Street encouraged by Google’s new AI offerings, BofA not worried about Nvidia

During the first day of its Cloud Next conference in Las Vegas, Google (GOOGL) introduced Axion, the company’s first Arm (ARM)-based, home grown CPU, designed for the data center. The unveiling was received well by Wall Street, with Citi noting Alphabet’s AI product cycle is accelerating and Wedbush saying it is “encouraged” by Google’s progress. With that said, BofA told investors to “not sweat” the competitive noise and that it sees Nvidia’s (NVDA) recent stock underperformance as a pause that can be “refreshing.”

AXION PROCESSORS: Google announced Google Axion Processors, the company’s first Arm-based CPUs designed for the data center. The company says that “Axion processors combine Google’s silicon expertise with Arm’s highest performing CPU cores to deliver instances with up to 30% better performance than the fastest general-purpose Arm-based instances available in the cloud today, up to 50% better performance and up to 60% better energy-efficiency than comparable current-generation x86-based instances.”

The company also said that “Gemini for Google Cloud is here.” Gemini for Google Cloud is a new generation of AI assistants for developers, Google Cloud services, and applications. These assist users in working and coding more effectively, gaining deeper data insights, navigating security challenges, and more. Google also announced its Gemini Code Assist. “It offers AI-powered assistance to help developers build applications with higher velocity and quality in popular code editors like VS Code and JetBrains. Gemini Code Assist is an enterprise-grade coding assistance solution that supports your private codebase wherever it lives.”

ACCELERATING AI PRODUCT CYCLE: Citi believes the Google Cloud Next event highlights Alphabet’s transition from experimental generative artificial intelligence features to fully established Gemini-enabled cloud offerings. The firm was impressed to hear that 60% of funded generative AI startups and nearly 90% of generative AI unicorns are Google Cloud customers and to date, 1M developers have built with Google’s generative AI tools. Citi believes Google’s AI product cycle is accelerating across both its core search and cloud businesses. It keeps a Buy rating on the shares with a $168 price target.

ENCOURAGING PROGRESS: Commenting on Google’s annual Cloud Next event, Wedbush says it comes away encouraged by the company’s progress, both in developing its AI capabilities and commercializing early enterprise solutions with new custom chips announced — Axion — and a long list of generative AI services and features within Google Cloud that should support growth over a multiyear period. Google’s most advanced model, Gemini 1.5 Pro offers significantly larger context windows versus leading competitors, and the firm is encouraged to see Gemini 1.5 Pro powering enterprise tools like Gemini Code Assist, as well as broader use cases across the Gemini family of models that should drive direct monetization. Wedbush has an Outperform rating on the shares with a price target of $175.

CLEARER LONG-TERM VISION: Google Cloud’s competitive vision and offering became clearer during Cloud Next, Morgan Stanley says in a research note following the first day of the event. Google’s deep technical integration across infrastructure, models, platform services, and applications represents a competitive advantage, it adds, noting that execution/go to market will be key to drive faster GCP growth.

Morgan Stanley believes Google demonstrated the benefits of its deep technical integration across each layer of the AI tech stack, which should help them better compete with AWS (AMZN) and Microsoft’s (MSFT) Azure going forward. While Alphabet will need to prove that this deep level of integration across the AI tech stack can translate into faster Google Cloud growth, the firm walks away more constructive on its competitive positioning. Morgan Stanley has an Overweight rating on Alphabet with a price target of $165 on the shares.

‘DON’T SWEAT’ COMPETITIVE NOISE: Google kicked-off its annual Cloud Next event with a keynote announcing new custom AI chips, platform-wide Gemini integration, strategic partnerships, and customer traction. The firm’s top takes are that Google’s hardware advances, Gemini progress, and AI driven app demos should help sentiment on AI capabilities, while ability to ground models with Google search/data was an appealing capability. Further, it thinks Cloud AI cycle could be underappreciated for the industry and a positive driver for Google stock given growing scale and potential contributor to higher overall revenue growth, increasing cloud margins aiding EPS, and growing stock valuation contribution on 2026 revenues.

In a separate note, BofA also pointed out that Nvidia shares were down about 11% on Tuesday from the all-time high of $950 hit on March 25 following the company’s GTC tradeshow, pointing out that the current selloff marks the ninth time Nvidia’s stock has declined about 10% or more since ChatGPT was launched in late 2022. While there are some market factors, the firm said that on a fundamental basis it has also heard of investor concern around rising competition, including Tuesday’s announcements from Google and Intel (INTC) and reducing lead-times that suggest decelerating demand. However, BofA advises investors “don’t sweat competitive and Asia supply chain noise” as it believes the fundamentals are solidly on track and periods of consolidation “tend to set the stock up for strong moves later.” BofA maintains a Buy rating and $1,100 price target on “top pick” Nvidia.

WHAT’S NOTABLE: At its Vision 2024 conference on Tuesday, Intel unveiled the Gaudi 3, an AI accelerator chip, saying it is both faster and more efficient than Nvidia H100 GPUs. The H100 is one of the company’s most popular chips. In comparison to Nvidia H100, the company said, Intel Gaudi 3 is projected to deliver 50% faster time-to-train on average across Llama models with 7B and 13B parameters, and GPT-3 175B parameter model. Additionally, Intel Gaudi 3 accelerator inference throughput is projected to outperform the H100 by 50% on average and 40% for inference power-efficiency averaged across Llama 7B and 70B parameters, and Falcon 180B parameter models.

Intel also outlined its strategy for open scalable AI systems, including hardware, software, frameworks and tools. Intel’s Gaudi 3 comes just a month after Nvidia debuted its next-generation AI platform called Blackwell.

PRICE ACTION: In morning trading, shares of Alphabet have dropped about 1% to $155.50, while Nvidia’s stock has gained almost 2% to $868.78.