Citrini观点：英伟达短期将受益于中国开源AI繁荣，结构性风险在于长期算力独立
Related
Aethir has launched its self-developed open-source LLM API platform, Aethir Mesh, enabling top-tier models such as DeepSeek V4 and Kimi K2.6 to run their inference processes entirely on Aethir GPUs.
BlockBeats reported on June 4th that Aethir officially released Aethir Mesh—a self-developed open-source large-model API platform running on Aethir's decentralized GPU infrastructure. It currently provides a unified API endpoint for direct connection to mainstream open-source models such as DeepSeek V4, Kimi K2.6, GLM-5.1, MiniMax-M2.5, and Qwen3.6-27B. Unlike aggregation routing solutions such as OpenRouter, Aethir Mesh directly provides inference services using Aethir's own computing power, eliminating resale premiums and third-party routing. Aethir Claw users can directly write the Mesh API Key into their agent configuration, achieving a full-stack closed loop from VPS hosting to model inference, ensuring that inference data never leaves the Aethir ecosystem. The platform is now live: https://agent.aethir.com/twitterCNmesh
Citrini analyst Jukan: MediaTek reportedly has secured Meta as an ASIC customer, following its TPU offerings.
According to a post on the X platform by Citrini analyst Jukan, Odaily Odaily, supply chain sources indicate that MediaTek has essentially secured a second ASIC customer besides Google. If all goes as planned, this customer will be Meta, previously rumored. MediaTek, as is customary, does not publicly comment on individual products, customer situations, or market rumors. While Qualcomm currently appears to have key customers such as Meta, Microsoft, and ByteDance, industry insiders believe that MediaTek's continued deepening of its cooperation with Google and its potential to secure a second customer puts it at a disadvantage. Semiconductor supply chain sources state that, considering the order prospects and generational transition pace of major cloud service providers for AI data centers and ASIC products, Google remains the most aggressive and proactive major customer. MediaTek not only has two products codenamed Zebrafish and Humufish, but according to market information and confirmation from industry insiders familiar with ASICs, its participation in the v9 generation Triggerfish is almost certain. This means that from the end of 2026 to 2028, and even into 2029, MediaTek can reliably receive revenue contributions from TPU mass production. Compared to Qualcomm's $15 billion cloud AI revenue target for 2029, it's only a matter of time before MediaTek, holding multiple TPU ASIC orders, reaches the $10 billion level. The industry is also focused on whether MediaTek can successfully secure its second major cloud service provider client. Based on earlier market information and recent supply chain confirmations, MediaTek is still actively collaborating with Meta on ASIC products, likely focusing on AI acceleration chips. Meta recently partnered with Arm and Qualcomm, but the products are all CPU-oriented, with no further concrete news regarding its self-developed AI acceleration chips. IC design industry insiders say that Meta's recent cloud AI development strategy has indeed been rather chaotic and unclear, with multiple adjustments to its internal chip development plans, seeking multiple partners and adopting different solutions in the CPU field alone. Regarding AI acceleration chips, although Meta has officially announced its collaboration with Broadcom, supply chain sources indicate that this has not interrupted the ongoing collaboration between MediaTek and Meta. IC design industry professionals also emphasized that MediaTek and Qualcomm's cloud AI development blueprints are still significantly different. MediaTek is concentrating all its resources on ASIC business, while Qualcomm plans to simultaneously promote customization and standardization, while also covering AI acceleration chips and CPU products.
Micron (MU) is more important than Nvidia! Citrini analysts say AI model inference performance relies more on memory than GPU.
According to Odaily Odaily, in response to the "Micron vs. Nvidia" debate discussed in the community, Jukan, an analyst at Citrini Research, posted on the X platform, stating, "While Micron is not Nvidia, its future importance may surpass Nvidia's. Inference is now directly linked to revenue, but performance improvements are not solely achieved by adding Nvidia GPUs. GPUs often remain idle with low utilization due to memory bottlenecks during inference. For inference, increasing memory is more valuable. The ROI of inference ultimately depends on memory, not GPUs. Therefore, why are people still focused on acquiring Micron through NVIDIA's framework? A more comprehensive approach is needed. Inference is memory."
Citrini analyst Jukan: Samsung Foundry achieved monthly profitability in June, its first since 2023.
According to a post on the X platform by Citrini analyst Jukan, as reported by Odaily Korean media, Samsung Foundry achieved monthly profitability in June, its first such month since 2023. With this positive profit in June, Samsung internally believes the likelihood of a profit turnaround starting in the third quarter has increased. The use of Samsung's 4nm process for HBM4 base dies is considered a key factor driving this turnaround, while improved yield rates are also seen as a major factor supporting the positive monthly profit. Industry sources estimate that Samsung's 4nm process yield has improved to approximately 80%.
Citrini researcher Jukan: Citibank report makes conservative assumptions; Micron Technology (MU)'s performance this quarter may exceed expectations.
According to Odaily Odaily, Jukan, a researcher at the well-known investment research firm Citrini, published an article on the X platform stating that after reading Citibank's earnings preview report, he believes that the institution's assumptions regarding Micron Technology's product pricing are too conservative. If the market consensus is at that level, Micron Technology may have exceeded expectations this quarter. Data shows that Micron Technology (MU) stock is currently trading at $1133.99. Previously, several institutions raised their target price for Micron Technology (MU) to between $1200 and $1500, among which: Citigroup raised its price target for Micron Technology to $1,200; Wedbush raised its price target for Micron Technology to $1,300. Deutsche Bank and investment bank Stifel have raised their target price for Micron to $1,500.
JPMorgan Chase: AI custom chip shipments may surpass GPU shipments in 2027, with Broadcom and Marvell poised for takeoff.
According to Mars Finance, on June 22nd, JPMorgan Chase stated that the custom ASIC market is entering a new growth cycle as large cloud computing companies and tech giants seek to reduce AI computing costs, improve energy efficiency, and move away from a single path dependence on general-purpose GPUs. Broadcom and Marvell are expected to be the biggest beneficiaries of this trend. In a recent semiconductor industry research report, JPMorgan analysts Harlan Sur and Mayur Ramdhani estimated that the digital AI ASIC market will reach approximately $60 billion to $70 billion by 2026, maintaining a compound annual growth rate of over 40% to 50% in the coming years. The report states that Broadcom currently holds approximately 80% to 85% of the high-end ASIC market share, with Marvell ranking second with approximately 10% to 12%. The rapid growth in AI computing demand is changing chip procurement structures. JPMorgan believes that customers such as Google, Amazon, Meta, Microsoft, OpenAI, and SoftBank/Arm are accelerating the development of their own or custom AI processors to achieve better performance, power consumption, and total cost of ownership. Unlike Nvidia and AMD's general-purpose GPUs, ASICs are typically designed for specific customers, software stacks, or platforms, making them more suitable for hyperscale cloud vendors with large-scale internal workloads. The report projects that Broadcom's AI revenue will grow significantly from approximately $20 billion in fiscal year 2025 to over $60 billion in fiscal year 2026, and track to over $150 billion in fiscal year 2027. Its project pipeline includes Google TPU, Meta MTIA, ByteDance AI video and network chips, OpenAI XPU, SoftBank/Arm XPU, and Anthropic's related TPU rack-mount solutions. For Marvell, JPMorgan Chase projects its data center revenue to grow from approximately $6.1 billion in 2025 to approximately $9.3 billion in 2026, reaching approximately $14.6 billion in 2027. Growth drivers include Amazon Trainium 3 and Trainium 4, Microsoft Maia, Google SmartNIC/DPU, CXL controllers, and 800G/1.6T optical DSPs, coherent lite, and early CPO solutions. The report also makes a key prediction: by 2027, the unit shipment of AI ASICs/XPUs will surpass that of GPUs. JPMorgan Chase projects total AI accelerator shipments to reach 23.3 million units in 2027, with GPUs accounting for 10.9 million units (47%) and ASICs/XPUs accounting for 12.5 million units (53%). This means that while GPUs will continue to grow, custom chips are likely to capture a larger share of new AI computing power deployments. JPMorgan Chase, citing Google/Broadcom's TPU7x Ironwood and Nvidia's Blackwell as examples, argues that AI ASICs are competitive in terms of cost-effectiveness and power efficiency. The report shows that the TPU7x Ironwood's FP8 computing power is close to that of the Nvidia B200/B300, but its estimated price is around $13,000, lower than the B200's $35,000 and the B300's $40,000; its computing power per dollar and computing power per watt are also superior to comparable GPUs. This assessment does not imply a rapid decline in Nvidia demand. Instead, it points to a divergence in AI infrastructure investment: GPUs continue to serve general training and inference needs, while cloud vendors' self-developed ASICs will achieve higher penetration rates in large-scale, stable, and predictable internal workloads. For investors, JPMorgan Chase's report reinforces the logic of the AI hardware chain diversifying from GPUs to ASICs, advanced packaging, HBM interfaces, SerDes, optical interconnects, and CPOs. If the predictions in the report come true, Broadcom and Marvell will not just be suppliers of AI networks or connectivity chips, but will become core platform companies in the next phase of AI computing architecture migration.