Recently, a U.S. official stated that Iran has launched missile and drone attacks against neighboring countries including Bahrain and Kuwait. Although there has been no significant impact on U.S. military facilities in the Middle East and no reports of U.S. personnel casualties, regional security premiums have significantly risen. At the same time, on June 27, 2026, Netanyahu publicly stated at a press conference that there is no space for a "two-state" solution between the Mediterranean and the Jordan River, and claimed that the Israeli public's perception of the "two-state solution" has changed, objectively further constraining the political space for Israeli-Palestinian peace negotiations. In the context of escalating geopolitical tensions, industry monitoring has detected that Google is deploying technology such as zero-copy MTP for the Pixel 9 and Pixel 10 series, resulting in over a 50% increase in the end-side Gemini Nano multi-token inference speed while simultaneously reducing memory usage. Combined with the trends revealed by OpenRouter data—indicating that the performance gap between open-source large models from China and the U.S. and closed-source cutting-edge models has narrowed to about 3 to 6 months with very low inference costs—this means that the end-side and open-source AI are simultaneously approaching closed-source cutting-edge models. The escalation of geopolitical conflicts and the setback of peace expectations typically raise global risk premiums, while accelerating AI technology and declining costs provide valuation support for technology and computing-related assets. These two clues will continue to constitute intertwined main variables for risk preference and technology asset pricing in the coming months.
Iran's Missile Attack Spread: Middle East Risks Rise Again
According to U.S. officials, Iran's recent missile and drone attacks on neighboring countries, including Bahrain and Kuwait, indicate that the scope of strikes has expanded from bilateral confrontations to a broader Gulf cooperation landscape, shifting geopolitical conflict from "localized friction" to a direct impact on regional security architecture. However, on a military level, no U.S. personnel casualties have been reported in the Middle East, and U.S. military facilities are described as "not significantly affected," which makes this operation appear more like a high-intensity signal release rather than an immediate trigger for large-scale confrontation.
For the market, the overlapping information of "missile attack spread" and "U.S. military not significantly affected" creates a contradictory risk perception: on one hand, the geopolitical security baseline is elevated, and investors need to reserve risk premiums for more frequent emergencies in asset allocation; on the other hand, the short-term lack of conditions for U.S. personnel casualties and critical facility damage reduces the probability of escalating into a regional war, causing risk assets to reflect shocks more through volatility amplification rather than one-sided trends. Historical experience shows that tensions in the Middle East often accompany fluctuations in oil prices and increasing demand for safe-haven assets; this cross-border strike could similarly push up the risk premium on energy-related assets, while amplifying the passive tracking of global risk assets to geopolitical "headlines," making such events a tail risk variable that needs to be continuously included in pricing frameworks.
Netanyahu Denies the Two-State Solution: Peace Prospects Stymied
On June 27, local time, Netanyahu publicly stated at a press conference that there is no space for a "two-state" solution between the Mediterranean and the Jordan River, emphasizing that public perception of the "two-state solution" among Israelis has changed. This statement effectively labels the "two-state solution" as "unrealistic" in terms of sovereignty and security narratives, whereas it has long been viewed as one of the mainstream international paths to alleviate the Israeli-Palestinian conflict. The current Israeli leadership publicly denying this path further narrows the already limited political and negotiation space, shifting the peace process from "suspension" to "reversal," and raising domestic political costs for any subsequent attempts to restart negotiations.
In this context, the probability of the Israeli-Palestinian conflict becoming prolonged and structurally stagnant is being re-evaluated by the market. The aforementioned cross-border strikes have pushed the risk of geopolitical conflict in the Middle East into investors' view, while Netanyahu's denial of the "two-state solution" signals that a "final resolution" to regional conflict is less likely to come onto the agenda in the foreseeable future, increasing the weight of conflict persistence. For global risk assessments, this change will be reflected in the embodiment of higher and more prolonged geopolitical risk premiums, passively increasing the sensitivity of energy, safe-haven assets, and assets highly correlated with regional risks in allocations. The escalation of geopolitical conflicts and the setback of peace prospects typically raises global risk premiums, while whether subsequent technology and computing-related assets can offset some of the risk elevation based on their fundamentals and valuation logic will become a key variable that needs continuous tracking in asset allocation.
Pixel 9 and 10 Speed Up 50% End-Side AI Acceleration
Against the backdrop of rising geopolitical risk premiums, a clear hedging signal monitored by industry observers is that Google is strongly betting on enhancing end-side AI efficiency in the upcoming Pixel 9 and Pixel 10 series: by deploying technologies such as zero-copy MTP, bringing down data transfer costs during model running, and more bandwidth being directly allocated to computing power itself. The result is that under the same hardware conditions, the multi-token inference speed of end-side Gemini Nano was raised by over 50%, while memory usage significantly decreased, meaning that the same phone can host more complex models without expanding capacity, or support more applications concurrently invoking the same end-side model.
From a techno-economic perspective, more efficient end-side AI rewrites the apportioning of mobile ecology and computing costs: on one hand, the acceleration of multi-token inference and memory compression brings more complex inference down from the cloud to the end, allowing applications to complete personalized inference and long context processing locally, reducing the frequency of calls to remote servers, thereby lowering cloud-side computing expenditures and significantly compressing latency; on the other hand, processing data locally with less uploading naturally strengthens privacy protection, making it attractive for scenarios emphasizing compliance and local data sovereignty. For the entire industry chain, such end-side optimizations represented by zero-copy MTP effectively "amplify" the available computing power of existing terminals without increasing hardware accounting costs, laying the groundwork for a redistribution of returns among terminal manufacturers, application developers, and computing power providers, and providing a quantifiable main line for efficiency improvement in the valuation of technology and computing-related assets.
OpenRouter Data: Open Source Models Approach Cutting Edge
From the "amplification" of end-side computing power to the reshuffling of competition on the model supply side, multi-model comparisons and usage data provided by aggregation service providers like OpenRouter give another clear signal: the performance gap between open-source large models from China and the U.S. and closed-source cutting-edge models has been compressed in practice to about a 3 to 6 month time lag. This indicates that the so-called exclusive period of "cutting-edge" is being rapidly shortened, and the lead of closed-source models on many general tasks resembles a rolling window settled quarterly rather than a historical incremental disparity measured in years.
When the performance gap is narrowed down to a few months, inference cost becomes a more critical decision variable. Currently, open-source large models are characterized by "extremely low cost" inference, making them highly attractive to price-sensitive companies and developers; once performance reaches the usable threshold, the cost curve will naturally drive a portion of workloads from closed-source to open-source, especially in scenarios with high concurrency, multi-token generation, and demands for local deployment. Coupled with the enhancements in end-side computing power, this combination of "performance catching up + extremely low cost" is accelerating the pace of replacing closed-source models globally and quietly changing the revenue distribution structure of computing infrastructure and AI asset valuation.
AI and Asset Pricing Under the Shadow of Conflict
Iran is accused of launching missile and drone attacks on neighboring countries including Bahrain and Kuwait, combined with Netanyahu's public denial of the "two-state solution," putting both security and political pressures on the Middle East. Even though U.S. military facilities are currently not significantly affected, and no personnel casualties have been reported, leaving short-term space for de-escalation, the medium to long-term upward pressure on global risk premiums remains challenging to absorb. With a shrinkage of risk preferences, Google boosts the end-side Gemini Nano multi-token inference speed by over 50% while compressing memory usage through technologies like zero-copy MTP in the Pixel 9 and Pixel 10, while OpenRouter data reveals that the performance gap between open-source large models from China and the U.S. and closed-source cutting-edge models has narrowed to about 3 to 6 months with very low inference costs, indicating that the marginal computing power demand of AI infrastructure is being repriced: the premium space for high-performance hardware increasingly relies on choices of technology stack and deployment forms rather than just pure accumulation of computing power. In this resonant pattern of "upward geopolitical uncertainty + AI acceleration and cost reduction," the weight of funds between safe-haven assets and growth assets will be rebalanced more frequently, forcing investors to account for the elevated security premium brought by conflicts, as well as the ongoing reshaping of technology and computing asset valuation centers due to model performance iterations and cost curve declines.
Join our community to discuss together and become stronger!
Exclusive Hyperliquid benefits for AiCoin: https://app.hyperliquid.xyz/join/AICOIN88
Exclusive Aster benefits for AiCoin: https://www.asterdex.com/zh-CN/referral/9C50e2
On-chain Telegram community: https://t.me/AiCoinWhaleData
On-chain community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin on-chain Twitter: https://x.com/aicoinwhaledata
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
