Smartotics Investment Daily - 2026-07-01

📈 Market Overview

Technology markets open H2 2026 with cautious optimism. The NASDAQ Composite enters July up 18% year-to-date, driven primarily by AI infrastructure and semiconductor names, though the gains have concentrated in fewer stocks as the year progresses. NVIDIA remains the bellwether, up 42% YTD, but its forward P/E of 32x has compressed from 45x in January as growth rates normalize from triple-digit to high-double-digit percentages. The semiconductor equipment sector faces crosscurrents: strong demand from AI/hyperscaler capex offsets weaker consumer electronics and automotive chip demand, creating a bifurcated market where AI-exposed equipment makers trade at premium valuations while diversified players languish. In venture capital, Q2 2026 robotics and AI funding totaled $18.7 billion globally, down 8% from Q1 but up 34% YoY, suggesting a normalization toward sustainable levels after the 2024-2025 frenzy. Chinese AI infrastructure startups continue to attract significant capital despite geopolitical headwinds, with storage, networking, and specialized AI chip companies collectively raising over $2 billion in Q2. The IPO pipeline for H2 looks robust, with at least four AI/robotics companies filing confidential S-1s in June, though recent new listings have traded mixed, suggesting investors are becoming more discerning about AI revenue quality versus AI marketing narratives.

💰 Funding Radar

1. Physical Intelligence (π) — $400M Series C at $2.4B Valuation

Source: Bloomberg, TechCrunch (May 2026)

Deal Details:

Why It Matters: This investment represents a bet on the “software eats robotics” thesis — that the value in robotics will shift from hardware to AI software, and that a general-purpose model can capture disproportionate value across the ecosystem. At $2.4 billion, Physical Intelligence is priced for a winner-take-most outcome in a market that doesn’t yet exist at scale. The strategic participation of NVIDIA (hardware) and robot manufacturers (deployment partners) suggests an ecosystem-building approach: embed π’s model as the default brain for multiple robot platforms, creating a de facto standard.

The $400M raise also signals that investors believe the embodied AI market will be large enough to support venture-scale returns. If the global robotics market reaches $300 billion by 2030 (current IFR projections), and software captures 15-20% of value, the addressable market for robot AI software would be $45-60 billion — sufficient to justify a $2.4B valuation if π captures significant share.

My Take: Investment Thesis: Physical Intelligence has the right team (top DeepMind robotics researchers), the right thesis (general-purpose over platform-specific), and the right timing (VLA models are approaching production readiness). However, the valuation implies execution risk that shouldn’t be ignored. Robot manufacturers have historically preferred building proprietary software stacks — convincing them to adopt a third-party model requires not just technical superiority but a business model that doesn’t threaten their margins.

Risk Factors:

Growth Potential: If π achieves 90%+ reliability by 2028, it could become the “Android of robotics” — powering hundreds of thousands of robots across dozens of platforms. At $100/robot/month, 500,000 deployed robots would generate $600M ARR, justifying a significantly higher valuation.

Recommendation: WATCH — The thesis is compelling but the valuation prices in execution that hasn’t happened yet. Wait for evidence of manufacturer adoption traction (3+ signed license deals with major robot OEMs) before investing at these levels.

2. Tongyou Technology (同有科技) — ¥1 Billion ($138M) for AI-Specific Storage

Source: 36Kr — “同有科技完成10亿元融资，聚焦AI存储赛道”

Deal Details:

Why It Matters: This fundraise highlights the AI infrastructure bottleneck shifting from compute to data. While GPU supply constraints have eased in 2026 (NVIDIA H200 and B100 widely available), AI training clusters are increasingly I/O-bound rather than compute-bound. Tongyou’s AI storage solutions address this by providing high-throughput, low-latency data pipelines that keep expensive GPUs fed with data — a critical economic consideration when GPU clusters cost $50,000+ per hour to operate. The ¥1 billion raise positions Tongyou to compete with Western storage vendors (Pure Storage, NetApp, Dell) in the Chinese AI infrastructure market, which is projected to reach $12 billion by 2028.

My Take: Investment Thesis: Tongyou occupies a strategically important niche in China’s AI infrastructure stack. As the US restricts advanced GPU exports to China, Chinese organizations are maximizing utilization of available GPUs — making I/O optimization a high-ROI investment. Tongyou’s AI storage revenue growing at 80% YoY suggests strong product-market fit.

Risk Factors: The AI storage market is attracting intense competition from both specialized startups and established vendors. Tongyou’s public listing provides capital access but also subjects it to quarterly earnings pressure that may conflict with the long investment cycles typical of enterprise infrastructure sales.

Recommendation: BUY — At approximately 3x revenue (based on estimated $1.2B valuation and $441M revenue), Tongyou trades at a discount to Western storage peers (4-6x revenue) despite faster growth, reflecting the China discount and smaller scale. The AI storage growth trajectory justifies a modest position.

Based on real news from 36Kr, WallStreetCN, and Hacker News.

Sources Referenced:

Disclaimer: This content is for informational purposes only and does not constitute investment advice.