Core Viewpoint - Inverse Lithography Technology (ILT) is emerging as a revolutionary method in chip design, potentially enhancing performance levels significantly, particularly through its application by TSMC and NVIDIA [2][8]. Group 1: Technology Overview - ILT is not a new technology but is being utilized in innovative ways by leading chip manufacturers [2][6]. - The technology addresses issues arising from the diffraction and distortion of extreme ultraviolet light as it passes through complex optical systems in advanced chip manufacturing [4]. - Traditional methods of mask design are iterative and additive, while ILT employs artificial intelligence to generate optical mask images pixel by pixel, resulting in unique and complex designs [5][11]. Group 2: Industry Implications - TSMC's upcoming N2 process (2nm process) will incorporate ILT technology, although it will initially be used for only a few mask layers [8]. - The application of ILT in future GPU products by companies like NVIDIA is expected to yield significant advancements, akin to generational upgrades in chip nodes without the complications of shorter wavelengths [8]. - The resulting lithography masks from ILT are described as visually striking and complex, paralleling the intricate workings of advanced AI models [11].

Core Insights - The article discusses the critical role of lithography technology in semiconductor manufacturing, particularly the challenges posed by the shrinking feature sizes of chips and the need for advanced correction techniques like Inverse Lithography Technology (ILT) [1][3]. Group 1: Lithography Technology - Lithography is essential for transferring integrated circuit patterns, facing challenges due to optical proximity effects as feature sizes decrease [1]. - Optical Proximity Correction (OPC) has evolved from rule-based to model-driven approaches, but traditional methods struggle with complex chip designs [1][3]. - ILT offers a new solution by mathematically reversing the lithography process, allowing for more precise compensation of optical distortions and etching effects [4][10]. Group 2: D2S, Inc and GPU Acceleration - D2S, Inc is a leader in GPU-accelerated solutions for semiconductor manufacturing, focusing on model-based processing technologies for nanoscale device fabrication [2]. - The introduction of GPU technology has transformed ILT from a theoretical concept to practical application, enabling full-chip processing that was previously limited by CPU capabilities [5][7]. - D2S's innovations, such as pixel-level dose correction (PLDC), enhance the accuracy and efficiency of mask writing processes [4][7]. Group 3: Market Trends and Future Outlook - The global market for computational lithography software is projected to grow from $1.268 billion in 2024 to $2.368 billion by 2031, driven by AI-enhanced solutions [9]. - AI is expected to complement existing technologies by optimizing model construction and accelerating ILT processes, thereby enhancing the overall efficiency of lithography [9][10]. - Full-chip ILT is seen as a foundational technology for future semiconductor generations, potentially enabling significant advancements in chip design and manufacturing processes [10][11].

Investment Rating - The report maintains a "Buy" rating for the company, indicating an expected price increase of over 15% in the next 6-12 months [4]. Core Insights - The company is expected to benefit as a leading AI chip manufacturer due to rapid hardware iteration and a rich software ecosystem, which enhances its competitive edge against rivals [4]. - The demand for AI computing power is anticipated to remain strong, driven by the complexity of models in the inference stage, which require significantly more computing resources compared to earlier generative AI models [2][4]. Summary by Sections Performance Review - The company held the GTC 2025 event on March 18, 2025, showcasing future product launches including GB300 (GB Ultra), Vera Rubin, Rubin Ultra GPUs, and CPO switches for Infiniband and Ethernet [1]. Operational Analysis - The transition from simple generative AI to assistant AI is expected to sustain demand for computing power, with inference requiring 100 times more tokens than before. The company anticipates strong customer demand, with major cloud providers expected to purchase 3.6 million Blackwell GPU dies in 2024 [2]. - Upcoming product releases include GB300 NVL72 in H2 2025, which will feature 288GB HBM3e memory and 1.5 times the computing power of GB200 NVL72. Vera Rubin NVL144 is expected in H2 2026, offering 3.3 times the computing power of GB300 NVL72, and Rubin Ultra NVL576 is projected for H2 2027, with 14 times the computing power of GB300 NVL72 [2]. Software Ecosystem - The company continues to enhance its software ecosystem, launching libraries tailored for various industries, such as cuLitho for lithography and CUDA-Q for quantum computing. Additionally, the introduction of the Dynamo system aims to improve GPU efficiency by assisting with prefill and decode tasks [3]. Profit Forecast and Valuation - The company forecasts net profits of $122.2 billion, $156.9 billion, and $177.9 billion for FY26, FY27, and FY28, respectively, with corresponding P/E ratios of 23, 18, and 16 [4][6].