Company Overview - Lattice Semiconductor has undergone significant changes in the past 17 months, including a restructuring to address inventory issues and declining revenue [3]. - The company faced a challenging inventory situation with six months of inventory in the channel and a revenue drop from $730 million to $500 million projected for 2024 [3]. Strategic Initiatives - A cost-cutting measure was implemented, resulting in a 14% restructuring to stabilize the company's financial position [3]. - The company has set a growth target of 25% for 2026, indicating a positive outlook for future performance [4].

Welcome to the Spartan UltraScale Plus SCU35 Technical Reference Design Demonstration. In this walkthrough, you will explore the AMD Spartan UltraScale Plus SCU35 Evaluation Kit and build the SCU35-IO Technical Reference Design, TRD. You will generate the hardware platform in the AMD Vivado Design Suite, develop and build the standalone software application in the AMD Vitis Unified IDE, create the device image, and finally run and validate the design on the target hardware.This demo requires the Spartan Ult ...

Welcome to the AMD Spartan UltraScale Plus SCU-35 evaluation kit, introduction and features demonstration. In this video, we will introduce the Spartan UltraScale Plus FPGA family, explore the Spartan UltraScale Plus architecture, introduce the SCU-35 evaluation kit, and examine its board components. Let's briefly define what a gate array is and what it means for a device to be field-programmable.A gate array is just a regular arrangement of logic gates. Field-programmable means that the device can be progr ...

Core Viewpoint - MicroCloud Hologram Inc. has proposed an innovative hardware acceleration technology that converts quantum tensor network algorithms into parallel computing circuits on field programmable gate arrays (FPGA), enabling efficient quantum spin model simulation on classical hardware [1][11]. Technology Development - The company focuses on algorithm-hardware co-design, breaking down tensor network algorithms into computational units that can be directly implemented in hardware, thus creating a high-density parallel scalable architecture using FPGA [5][11]. - HOLO's technology achieves a performance improvement of 1.7 times faster than traditional CPU computations and more than 2 times improved energy efficiency [11]. Implementation Details - A Hierarchical Tensor Contraction Pipeline has been constructed, consisting of three main layers: input and scheduling layer, core computing layer, and output and reduction layer [7][8]. - The core computing layer utilizes multiple MAC Arrays to support tensor contraction operations, achieving pipeline-level parallelism for floating-point operations [8]. Future Plans - The company aims to continue developing hardware implementations for various quantum computing core modules, including quantum variational algorithms and quantum machine learning models, to establish a comprehensive quantum algorithm acceleration ecosystem [12]. Financial Position - MicroCloud Hologram Inc. has cash reserves exceeding 3 billion RMB and plans to invest over 400 million USD in the development of blockchain, quantum computing, and other frontier technologies [13].

Core Insights - The communication industry is identified as the largest application field for FPGA technology, highlighting its significance in the sector [1] - Anlu Technology's FPGA chips are noted for their independent research and development, leading product characteristics, and timely expert services, which have led to widespread adoption among major domestic communication manufacturers [1] - The future development and commercialization of advanced communication technologies are expected to further enhance the application prospects of FPGA [1] Company Summary - Anlu Technology has successfully positioned its FPGA chips in the domestic communication market, particularly in applications such as 5G and F5G networks [1] - The company emphasizes its commitment to innovation and service, which contributes to its competitive edge in the FPGA market [1] Industry Summary - The FPGA market is poised for growth due to advancements in communication technologies, indicating a robust future demand [1] - The increasing complexity and requirements of modern communication networks are likely to drive further adoption of FPGA solutions [1]

4/Next Integration Phase: Began integrating the optimized FPGA LinearGKR modules into the full GKR proof pipeline, preparing for end-to-end hardware verification and unified GPU–FPGA co-design testing.Stay tuned for more updates. ...

Technology & Performance Improvement - FPGA is used to resolve the efficiency bottleneck in GPU's random accumulation writes during the prepare phase [1] - Effective bandwidth increased from 1 TB/s to 6 TB/s by using FPGA, a 500% increase [1]

Core Viewpoint - Shenzhen Zhishang Technology Co., Ltd. plans to acquire 99.8555% equity of Shenzhen Hengyang Data Co., Ltd. The independent financial advisor, Wukuang Securities, has provided a special review opinion addressing key issues related to the target asset's operations and financial status, indicating high reliance on major client Alibaba and a significant decline in gross margin for smart computing products, although the company asserts that the collaboration is sustainable and does not pose significant adverse effects on ongoing operations [1][2]. Group 1: Business Model and Revenue - Hengyang Data focuses on the R&D and sales of smart computing and data processing products, with core products including AI computing clusters and integrated intelligent computing machines. The business model is centered on R&D design, with production mainly outsourced to external manufacturers. Revenue for the years 2023, 2024, and Q1 2025 is projected at 215 million, 455 million, and 178 million respectively, with smart computing product revenue increasing from 32.83% to 91.48% of total revenue [2][4]. - The growth of Hengyang Data's business is heavily dependent on orders from Alibaba and its designated manufacturer, Xinhua San. Sales to Xinhua San amounted to 101 million and 56 million in the last year and the first quarter of 2025, accounting for 33.97% and 30.68% of total revenue [2][4]. Group 2: Financial Performance - Hengyang Data's revenue shows significant volatility, with smart computing product revenue reaching 264 million in 2024, a 271.83% increase from 2023, but dropping to 163 million in Q1 2025. The company attributes this fluctuation to product delivery cycles and customer procurement rhythms [4][6]. - The concentration of customers is notable, with sales to Alibaba increasing from 26.84% in 2023 to 79.81% in Q1 2025. Despite the high dependency on a major client, Wukuang Securities notes that Alibaba, as a leading player in the domestic cloud computing market, plans to invest 380 billion in AI infrastructure over the next three years, suggesting a sustainable partnership for Hengyang Data [4][6]. Group 3: Gross Margin Trends - The gross margin for sales to Alibaba has been declining, with figures of 55.39%, 28.38%, and 9.91% over the reporting periods. This decline is attributed to price adjustments following the transition to large-scale delivery and market competition, aligning with the industry norm of "high initial, stable later" gross margins for customized products [3][4]. - Overall gross margin for Hengyang Data's smart computing business has also decreased from 55.48% to 12.65% during the reporting period. The company anticipates that the launch of new products like NSA.A5 and Beidou Seven Stars 2.0 will alleviate these adverse factors by 2026 [4][6]. Group 4: Production and Supply Chain - Hengyang Data employs a "R&D + sales" business model, outsourcing low-value-added production processes to external manufacturers. The top two suppliers, NationGate and Zhuhai Yibo Technology, accounted for 85.68%, 98.61%, and 98.39% of procurement in the reporting periods [5][6]. - The company maintains that its pricing with suppliers is based on competitive bidding and that it will continue to focus on R&D while outsourcing production. As of March 2025, the company's inventory value stands at 172 million, primarily related to a data application solution project for a major client, which is expected to be delivered within 2025 [5][6].

Performance Metrics - Achieved an estimated throughput of 66,700 proofs per second on FPGA during single-round sumcheck evaluation across four fields [1]

FPGA Development - Fiat-Shamir (FS) 状态机在 FPGA 上的设计已启动 [1] - 目标时钟频率为 250 MHz [1] - 流水线阶段和控制逻辑正在积极开发中，以满足 Versal HBM 器件上的时序收敛 [1] Project Updates - 每周四更新项目进展 [1]