterprise software accounts such as Microsoft Corp. and DoorDash Inc. on its customer roster. The buyout delivers a rapid return to these early financial backers while shifting the startup’s research group into Palo Alto Networks as enterprise demand for open-weight model infrastructure accelerates across customer accounts [1].

Applied Compute Model Platform

The software platform simplifies model operations. Applied Compute provides an autoscaling inference engine designed to handle unpredictable request volumes together with a model routing system called Shift-1 that directs queries to optimal neural network backends. Engineering teams use these integrated components to train, serve, and improve their own models within a single unified environment [1].

While hardware developers like SiMa.ai develop physical AI chips for edge processing, Applied Compute focuses entirely on software workflows for post-training and running open models. Enterprise teams face steep technical barriers. Training open-weight models on proprietary corporate data demands complex coordination of memory buffers, parameter weights, and computing clusters, which often forces companies to build custom tooling or hire rare machine learning infrastructure specialists. The Applied Compute platform automates these background procedures so that software groups can train, tune, and evaluate their own models without having to assemble scattered open-source tools from scratch. Patil and his research colleagues designed the framework to make open models as straightforward to run as commercial closed-source APIs in production business environments [1].

Palo Alto Networks headquarters and cybersecurity platform operations
The Register reported on the agreement by Palo Alto Networks to acquire Applied Compute. (Credit: The Register)

Open weights offer organizations greater operational independence. By maintaining direct access to model weights, companies can inspect data handling steps and run workloads on private servers rather than routing sensitive information through external hosted endpoints. Applied Compute removes deployment hurdles by packaging open models in secure, automated serving infrastructure for production systems [1].

Hybrid Models and Enterprise Security

Hybrid AI architectures are gaining ground. Palo Alto Networks framed the acquisition as a long-term bet that businesses will operate a mixture of proprietary cloud systems and internal open-weight models tuned to their own private data. The company noted that no major cybersecurity provider currently delivers security across open-weight training and inference workflows [1].

Executive leaders argue that protecting model layers is essential as generative tools enter core production workflows. “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 [1]. “Applied Compute has built the best post-training and inference platform we’ve seen, and its research team is exceptional.” Arora has guided the company since June 2018. Before taking the Chairman and CEO role at Palo Alto Networks, Arora served as a senior executive at Google and SoftBank, directing international technology operations and financial expansion before turning to cybersecurity [3].

Computing trends point to open architectures. Industry reporting from The Register highlights growing enterprise interest in high-performance computing clusters and open models as organizations seek alternatives to commercial cloud dependencies [2]. Pairing Applied Compute with existing defenses allows Palo Alto Networks to provide verified guardrails for enterprise model deployments [1].

Palo Alto Networks event booth and company branding
SiliconANGLE covered the acquisition of Applied Compute and fiscal earnings. (Credit: SiliconANGLE)

Record First-Quarter Earnings

The purchase announcement arrived on the same day as fiscal first-quarter financial results for the period ended October 31, 2026, which came in well ahead of Wall Street estimates. Financial results showed broad revenue gains across enterprise security contracts. Palo Alto Networks posted revenue of $3.44 billion, marking a 39% increase from the prior-year period and beating the $3.31 billion consensus forecast compiled by FactSet analysts. Adjusted earnings per share climbed to $1.05. That adjusted profit figure improved from 93 cents per share reported in the same quarter last year, comfortably topping the 97 cents per share projected by market analysts [1].

Subscription metrics also moved sharply higher. Next-generation security annual recurring revenue grew 66% year-over-year to $9.73 billion, while remaining performance obligations rose 35% to reach $21 billion. Unadjusted net income reached $482 million, up from $334 million a year earlier, helped by an income tax benefit of $457 million during the quarter [1].

Full-year revenue expectations were raised upward. Chief Financial Officer Dipak Golechha said that given the current threat environment and strong market demand, the company is “materially raising” its financial guidance across coming quarters. For the fiscal second quarter, the company anticipates revenue between $3.53 billion and $3.55 billion with adjusted earnings of $1.01 to $1.03 per share. Annual revenue targets rose above previous projections. Full-year revenue is now forecast between $14.52 billion and $14.57 billion, up from the $14.10 billion to $14.20 billion range provided in September, with adjusted earnings expected between $4.26 and $4.30 per share [1].

Palo Alto Networks Background and Scale

Nir Zuk founded the company. A former engineer at Check Point and NetScreen Technologies, Zuk previously directed software development in Unit 8200, an intelligence branch of the Israel Defense Forces. Palo Alto Networks launched in 2005 in Santa Clara, California, raising $260 million during its 2012 New York Stock Exchange debut before shifting its listing to Nasdaq in October 2021 [3].

International market expansion continued through 2026. In February 2026, the company joined the Tel Aviv Stock Exchange as a dual-listed entity, and by September 2026 it became the first enterprise to hit the NIS 1 trillion threshold on that market. Global market capitalization crossed $200 billion for the first time in May 2026. Customer reach spans leading global enterprises across multiple industry sectors. Today Palo Alto Networks serves more than 70,000 client organizations across more than 150 countries, including 85 of the Fortune 100, while hosting the Ignite security conference and partnering with the World Economic Forum [3].

Targeted acquisitions built out the platform. The company bought Evident.io for $300 million in 2018 to launch Prisma Cloud, Demisto for $560 million in 2019, CyberArk for $25 billion in 2026, and Chronosphere for $3.35 billion in 2026. Meanwhile, its in-house research group Unit 42 tracks global malware threats and shares intelligence through the Cyber Threat Alliance founded in 2014 with Fortinet, McAfee, and Symantec to strengthen defenses worldwide [3].

Sources
  1. ONLINE NEWS Riley, D. (2026, November 18). Palo Alto Networks bets $4.5B on AI startup Applied Compute as earnings top forecasts. SiliconANGLE. [Article Link]
  2. ONLINE NEWS Claburn, T. (2026, November 19). Palo Alto Networks opens wallet for open-weight training biz. The Register. [Article Link]
  3. WEBSITE Wikipedia contributors. (2026). Palo Alto Networks. Wikipedia. [Article Link]
Leave a Comment

Related Posts
Total
0
Share