AI Infrastructure & Investment - Hyperscalers are focusing on building comprehensive infrastructure including land and power, not just data [1] - Meta is projected to have $100 billion in CapEx for 2026, indicating significant investment in infrastructure [2] - Partnerships are becoming crucial for hyperscalers to manage financing, land acquisition, and power requirements, as their core competency lies in application development and algorithms [2][3] AI Compute & Usage - Google's token processing for generative AI has increased 100x over the past year, reaching 1 quadrillion (1,000 trillion) tokens per month [6] - Generative AI is rapidly growing, with monthly active users approaching 1 billion [6] - The surge in AI capabilities is leading to increased "vibe coding" [5] - Compute infrastructure demand is driven by the need for GPUs (e.g., Nvidia) and specialized AI accelerators (e.g., Google's Zella, Meta's AI LAN) [7] AI Development & Efficiency - AI is saving developers significant time in coding, as demonstrated by Jeopardy Five [9] - The scale of compute is magnifying, with some usage potentially being wasteful if not contributing to productive or enterprise use cases [8]

Technology & Innovation - Nvidia made compute affordable [1] - stwo is making proofs affordable [1]

Financial Data and Key Metrics Changes - Q2 revenue was $148.7 million, representing a 12% year-over-year growth, an improvement from 8% in Q1 [15][30] - Gross margin improved to 59%, a 170 basis point gain quarter-over-quarter [16][32] - Operating loss was $4.6 million, better than the guidance midpoint of $6 million [17][33] - Cash flow from operations was $26 million, or 17% of revenue, indicating strong cash management [18][35] - Free cash flow for Q2 was $10.9 million, a significant improvement from negative $18.5 million in Q2 2024 [36] Business Line Data and Key Metrics Changes - Security revenue reached a record high of $29.3 million, accounting for 20% of total revenue, with a 15% year-over-year growth [16][30] - Network services revenue was $114.9 million, growing 10% year-over-year [29] - Other products revenue grew 60% year-over-year, primarily driven by compute products [30] Market Data and Key Metrics Changes - Total customer count increased to 3,097, with enterprise customer count at 622, reflecting a 27 increase from the previous quarter [19] - LTM Net Revenue Retention (NRR) improved to 104%, up from 100% in Q1 [19][31] - Revenue from customers outside the top 10 grew 17% year-over-year, outpacing overall revenue growth [20][30] Company Strategy and Development Direction - The company aims to accelerate growth and drive towards profitability, focusing on customer needs and product responsiveness [9][10] - Three pillars of expansion include targeting performance-focused customers, cross-selling within the existing customer base, and geographic expansion [20][22] - A new leadership position was created to drive opportunities in the APJ region, indicating a focus on international revenue growth [23] Management's Comments on Operating Environment and Future Outlook - Management expressed optimism about the pricing environment, citing increased discipline in negotiations and a stable macro environment [48][50] - The company anticipates continued double-digit growth rates year-over-year for Q3 revenue [18] - Guidance for 2025 revenue was raised to a range of $594 million to $602 million, reflecting a 10% annual growth at the midpoint [41] Other Important Information - The company announced leadership changes, with a new CFO, Richard Wong, joining the team [11][12] - The company is focused on improving customer acquisition and reducing onboarding friction, with a 50% increase in packages sold year-over-year [24] Q&A Session Summary Question: Is the management change a new chapter for Fastly? - Management acknowledged the change as a new chapter and emphasized the opportunity to build on established momentum and accelerate results [44][46] Question: What is the outlook for the pricing environment? - Management noted an improving pricing environment due to increased discipline in negotiations and stability in the macro environment [47][50] Question: Can you provide insights on the cross-sell initiative? - Management highlighted strong customer interest in new products and the impact of the platform strategy on cross-selling opportunities [53][55] Question: What is driving the volatility in security revenue growth? - Management explained that volatility was due to historical trends and customer concentration, but stability is expected moving forward [78][81] Question: Can you elaborate on competitive displacements in DDoS? - Management indicated that new DDoS and bot mitigation products have led to consolidations from third-party vendors, enhancing customer experience [82]

Market Trend & Problem Statement - AI 将与未来的一切融合，对 GPU 和数据中心的需求正在爆炸式增长 [4] - 到 2030 年，需要比现在快四倍的速度建造多四倍的数据中心 [5] - 仅在美国，到 2030 年数据中心供应缺口将超过 15 吉瓦 [8] - 企业和公司 GPU 的空闲时间占 80% [9] - 构建数据中心面临挑战，例如成本高昂（第一个星际之门数据中心耗资超过 10 亿美元），连接电网速度慢（等待时间长达 7 年才能将 100 兆瓦的设施连接到北弗吉尼亚州的电网） [6][7] - GPU 和数据中心消耗了美国总用电量的 4%，并且环境可持续性较差，导致大量的二氧化碳排放 [8] Proposed Solution & Hyperbolic's Approach - 行业需要构建一个 GPU 市场或聚合层，以聚合不同的数据中心和 GPU 提供商，从而解决 GPU 用户的问题 [10] - Hyperbolic 正在构建一个名为 HyperDOS（Hyperbolic Distributed Operating System）的全球编排层，它类似于 Kubernetes 软件，允许任何集群在安装软件后成为网络中的一个集群 [11] - 用户可以通过多种方式租用 GPU，例如现货实例、按需、长期预留或托管模型 [11] - Hyperbolic 的 GPU 市场 H100 的 GPU 成本为每小时 0.99 美元，而 Google 的按需 GPU 成本为 11 美元 [13] - 通过统一的分销渠道，可以大幅降低价格 [13][14] - Hyperbolic 正在构建一个统一的平台，初创公司或公司不再需要审查不同的数据中心，只需选择评级高或价格最优的数据中心即可，还将对 GPU 的性能进行基准测试 [16] Benefits & Cost Savings - 通过 GPU 市场，可以节省 50% 到 75% 的成本 [13] - 通过 Hyperbolic，可以将成本从 4380 万美元降低到 690 万美元，节省 6 倍 [19] - 通过增加计算量，可以提高模型的质量，在相同的预算下，生产力可以提高 6 倍 [20] - 通过将闲置的 GPU 出售给其他人，可以帮助其他人获得更便宜的 GPU [20] Future Vision - GPU 市场将发展成为不同 AI 工作负载的一体化平台，包括 AI 推理（在线和离线）和训练作业 [21] - 行业应该更好地重用和回收那些闲置的计算资源，而不是仅仅关注构建数据中心，因为这会消耗大量能源和占用大量土地 [21]

$6 billion in a deal to equip a new data center in the state. Joining us now in a first on CNBC interview is core CEO Michael Intrator in Pittsburgh, along with our own Brian Sullivan. Brian.>> Yeah, John, thank you very much. I'm going to cross over into your world a little bit and talk about technology and data centers. And we are joined by Mike and trader.He is the CEO of core. We've Mike, thank you very much for joining us. An event with President Trump just wrapped downstairs.It was you and a bunch of ...

AI Infrastructure and Compute Demand - OpenAI recognizes the shift from compute for model training to compute for model usage, highlighting the need for significant infrastructure investment [1][2] - The company is experiencing high demand for its AI services, constrained by available compute resources, and could utilize twice its current capacity to improve product offerings [10] - OpenAI estimates needing $500 billion in compute capacity over the next few years to support growth projections, but is unsure if it could profitably deploy $1 trillion [11] - The company acknowledges a GPU shortage impacting its ability to handle demand surges and launch new features [13][16][17][18] - More compute directly translates to the ability to provide more AI capabilities for various applications [19] Strategic Partnerships and Investments - OpenAI is collaborating with Microsoft for compute resources, but believes the scale of its needs exceeds what any single company can provide [3][6] - The company has partnered with SoftBank for financial support and Oracle for technical expertise in the Stargate project [6] - OpenAI has raised tens of billions of dollars in funding and expresses confidence in its financial sustainability and profitability [12] Technological Advancements and Competition - OpenAI emphasizes the importance of both the infrastructure layer and the top-of-stack (AI models) as its competitive edge [20] - The company acknowledges the existence of other good AI models but highlights that ChatGPT has significantly more users than any other AI service [20] - OpenAI is focused on optimizing all aspects of its AI infrastructure, including chips, energy sources, and algorithms [33] Future Vision and Societal Impact - OpenAI's grand vision is to provide people with tools to enhance their capabilities and creativity [21][22] - The company is particularly excited about the potential of AI for scientific discovery, anticipating significant progress in 2026 [22][23] - OpenAI acknowledges the potential for AI to displace jobs but believes it will also create new opportunities, albeit at a rapid pace [25][26][27][28] - The company aims to make decisions that benefit humanity as a whole, particularly with the advent of humanoid robots [39]

Competitive Advantage - Data is the primary pillar for a durable competitive advantage among model providers [1] - Algorithms are IP that the industry will eventually learn [2] - Compute can be acquired by spending more money [2] - Data is an area where a long-term sustainable competitive advantage can be produced [2]