📡 Physix Frontier · Alpha News Source Draft

August 25, 2026 · Tuesday

Coverage Window: Global 24 hours (Up to US Eastern Close 8/24 + Asia-Pacific Session 8/24)

---

🧠 I. Large AI Models

DeepSeek Releases Native Multimodal Vision Model — V4 Roadmap Fills Key Gap in Image-Text Understanding

According to TMTPost, DeepSeek launched a Vision model supporting native multimodality, part of the subsequent V4 version capabilities. Founder Liang Wenfeng previously stated in investor exchanges that multimodality is a component rather than the main line for intelligence ceilings, but subsequent V4 versions would natively support it. Now this judgment has materialized. Native multimodality means image and video understanding capabilities enter the model base directly, no longer adapted via external plugins, allowing C-end applications and API product matrices to call them synchronously. For the open-source ecosystem, DeepSeek's multimodal weights have long been one of the community's most important bases. This update will drive a round of updates in downstream applications and fine-tuning ecosystems. Domestic large model competition thus extends from text to multimodality, heating up the battle for C-end assistant entry points. Multimodal inference consumes more compute, making marginal changes in inference-side hardware demand worth tracking. For the industry chain, multimodal weights require significantly higher VRAM and inference throughput than pure text, further increasing the unit value of compute demand. Adaptation progress of domestic inference chips and accelerator cards will also receive more attention. From a product rhythm perspective, the release of the Vision model means DeepSeek's support for C-end multimodal applications is ready. Image understanding and visual Q&A products can directly call native capabilities without stitching external vision modules. A batch of lightweight multimodal applications is expected to land quickly.

Research Institute Claims China-Linked Hackers Using DeepSeek to Enhance Attack Capabilities — AI Security Agenda Extends from Content Side to Offense-Defense Confrontation

Bloomberg cited researchers' reports stating that hacker groups with Chinese backgrounds have begun using open-source large models like DeepSeek to improve the efficiency and automation levels of cyberattacks. Signs of model assistance appeared in attack writing, obfuscation, and targeting stages. The spillover of open-source model capabilities into attack scenarios expands AI security discussions from content safety to offense-defense confrontation. National regulations on acquiring and using open-source weight models may tighten. For the security industry, demand for red team detection, threat intelligence, and endpoint protection rises accordingly. For the open-source ecosystem, controversies over model capability misuse may affect DeepSeek's overseas ecosystem expansion pace. This will be a long-term balance issue between open source and security. From an investment perspective, AI security shifts from compliance cost to real demand. Order visibility in sub-sectors like threat intelligence, model red teaming, and identity security is improving. Meanwhile, discussions around export and hosting policies for open-source models will increase. Classification management of model weights by countries may become the next regulatory agenda item, raising compliance costs for the open-source community.

• Mistral Partners with Saudi Arabia's Humain — Both sides join hands to advance sovereign AI infrastructure construction in Saudi Arabia and the region. European head large model vendors accelerate Middle East layout. Sovereign AI moves from slogan to orders, adding another piece to the AI going-global competitive map.
• Microsoft AI Business Highly Concentrated on Three Major Clients — OpenAI, TikTok, and Meta prop up Microsoft's AI revenue pie. Client concentration risk attracts market attention. Once any one adjusts procurement strategies or substitutes with self-research, the impact on Azure AI revenue will be direct. Investors begin treating concentration as a pricing variable for Microsoft's AI narrative.

---

💻 II. AI Software

• Porsche Signs $1.5 Billion AI Mega-Deal — Multi-year cooperation with India's Tata Consultancy Services, outsourcing intelligent and digital capability building to service providers on a large scale. Traditional automakers' AI transformation doesn't pursue building full capabilities in-house but packages compute, models, and software development. The scale and rhythm of such mega-deals are windows to observe corporate AI spending willingness. For TCS, this is an important anchor order for globalizing its AI business, providing new growth footnotes for the Indian software service industry's AI transformation. Deep binding between automakers and service providers is becoming a new landscape for European automotive digitization, changing the cost structure of software-defined cars.
• UK Digital Investment Surges to Dot-Com Bubble Era Levels — Driven by AI deployment, UK enterprise IT spending and digital infrastructure investment scale hit highs since the turn of the century. Europe's AI capex cycle remains in an expansion channel, resonating with the US. Data centers, cloud services, and enterprise software all benefit, while also raising market expectations for investment return realization.

---

🤖 III. Humanoid Robots

• Leapmotor Officially Announces Embodied Intelligence Robot Layout — The company believes NEV makers with full-domain self-research capabilities are fully qualified to do embodied intelligence. Plans exist, and specific information will be announced soon. Car manufacturing and robots share chassis, motors, and supply chains. The logic of automakers crossing over to embodied intelligence is being verified by more manufacturers. Leapmotor is another mainstream new force explicitly stating this after Xpeng. Whether robot business can move from planning to delivery depends on vehicle cash flow supporting long-cycle R&D. This is a common constraint for automakers doing embodied intelligence. For component suppliers, automaker entry means the robot industry chain gains another customer with mass production discipline. Supply chain scaling processes are expected to accelerate.
• Inovance Technology H1 Net Profit Attributable to Parent 2.81 Billion Yuan — Revenue 24.675 billion yuan, YoY growth 20.31%. Net profit declined 5.35% YoY. Industrial control main business remains stable, but margins are under pressure. As a core robot component supplier, Inovance's order and capacity trends are viewed as reference indicators for industry prosperity. Its H1 performance confirms automation demand is still gently recovering. Structurally, motor, electronic control, and power supply demands brought by NEVs and AI servers are providing second growth curves for traditional industrial control leaders. The extent of gross margin recovery in H2 is worth tracking.
• A Dexterous Hand Polished for 14 Months Before Launch — Lingqiao Intelligent, a company affiliated with Shanghai Jiao Tong University, states that the bottleneck of embodied intelligence is shifting from understanding to execution. Dexterous manipulation is the next watershed. Robots can see and hear but fail to twist bottle caps well. Fine manipulation capability at the execution level becomes the industry's next competitive focus. As the most direct human-machine interaction terminal, the commercialization rhythm of dexterous hands begins to be re-examined by outsiders.

---

🚗 IV. Autonomous Driving

• Qianli Technology H1 Revenue 5.12 Billion Yuan — Smart driving business covers 16 vehicle models. Net profit attributable to parent surged 151% YoY. The trend of smart driving solution commercialization volume ramping up continues. Amidst price wars and automaker self-research coexisting, third-party smart driving suppliers achieving both revenue growth and profit improvement indicates high-level smart driving penetration is spreading from head automakers outward. Mass production coverage of 16 models means marginal costs of smart driving solutions are declining. The industry shifts from selling solutions to selling scaled services, improving profit models. Qianli Technology's interim report also shows that after smart driving business moves from designation to mass production delivery, revenue recognition rhythm stabilizes, providing references for similar smart driving suppliers.
• Xpeng Group H1 Net Loss 3.121 Billion Yuan — Loss expanded 173.35% YoY. Revenue 32.777 billion yuan, down 3.8% YoY. Investments in new model launches and overseas expansion continue to suppress the income statement. However, management simultaneously advances robot business financing and smart driving tech releases. The market will watch H2 delivery rhythm's support for loss reduction and whether new businesses can open second growth curves.

---

🌐 V. Physical AI & Compute Infrastructure

Optical Module Leader Innolight Sees Inventory Surge — Inventory Concerns Emerge Under AI Boom

Bloomberg reported that Eoptolink (Innolight), one of the biggest beneficiaries of AI optical module demand, stockpiled heavily during the boom cycle, causing inventory scale to climb sharply. The optical module segment has maintained high prosperity for multiple consecutive quarters. The market begins to reassess inventory digestion rhythm and price hike sustainability. Inventory surge has two interpretations: locking capacity for delivery preparation, or early signals of demand overdraft. Bulls and bears can find arguments in both. A-share optical module duo stocks adjust at highs. Sector sentiment shifts from unilateral optimism to divergence. If subsequent inventory destocking falls short of expectations, valuation center has room for downward revision. If demand continues to exceed expectations, stockpiling instead constitutes delivery advantage. Divergence itself shows compute hardware trading has entered the verification period of prosperity realization. On the capital side, sector valuations are at highs after continuous rises. Any inventory and price signals will be amplified. Holding volatility is unlikely to drop in the short term. From position structure, optical modules are one of the heaviest allocations for institutions in AI hardware. Reflexivity of crowded trading means stock price sensitivity to negative information rises significantly. Marginal changes in single-product inventory data could trigger rebalancing.

• Meta Releases MetaRoCE — New RDMA transmission scheme built for AI-scale Ethernet, reducing GPU cluster data transfer losses. The larger the cluster scale, the more the network layer becomes a training efficiency bottleneck. Technical competition in AI infrastructure extends from chips to interconnects and network protocols.
• TerraPower CEO Says AI and Hyperscale Data Centers Are Driving Nuclear Power Demand — Advanced nuclear energy company invested by Bill Gates judges that AI power gaps will accelerate nuclear commercialization. Power supply is becoming a hard constraint on compute expansion. Order cycles for nuclear and gas power plants are being repriced by the market.
• Australia's Data Center Electricity Consumption Expected to Increase ~7x by 2036 — AI drives power infrastructure investment. Countries' planning for data center power supply enters substantive stage. Triple constraints of compute, power, and land begin to manifest. Capital expenditure cycle for power infrastructure is expected to lengthen in sync with AI capex cycle.

---

📈 VI. Macro & Market Data

Nvidia Senior Manager Allegedly Involved in Supermicro Smuggling Case — AI Server Export Compliance Risks Heat Up

Ars Technica reported that a senior Nvidia manager is allegedly linked to Supermicro's suspected smuggling plan to funnel AI servers to China, bypassing US export controls. US export controls on AI chips to China enter deep implementation waters. Executive involvement reinforces compliance risk narratives and indirectly confirms the huge demand gap for advanced compute in the Chinese market. For Nvidia, regulatory and reputation uncertainty rises, putting short-term stock pressure. For domestic compute industry chain, substitution logic gains new support. Verification and procurement rhythm of domestic GPUs may accelerate due to this. For Supermicro, the event adds to previous governance issues. Repairing compliance costs and customer trust for its server business still takes time. For other suppliers in the export control chain, this case is also a warning. Manufacturers with unsound compliance systems face risks of exclusion from mainstream client lists.

Pinduoduo Q2 Adjusted Profit Beats Expectations — Another Highlight in Chinese Concept Stock Earnings Season

Pinduoduo released its Q2 2026 earnings. Adjusted profit exceeded market expectations. Quarterly revenue 112.4 billion yuan, profit 27.2 billion yuan, benefiting from the release of effects from merchant support and platform governance upgrades. Under dual pressure of Temu overseas expansion and domestic e-commerce competition, profitability resilience