Flybridge AI Index: October 2024 Update
We thank Aiera, an AI-powered equity research platform, for accelerating our data and insight extraction process.
Summary:
In October, the Flybridge AI Index returned 2%. For the prior 12 months, the Index has returned 43% to its current level of 112% since January of 2023.
Of the 32 active companies in the Index, 17 gained in October and 15 declined in the month. Significant gainers include Astera Labs (+38%), and Palantir (+14%). Significant decliners include Super Micro (-28%), and Simulation Plus (-11%).
The Q3 earnings season has kicked off, with 11 companies reporting so far. Early results indicate initial value generation from generative AI at the application level.
Early signs of true value creation!
As the IndexAI shows, the early winners of this GenAI wave have mainly been in the compute layer. However, the critical $600 billion question remains: will all the investment in AI infrastructure, especially GPU spending, translate into sustained revenue and real value for companies and end users? At the start of this Q3 earnings season, we're beginning to see signs that could help answer this question.
ServiceNow, for whom we did a deep dive last month, has reached over $70 million in ACVs from NowAssist within just 16 months of launch. Meta, leveraging AI to enhance content recommendations, has seen an 8% increase in time spent on Facebook and 6% on Instagram, with advertisers reporting a 7% boost in conversion rates through Meta's AI tools. In a major announcement, Sundar Pichai shared in Google’s earnings call that over 25% of the company's code is now generated via AI.
Although we are still early in the trend and cannot yet answer the $600 billion question, these initial signs spark enthusiasm about the massive value generative AI is poised to create.
AI Insights and News:
Tech Companies Q2 2024 Earnings Highlights:
TSMC reported their Q3-24 earnings. Some interesting insights from the month included
Management noted experiencing exceptionally high demand for AI-related processing power and expressed confidence in sustained growth.
They highlighted the use of AI and ML in their fabrication and R&D operations, noting that even a 1% productivity gain translates to $30 million, showcasing a clear ROI benefit of AI.
Despite initial challenges, TSMC reported a 4% higher production yield at its Arizona site compared to similar sites in Taiwan—a promising development for the U.S. in its efforts to reduce reliance on Taiwan. (Bloomberg)
OpenAI is developing its own semiconductor and plans to partner with TSMC for manufacturing. (Reuters)
IBM reported their Q3-24 earnings. Some interesting insights include
IBM’s generative AI business has surpassed $3 billion to date, with over $1 billion added in the recent quarter. About 80% of this comes from consulting and 20% from software signings.
The Granite 3.0 family of open-source AI models has launched with 1B to 8B parameters, trained on over 12 trillion tokens. It includes three variants: Granite Dense (enterprise-focused), Granite MoE (CPU/device efficient), and Granite Accelerator (optimized for fast inference). These models improve efficiency by about 90% compared to larger models and support applications like RAG, summarization, entity extraction, classification, code generation/translation, and tool integration. (IBM)
ServiceNow reported their Q3-24 earnings. Some interesting insights include
Last month, we took an in-depth look at ServiceNow. Despite potential disruption from AI-native competitors, ServiceNow’s early results are impressive. In last week’s earnings, they reported that NowAssist already has 36 customers with $1M ACV, 6 with $5M ACV, and 2 with $10M+ ACV, totaling over $70M+ in annual revenue—a remarkable achievement just 16 months after launch.
They estimate that AI-driven opportunities will expand their addressable market by an additional $500 billion, effectively doubling it.
One customer achieved a 75% improvement in case deflection rates with AI-driven ITSM and HRSD tools, while another saw a 30% increase in self-service adoption.
Adobe
Firefly Services, a platform for automated content creation, has seen a 3x increase in API calls quarter-over-quarter, highlighting strong enterprise demand.
At Adobe Max 2024, Adobe unveiled a new Firefly video model for generating videos from text and image prompts, along with the Firefly Image 3 Model, which generates images four times faster. As before, these models are trained exclusively on licensed content. A key theme was that these generative AI tools aim to enhance—not replace—creators and designers, with a strong focus on video editing. (Adobe)
Intel reported their Q3-24 earnings. Some interesting insights include
IBM is set to deploy Gaudi 3 as a service on IBM Cloud.
Transition from Gaudi 2 to Gaudi 3, alongside software ease-of-use challenges, led to slower adoption than anticipated, which ultimately impacted the revenue target. As a result, Intel adjusted its initial 2024 revenue target of $500 million for Gaudi due to these adoption delays.
InflectionAI new Inflection 3.0 model family is powered by Intel's Gaudi 3 AI accelerators. (Inflection)
AMD reported their Q3-24 earnings. Some interesting insights include
The company raised its 2024 data center GPU revenue projection to over $5 billion, up from the $4.5 billion forecasted in July, reflecting significant growth driven by strong demand and key customer milestones.
Anticipates the AI accelerator TAM in data centers to grow over 60% annually, reaching $500 billion by 2028.
Microsoft and Meta have expanded their use of MI300X GPUs for inferencing and training, with Meta exclusively using MI300X for the demanding inferencing tasks of their Llama 4 model.
AMD launched the MI325X this quarter, delivering 20% higher inferencing performance than its H200 competitor. The MI355 is scheduled for release in the second half of 2025, with the MI400 following in 2026.
OpenAI plans to incorporate AMD's new MI300X processors into its training infrastructure, reducing its reliance on Nvidia GPUs.
Meta reported their Q3-24 earnings. Some interesting insights include
Meta AI has quickly hit a major milestone, reaching 500 million users since its launch last fall
Enhancements to AI-driven content recommendations contributed to an 8% increase in time spent on Facebook and a 6% increase on Instagram in 2024.
More than 1 million advertisers used Meta’s generative AI tools to create over 15 million ads in the last month. Advertisers who leveraged AI-driven image generation tools saw a 7% increase in conversion rates.
Meta sees potential for monetizing Meta AI through commercial queries in the future, though the current focus remains on optimizing user experience.
Finalizing a new AI supercomputing cluster with over 100,000 Nvidia H100 chips to train its upcoming Llama 4 language model. (The Information)
Google reported their Q3-24 earnings. Some interesting insights from the month included
Google now uses AI to write over 25% of its code, with human engineers reviewing and approving it.
Google Cloud AI revenue grew 35% YoY, driven by strong demand for generative AI (Gen AI) solutions.
Gemini API calls have grown 14x over six months
Reported a 90% cost reduction for AI search queries over 18 months due to model, hardware, and engineering enhancements.
Microsoft reported their Q3-24 earnings. Some interesting insights from the month included
Microsoft's AI business is projected to surpass a $10 billion annual revenue run rate by Q2 FY25, marking it as the fastest-growing business in the company's history.
Nearly 70% of Fortune 500 companies now use Microsoft 365 Copilot, which Microsoft described as its most rapidly adopted new suite.
GitHub Copilot's enterprise customers grew by 55% quarter-over-quarter. Companies like AMD and Flutter Entertainment are customizing Copilot for their codebases
Note: These insights are focused on AI developments and priorities as discussed in management publications, earnings calls, and other company announcements.
Performance Overview:
Inception (January 2023) to date returns for the Flybridge AI Index are 112%. In comparison, over the same time period the Cloud Bessemer Index returned 35%, the S&P 500 returned 51%, the Nasdaq returned 78%, and the F-Prime Fintech Index returned 119%.
The median NTM revenue multiple was 8.3x.
The median quarterly YoY revenue growth rate was 16%.
The median LTM net income margin was 13%.
Additions/Deletions:
No new additions were made to the AI Index in October, but we added Datadog to the list of potential future entries. Although Datadog initially took a conservative approach with GenAI-related products, they recently launched a monitoring platform for tracking OpenAI usage within Datadog to analyze usage patterns and optimize performance. Management has indicated plans to be more aggressive in developing new products in this area, so we’ll be monitoring and likely adding them in the coming months.
The information in the Index is for informational purposes only. Flybridge is not offering investment advisory services, nor is it offering to sell securities.
