Sharpness Aware Minimization (SAM) is an optimization algorithm used in machine learning that aims to improve model generalization. The method seeks to find model parameters that are located in regions of the loss landscape with uniformly low loss values, rather than parameters that only achieve a minimal loss value at a single point. This approach is described as finding "flat" minima instead of "sharp" ones. The rationale is that models trained this way are less sensitive to variations between training and test data, which can lead to better performance on unseen data. The algorithm was introduced in a 2020 paper by a team of researchers including Pierre Foret, Ariel Kleiner, Hossein Mobahi, and Behnam Neyshabur. == Underlying Principle == SAM modifies the standard training objective by minimizing a "sharpness-aware" loss. This is formulated as a minimax problem where the inner objective seeks to find the highest loss value in the immediate neighborhood of the current model weights, and the outer objective minimizes this value: min w max ‖ ϵ ‖ p ≤ ρ L train ( w + ϵ ) + λ ‖ w ‖ 2 2 {\displaystyle \min _{w}\max _{\|\epsilon \|_{p}\leq \rho }L_{\text{train}}(w+\epsilon )+\lambda \|w\|_{2}^{2}} In this formulation: w {\displaystyle w} represents the model's parameters (weights). L train {\displaystyle L_{\text{train}}} is the loss calculated on the training data. ϵ {\displaystyle \epsilon } is a perturbation applied to the weights. ρ {\displaystyle \rho } is a hyperparameter that defines the radius of the neighborhood (an L p {\displaystyle L_{p}} ball) to search for the highest loss. An optional L2 regularization term, scaled by λ {\displaystyle \lambda } , can be included. A direct solution to the inner maximization problem is computationally expensive. SAM approximates it by taking a single gradient ascent step to find the perturbation ϵ {\displaystyle \epsilon } . This is calculated as: ϵ ( w ) = ρ ∇ L train ( w ) ‖ ∇ L train ( w ) ‖ 2 {\displaystyle \epsilon (w)=\rho {\frac {\nabla L_{\text{train}}(w)}{\|\nabla L_{\text{train}}(w)\|_{2}}}} The optimization process for each training step involves two stages. First, an "ascent step" computes a perturbed set of weights, w adv = w + ϵ ( w ) {\displaystyle w_{\text{adv}}=w+\epsilon (w)} , by moving towards the direction of the highest local loss. Second, a "descent step" updates the original weights w {\displaystyle w} using the gradient calculated at these perturbed weights, ∇ L train ( w adv ) {\displaystyle \nabla L_{\text{train}}(w_{\text{adv}})} . This update is typically performed using a standard optimizer like SGD or Adam. == Application and Performance == SAM has been applied in various machine learning contexts, primarily in computer vision. Research has shown it can improve generalization performance in models such as Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) on image datasets including ImageNet, CIFAR-10, and CIFAR-100. The algorithm has also been found to be effective in training models with noisy labels, where it performs comparably to methods designed specifically for this problem. Some studies indicate that SAM and its variants can improve out-of-distribution (OOD) generalization, which is a model's ability to perform well on data from distributions not seen during training. Other areas where it has been applied include gradual domain adaptation and mitigating overfitting in scenarios with repeated exposure to training examples. == Limitations == A primary limitation of SAM is its computational cost. By requiring two gradient computations (one for the ascent and one for the descent) per optimization step, it approximately doubles the training time compared to standard optimizers. The theoretical convergence properties of SAM are still under investigation. Some research suggests that with a constant step size, SAM may not converge to a stationary point. The accuracy of the single gradient step approximation for finding the worst-case perturbation may also decrease during the training process. The effectiveness of SAM can also be domain-dependent. While it has shown benefits for computer vision tasks, its impact on other areas, such as GPT-style language models where each training example is seen only once, has been reported as limited in some studies. Furthermore, while SAM seeks flat minima, some research suggests that not all flat minima necessarily lead to good generalization. The algorithm also introduces the neighborhood size ρ {\displaystyle \rho } as a new hyperparameter, which requires tuning. == Research, Variants, and Enhancements == Active research on SAM focuses on reducing its computational overhead and improving its performance. Several variants have been proposed to make the algorithm more efficient. These include methods that attempt to parallelize the two gradient computations, apply the perturbation to only a subset of parameters, or reduce the number of computation steps required. Other approaches use historical gradient information or apply SAM steps intermittently to lower the computational burden. To improve performance and robustness, variants have been developed that adapt the neighborhood size based on model parameter scales (Adaptive SAM or ASAM) or incorporate information about the curvature of the loss landscape (Curvature Regularized SAM or CR-SAM). Other research explores refining the perturbation step by focusing on specific components of the gradient or combining SAM with techniques like random smoothing. Theoretical work continues to analyze the algorithm's behavior, including its implicit bias towards flatter minima and the development of broader frameworks for sharpness-aware optimization that use different measures of sharpness.
Google Mobile Services
Google Mobile Services (GMS) is a collection of proprietary applications and application programming interfaces (APIs) services from Google that are typically pre-installed on the majority of Android devices, such as smartphones, tablets, and smart TVs. GMS is not a part of the Android Open Source Project (AOSP), which means an Android manufacturer needs to obtain a license from Google in order to legally pre-install GMS on an Android device. This license is provided by Google without any licensing fees except in the EU. == Core applications == The following are core applications that are part of Google Mobile Services: Google Search Google Chrome YouTube Google Play Google Drive Gmail Google Meet Google Maps Google Photos Google TV YouTube Music === Historically === Google+ Google Hangouts Google Wallet Google Play Magazines Google Play Music Google Play Movies & TV Google Duo == Reception, competitors, and regulators == === FairSearch === Numerous European firms filed a complaint to the European Commission stating that Google had manipulated their power and dominance within the market to push their Services to be used by phone manufacturers. The firms were joined under the name FairSearch, and the main firms included were Microsoft, Expedia, TripAdvisor, Nokia and Oracle. FairSearch's major problem with Google's practices was that they believed Google were forcing phone manufacturers to use their Mobile Services. They claimed Google managed this by asking these manufacturers to sign a contract stating that they must preinstall specific Google Mobile Services, such as Maps, Search and YouTube, in order to get the latest version of Android. Google swiftly responded stating that they "continue to work co-operatively with the European Commission". === Aptoide === The third-party Android app store Aptoide also filed an EU competition complaint against Google once again stating that they are misusing their power within the market. Aptoide alleged that Google was blocking third-party app stores from being on Google Play, as well as blocking Google Chrome from downloading any third-party apps and app stores. As of June 2014, Google had not responded to these allegations. === Abuse of Android dominance === In May 2019, Umar Javeed, Sukarma Thapar, Aaqib Javeed vs. Google LLC & Ors. the Competition Commission of India ordered an antitrust probe against Google for abusing its dominant position with Android to block market rivals. In Prima Facie opinion the commission held that mandatory pre-installation of the entire Google Mobile Services (GMS) suite, under Mobile Application Distribution Agreements (MADA), amounts to the imposition of unfair conditions on the device manufacturers. === EU antitrust ruling === On July 18, 2018, the European Commission fined Google €4.34 billion for breaching EU antitrust rules which resulted in a change of licensing policy for the GMS in the EU. A new paid licensing agreement for smartphones and tablets shipped into the EEA was created. The change is that the GMS is now decoupled from the base Android and will be offered under a separate paid licensing agreement. === Privacy policy === At the same time, Google faced problems with various European data protection agencies, most notably In the United Kingdom and France. The problem they faced was that they had a set of 60 rules merged into one, which allowed Google to "track users more closely". Google once again came out and stated that their new policies still abide by European Union laws. === Android distributions without Google Mobile Services === After surveillance and privacy concerns, several custom android distributions have been implemented, such as GrapheneOS, LineageOS, CalyxOS, iodéOS or /e/OS, and they come either without any GMS installed by default or with microG, that adds a compatibility layer.
DeepSeek
Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence (AI) company that develops large language models (LLMs). Based in Hangzhou, Zhejiang, DeepSeek is owned and funded by High-Flyer, a Chinese hedge fund. DeepSeek was founded in July 2023 by Liang Wenfeng, the co-founder of High-Flyer, who also serves as the CEO for both of the companies. The company launched an eponymous chatbot alongside its DeepSeek-R1 model in January 2025. DeepSeek-R1 provided responses comparable to other contemporary large language models, such as OpenAI's GPT-4 and o1. Its training cost was reported to be significantly lower than other LLMs. The company claims that it trained its V3 model for US$6 million—far less than the US$100 million cost for OpenAI's GPT-4 in 2023—and using approximately one-tenth the computing power consumed by Meta's comparable model, Llama 3.1. DeepSeek's success against larger and more established rivals has been described as "upending AI". DeepSeek's models are described as "open-weight", meaning the exact parameters are openly shared, but the training data is not openly licensed. Since the January 2025 debut of DeepSeek-R1, the company has made its new models available under free and open-source software licenses, primarily the MIT License. The company reportedly recruits AI researchers from top Chinese universities and also hires from outside traditional computer science fields to broaden its models' knowledge and capabilities. DeepSeek significantly reduced training expenses for their R1 model by incorporating techniques such as mixture of experts (MoE) layers. The company also trained its models during ongoing trade restrictions on AI chip exports to China, using weaker AI chips intended for export and employing fewer units overall. Observers say this breakthrough sent "shock waves" through the industry which were described as triggering a "Sputnik moment" for the US in the field of artificial intelligence, particularly due to its open-source, cost-effective, and high-performing AI models. This threatened established AI hardware leaders such as Nvidia; Nvidia's share price dropped sharply, losing US$600 billion in market value, the largest single-company decline in U.S. stock market history. == History == === Founding and early years (2016–2023) === In February 2016, High-Flyer was co-founded by AI enthusiast Liang Wenfeng, who had been trading since the 2008 financial crisis while attending Zhejiang University. The company began stock trading using a GPU-dependent deep learning model on 21 October 2016; before then, it had used CPU-based linear models. By the end of 2017, most of its trading was driven by AI. Liang established High-Flyer as a hedge fund focused on developing and using AI trading algorithms, and by 2021 the firm was using AI exclusively, often using Nvidia chips. In 2019, the company began constructing its first computing cluster, Fire-Flyer, at a cost of 200 million yuan; it contained 1,100 GPUs interconnected at 200 Gbit/s and was retired after 1.5 years in operation. By 2021, Liang had started buying large quantities of Nvidia GPUs for an AI project, reportedly obtaining 10,000 Nvidia A100 GPUs before the United States restricted chip sales to China. Computing cluster Fire-Flyer 2 began construction in 2021 with a budget of 1 billion yuan. It was reported that in 2022, Fire-Flyer 2's capacity had been used at over 96%, totaling 56.74 million GPU hours. 27% was used to support scientific computing outside the company. During 2022, Fire-Flyer 2 had 5,000 PCIe A100 GPUs in 625 nodes, each containing 8 GPUs. At the time, it exclusively used PCIe instead of the DGX version of A100, since at the time the models it trained could fit within a single 40 GB GPU VRAM and so there was no need for the higher bandwidth of DGX (i.e., it required only data parallelism but not model parallelism). Later, it incorporated NVLinks and NCCL (Nvidia Collective Communications Library) to train larger models that required model parallelism. On 14 April 2023, High-Flyer announced the launch of an artificial general intelligence (AGI) research lab, stating that the new lab would focus on developing AI tools unrelated to the firm's financial business. Two months later, on 17 July 2023, that lab was spun off into an independent company, DeepSeek, with High-Flyer as its principal investor and backer. Venture capital investors were reluctant to provide funding, as they considered it unlikely that the venture would be able to quickly generate an "exit". === Model releases since 2023 === DeepSeek released its first model, DeepSeek Coder, on 2 November 2023, followed by the DeepSeek-LLM series on 29 November 2023. In January 2024, it released two DeepSeek-MoE models (Base and Chat), and in April 3 DeepSeek-Math models (Base, Instruct, and RL). DeepSeek-V2 was released in May 2024, followed a month later by the DeepSeek-Coder V2 series. In September 2024, DeepSeek V2.5 was introduced and revised in December. On 20 November 2024, the preview of DeepSeek-R1-Lite became available via chat. In December, DeepSeek-V3-Base and DeepSeek-V3 (chat) were released. On 20 January 2025, DeepSeek launched the DeepSeek chatbot—based on the DeepSeek-R1 model—free for iOS and Android. By 27 January, DeepSeek surpassed ChatGPT as the most downloaded freeware app on the iOS App Store in the United States, triggering an 18% drop in Nvidia's share price. On 24 March 2025, DeepSeek released DeepSeek-V3-0324 under the MIT License. On 28 May 2025, DeepSeek released DeepSeek-R1-0528 under the MIT License. The model has been noted for more tightly following official Chinese Communist Party ideology and censorship in its answers to questions than prior models. On 21 August 2025, DeepSeek released DeepSeek V3.1 under the MIT License. This model features a hybrid architecture with thinking and non-thinking modes. It also surpasses prior models like V3 and R1, by over 40% on certain benchmarks like SWE-bench and Terminal-bench. It was updated to V3.1-Terminus on 22 September 2025. V3.2-Exp was released on 29 September 2025. It uses DeepSeek Sparse Attention, a more efficient attention mechanism based on previous research published in February. DeepSeek-V3.2 was released on 1 December 2025, alongside a DeepSeek-V3.2-Speciale variant that focused on reasoning. In February 2026, Anthropic accused DeepSeek of using thousands of fraudulent accounts to generate millions of conversations with Claude to train its own large language models. In April 2026, investors began speaking with DeepSeek for a $300 million funding round, which would bring DeepSeek to a total valuation of $10 billion. On April 24, 2026, DeepSeek released a preview of its V4 series, including the 1.6-trillion parameter DeepSeek-V4-Pro and the 284-billion parameter DeepSeek-V4-Flash, both featuring a 1-million token context window, under the MIT License. DeepSeek's V4 LLM has been adopted by key semiconductor manufacturers and artificial intelligence chipmakers such as Huawei and Cambricon. == Company operation == DeepSeek is headquartered in Hangzhou, Zhejiang, and is owned and funded by High-Flyer. Its co-founder, Liang Wenfeng, serves as CEO. As of May 2024, Liang personally held an 84% stake in DeepSeek through two shell corporations. === Strategy === DeepSeek has stated that it focuses on research and does not have immediate plans for commercialization. This posture also means it can skirt certain provisions of China's AI regulations aimed at consumer-facing technologies. DeepSeek's hiring approach emphasizes skills over lengthy work experience, resulting in many hires fresh out of university. The company likewise recruits individuals without computer science backgrounds to expand the range of expertise incorporated into the models, for instance in poetry or advanced mathematics. According to The New York Times, dozens of DeepSeek researchers have or have previously had affiliations with People's Liberation Army laboratories and the Seven Sons of National Defence. Due to the impact of United States restrictions on chips, DeepSeek refined its algorithms to maximise computational efficiency and thereby leveraged older hardware and reduced energy consumption. DeepSeek also expanded on the African continent as it offers more affordable and less power-hungry AI solutions. The company has bolstered African language models and generated a number of startups, for example in Nairobi. Along with Huawei's storage and cloud computing services, the impact on the tech scene in sub-saharan Africa is considerable. DeepSeek offers local data sovereignty and more flexibility compared to Western AI platforms. == Training framework == High-Flyer/DeepSeek had operated at least two primary computing clusters: Fire-Flyer (萤火一号) and Fire-Flyer 2 (萤火二号). Fire-Flyer 1 was constructed in 2019 and was retired after 1.5 years of operation. Fi
Andrew Ng
Andrew Yan-Tak Ng (Chinese: 吳恩達; born April 18, 1976) is a British-American computer scientist and technology entrepreneur focusing on machine learning and artificial intelligence (AI). Ng was a cofounder and head of Google Brain and was the former Chief Scientist at Baidu. Ng is an adjunct professor at Stanford University (formerly associate professor and Director of its Stanford AI Lab or SAIL). Ng has also worked in online education, cofounding Coursera and DeepLearning.AI. He has spearheaded many efforts to "democratize deep learning" teaching over 8 million students through his online courses. Ng is renowned globally in computer science, recognized in Time magazine's 100 Most Influential People in 2012 and Fast Company's Most Creative People in 2014. His influence extends to being named in the Time100 AI Most Influential People in 2023. In 2018, he launched and currently heads the AI Fund, initially a $175-million investment fund for backing artificial intelligence startups. He has founded Landing AI, which provides AI-powered SaaS products. On April 11, 2024, Amazon announced Ng's appointment to its board of directors. == Early life and education == Andrew Yan-Tak Ng was born in London, in 1976 to Ronald Paul Ng, a hematologist and lecturer at UCL Medical School, and Tisa Ho, an arts administrator working at the London Film Festival. His parents were both immigrants from Hong Kong. His family moved back to Hong Kong and he spent his early childhood there. In 1984 he and his family moved to Singapore. Ng attended and graduated from Raffles Institution. In 1997, he earned his undergraduate degree with a triple major in computer science, statistics, and economics from Carnegie Mellon University in Pittsburgh, Pennsylvania. Between 1996 and 1998 he also conducted research on reinforcement learning, model selection, and feature selection at the AT&T Bell Labs. In 1998, Ng earned his master's degree in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology (MIT) in Cambridge, Massachusetts. At MIT, he built the first publicly available, automatically indexed web-search engine for research papers on the web. It was a precursor to CiteSeerX/ResearchIndex, but specialized in machine learning. In 2002, he received his Doctor of Philosophy (Ph.D.) in Computer Science from the University of California, Berkeley, under the supervision of Michael I. Jordan. His thesis is titled "Shaping and policy search in reinforcement learning" and is well-cited to this day. == Career == === Academia and teaching === Ng started working as an assistant professor at Stanford University in 2002 and as an associate professor in 2009. Ng is a professor at Stanford University departments of Computer Science and electrical engineering. He served as the director of the Stanford Artificial Intelligence Laboratory (SAIL), where he taught students and undertook research related to data mining, big data, and machine learning. His machine learning course CS229 at Stanford is the most popular course offered on campus with over 1,000 students enrolling some years. As of 2020, three of the most popular courses on Coursera are Ng's: Machine Learning (#1), AI for Everyone (#5), Neural Networks and Deep Learning (#6). In 2008, his group at Stanford was one of the first in the US to start advocating the use of GPUs in deep learning. The rationale was that an efficient computation infrastructure could speed up statistical model training by orders of magnitude, ameliorating some of the scaling issues associated with big data. At the time it was a controversial and risky decision, but since then and following Ng's lead, GPUs have become a cornerstone in the field. Since 2017, Ng has been advocating the shift to high-performance computing (HPC) for scaling up deep learning and accelerating progress in the field. In 2012, along with Stanford computer scientist Daphne Koller he cofounded and was CEO of Coursera, a website that offers free online courses to everyone. It took off with over 100,000 students registered for Ng's popular CS229A course. Today, several million people have enrolled in Coursera courses, making the site one of the leading massive open online courses (MOOCs) in the world. === Industry === From 2011 to 2012, he worked at Google, where he founded and directed the Google Brain Deep Learning Project with Jeff Dean, Greg Corrado, and Rajat Monga. In 2014, he joined Baidu as chief scientist, and carried out research related to big data and AI. There he set up several research teams for things like facial recognition and Melody, an AI chatbot for healthcare. He also developed for the company the AI platform called DuerOS and other technologies that positioned Baidu ahead of Google in the discourse and development of AI. In March 2017, he announced his resignation from Baidu. He soon afterward launched DeepLearning.AI, an online series of deep learning courses (including the AI for Good Specialization). Then Ng launched LandingAI, which provides AI-powered SaaS products. In January 2018, Ng unveiled the AI Fund, raising $175 million to invest in new startups. In November 2021, LandingAI secured a $57 million round of series A funding led by McRock Capital, to help enterprises adopt AI. In October 2024, Ng's AI Fund made its first investment in India, backing AI healthcare startup Jivi, which uses AI for diagnoses, treatment recommendations, and administrative tasks. The investment highlights the growth of India's AI sector, expected to reach $22 billion by 2027. === Research === Ng researches primarily in machine learning, deep learning, machine perception, computer vision, and natural language processing; and is one of the world's most famous and influential computer scientists. He's frequently won best paper awards at academic conferences and has had a huge impact on the field of AI, computer vision, and robotics. During graduate school, together with David M. Blei and Michael I. Jordan, Ng co-authored the influential paper that introduced latent Dirichlet allocation (LDA) for his thesis on reinforcement learning for drones. His early work includes the Stanford Autonomous Helicopter project, which developed one of the most capable autonomous helicopters in the world. He was the leading scientist and principal investigator on the STAIR (Stanford Artificial Intelligence Robot) project, which resulted in Robot Operating System (ROS), a widely used open source software robotics platform. His vision to build an AI robot and put a robot in every home inspired Scott Hassan to back him and create Willow Garage. He is also one of the founding team members for the Stanford WordNet project, which uses machine learning to expand the Princeton WordNet database created by Christiane Fellbaum. In 2011, Ng founded the Google Brain project at Google, which developed large-scale artificial neural networks using Google's distributed computing infrastructure. Among its notable results was a neural network trained using deep learning algorithms on 16,000 CPU cores, which learned to recognize cats after watching only YouTube videos, and without ever having been told what a "cat" is. The project's technology is also currently used in the Android operating system's speech recognition system. === Views on AI === Ng thinks that the real threat is contemplating the future of work: "Rather than being distracted by evil killer robots, the challenge to labor caused by these machines is a conversation that academia and industry and government should have." He has emphasized the importance of expanding access to AI education, stating that empowering people around the world to use AI tools is essential to building AI applications. In a December 2023 Financial Times interview, Ng highlighted concerns regarding the impact of potential regulations on open-source AI, emphasizing how reporting, licensing, and liability risks could unfairly burden smaller firms and stifle innovation. He argued that regulating basic technologies like open-source models could hinder progress without markedly enhancing safety. Ng advocated for carefully designed regulations to prevent obstacles to the development and distribution of beneficial AI technologies. In a June 2024 interview with the Financial Times, Ng expressed concerns about proposed AI legislation in California that would have required developers to implement safety mechanisms such as a "kill switch" for advanced models. He described the bill as creating "massive liabilities for science-fiction risks" and said it "stokes fear in anyone daring to innovate." Other critics argued the bill would impose burdens on open-source developers and smaller AI companies. The bill was ultimately vetoed by Governor Gavin Newsom in September 2024. == Online education: massive open online course == In 2011, Stanford launched a total of three massive open online course (MOOCs) on machine learning (CS229a), databases, and AI, taught by Ng
DeepSeek
Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence (AI) company that develops large language models (LLMs). Based in Hangzhou, Zhejiang, DeepSeek is owned and funded by High-Flyer, a Chinese hedge fund. DeepSeek was founded in July 2023 by Liang Wenfeng, the co-founder of High-Flyer, who also serves as the CEO for both of the companies. The company launched an eponymous chatbot alongside its DeepSeek-R1 model in January 2025. DeepSeek-R1 provided responses comparable to other contemporary large language models, such as OpenAI's GPT-4 and o1. Its training cost was reported to be significantly lower than other LLMs. The company claims that it trained its V3 model for US$6 million—far less than the US$100 million cost for OpenAI's GPT-4 in 2023—and using approximately one-tenth the computing power consumed by Meta's comparable model, Llama 3.1. DeepSeek's success against larger and more established rivals has been described as "upending AI". DeepSeek's models are described as "open-weight", meaning the exact parameters are openly shared, but the training data is not openly licensed. Since the January 2025 debut of DeepSeek-R1, the company has made its new models available under free and open-source software licenses, primarily the MIT License. The company reportedly recruits AI researchers from top Chinese universities and also hires from outside traditional computer science fields to broaden its models' knowledge and capabilities. DeepSeek significantly reduced training expenses for their R1 model by incorporating techniques such as mixture of experts (MoE) layers. The company also trained its models during ongoing trade restrictions on AI chip exports to China, using weaker AI chips intended for export and employing fewer units overall. Observers say this breakthrough sent "shock waves" through the industry which were described as triggering a "Sputnik moment" for the US in the field of artificial intelligence, particularly due to its open-source, cost-effective, and high-performing AI models. This threatened established AI hardware leaders such as Nvidia; Nvidia's share price dropped sharply, losing US$600 billion in market value, the largest single-company decline in U.S. stock market history. == History == === Founding and early years (2016–2023) === In February 2016, High-Flyer was co-founded by AI enthusiast Liang Wenfeng, who had been trading since the 2008 financial crisis while attending Zhejiang University. The company began stock trading using a GPU-dependent deep learning model on 21 October 2016; before then, it had used CPU-based linear models. By the end of 2017, most of its trading was driven by AI. Liang established High-Flyer as a hedge fund focused on developing and using AI trading algorithms, and by 2021 the firm was using AI exclusively, often using Nvidia chips. In 2019, the company began constructing its first computing cluster, Fire-Flyer, at a cost of 200 million yuan; it contained 1,100 GPUs interconnected at 200 Gbit/s and was retired after 1.5 years in operation. By 2021, Liang had started buying large quantities of Nvidia GPUs for an AI project, reportedly obtaining 10,000 Nvidia A100 GPUs before the United States restricted chip sales to China. Computing cluster Fire-Flyer 2 began construction in 2021 with a budget of 1 billion yuan. It was reported that in 2022, Fire-Flyer 2's capacity had been used at over 96%, totaling 56.74 million GPU hours. 27% was used to support scientific computing outside the company. During 2022, Fire-Flyer 2 had 5,000 PCIe A100 GPUs in 625 nodes, each containing 8 GPUs. At the time, it exclusively used PCIe instead of the DGX version of A100, since at the time the models it trained could fit within a single 40 GB GPU VRAM and so there was no need for the higher bandwidth of DGX (i.e., it required only data parallelism but not model parallelism). Later, it incorporated NVLinks and NCCL (Nvidia Collective Communications Library) to train larger models that required model parallelism. On 14 April 2023, High-Flyer announced the launch of an artificial general intelligence (AGI) research lab, stating that the new lab would focus on developing AI tools unrelated to the firm's financial business. Two months later, on 17 July 2023, that lab was spun off into an independent company, DeepSeek, with High-Flyer as its principal investor and backer. Venture capital investors were reluctant to provide funding, as they considered it unlikely that the venture would be able to quickly generate an "exit". === Model releases since 2023 === DeepSeek released its first model, DeepSeek Coder, on 2 November 2023, followed by the DeepSeek-LLM series on 29 November 2023. In January 2024, it released two DeepSeek-MoE models (Base and Chat), and in April 3 DeepSeek-Math models (Base, Instruct, and RL). DeepSeek-V2 was released in May 2024, followed a month later by the DeepSeek-Coder V2 series. In September 2024, DeepSeek V2.5 was introduced and revised in December. On 20 November 2024, the preview of DeepSeek-R1-Lite became available via chat. In December, DeepSeek-V3-Base and DeepSeek-V3 (chat) were released. On 20 January 2025, DeepSeek launched the DeepSeek chatbot—based on the DeepSeek-R1 model—free for iOS and Android. By 27 January, DeepSeek surpassed ChatGPT as the most downloaded freeware app on the iOS App Store in the United States, triggering an 18% drop in Nvidia's share price. On 24 March 2025, DeepSeek released DeepSeek-V3-0324 under the MIT License. On 28 May 2025, DeepSeek released DeepSeek-R1-0528 under the MIT License. The model has been noted for more tightly following official Chinese Communist Party ideology and censorship in its answers to questions than prior models. On 21 August 2025, DeepSeek released DeepSeek V3.1 under the MIT License. This model features a hybrid architecture with thinking and non-thinking modes. It also surpasses prior models like V3 and R1, by over 40% on certain benchmarks like SWE-bench and Terminal-bench. It was updated to V3.1-Terminus on 22 September 2025. V3.2-Exp was released on 29 September 2025. It uses DeepSeek Sparse Attention, a more efficient attention mechanism based on previous research published in February. DeepSeek-V3.2 was released on 1 December 2025, alongside a DeepSeek-V3.2-Speciale variant that focused on reasoning. In February 2026, Anthropic accused DeepSeek of using thousands of fraudulent accounts to generate millions of conversations with Claude to train its own large language models. In April 2026, investors began speaking with DeepSeek for a $300 million funding round, which would bring DeepSeek to a total valuation of $10 billion. On April 24, 2026, DeepSeek released a preview of its V4 series, including the 1.6-trillion parameter DeepSeek-V4-Pro and the 284-billion parameter DeepSeek-V4-Flash, both featuring a 1-million token context window, under the MIT License. DeepSeek's V4 LLM has been adopted by key semiconductor manufacturers and artificial intelligence chipmakers such as Huawei and Cambricon. == Company operation == DeepSeek is headquartered in Hangzhou, Zhejiang, and is owned and funded by High-Flyer. Its co-founder, Liang Wenfeng, serves as CEO. As of May 2024, Liang personally held an 84% stake in DeepSeek through two shell corporations. === Strategy === DeepSeek has stated that it focuses on research and does not have immediate plans for commercialization. This posture also means it can skirt certain provisions of China's AI regulations aimed at consumer-facing technologies. DeepSeek's hiring approach emphasizes skills over lengthy work experience, resulting in many hires fresh out of university. The company likewise recruits individuals without computer science backgrounds to expand the range of expertise incorporated into the models, for instance in poetry or advanced mathematics. According to The New York Times, dozens of DeepSeek researchers have or have previously had affiliations with People's Liberation Army laboratories and the Seven Sons of National Defence. Due to the impact of United States restrictions on chips, DeepSeek refined its algorithms to maximise computational efficiency and thereby leveraged older hardware and reduced energy consumption. DeepSeek also expanded on the African continent as it offers more affordable and less power-hungry AI solutions. The company has bolstered African language models and generated a number of startups, for example in Nairobi. Along with Huawei's storage and cloud computing services, the impact on the tech scene in sub-saharan Africa is considerable. DeepSeek offers local data sovereignty and more flexibility compared to Western AI platforms. == Training framework == High-Flyer/DeepSeek had operated at least two primary computing clusters: Fire-Flyer (萤火一号) and Fire-Flyer 2 (萤火二号). Fire-Flyer 1 was constructed in 2019 and was retired after 1.5 years of operation. Fi
Computer Graphics: Principles and Practice
Computer Graphics: Principles and Practice is a textbook written by James D. Foley, Andries van Dam, Steven K. Feiner, John Hughes, Morgan McGuire, David F. Sklar, and Kurt Akeley and published by Addison–Wesley. First published in 1982 as Fundamentals of Interactive Computer Graphics, it is widely considered a classic standard reference book on the topic of computer graphics. It is sometimes known as the bible of computer graphics (due to its size). == Editions == === First Edition === The first edition, published in 1982 and titled Fundamentals of Interactive Computer Graphics, discussed the SGP library, which was based on ACM's SIGGRAPH CORE 1979 graphics standard, and focused on 2D vector graphics. === Second Edition === The second edition, published 1990, was completely rewritten and covered 2D and 3D raster and vector graphics, user interfaces, geometric modeling, anti-aliasing, advanced rendering algorithms and an introduction to animation. The SGP library was replaced by SRGP (Simple Raster Graphics Package), a library for 2D raster primitives and interaction handling, and SPHIGS (Simple PHIGS), a library for 3D primitives, which were specifically written for the book. === Second Edition in C === In the second edition in C, published in 1995, all examples were converted from Pascal to C. New implementations for the SRGP and SPHIGS graphics packages in C were also provided. === Third Edition === A third edition covering modern GPU architecture was released in July 2013. Examples in the third edition are written in C++, C#, WPF, GLSL, OpenGL, G3D, or pseudocode. == Awards == The book has won a Front Line Award (Hall of Fame) in 1998.
Open Mind Common Sense
Open Mind Common Sense (OMCS) is an artificial intelligence project based at the Massachusetts Institute of Technology (MIT) Media Lab whose goal is to build and utilize a large commonsense knowledge base from the contributions of many thousands of people across the Web. It has been active from 1999 to 2016. Since its founding, it has accumulated more than a million English facts from over 15,000 contributors in addition to knowledge bases in other languages. Much of OMCS's software is built on three interconnected representations: the natural language corpus that people interact with directly, a semantic network built from this corpus called ConceptNet, and a matrix-based representation of ConceptNet called AnalogySpace that can infer new knowledge using dimensionality reduction. The knowledge collected by Open Mind Common Sense has enabled research projects at MIT and elsewhere. == History == The project was the brainchild of Marvin Minsky, Push Singh, Catherine Havasi, and others. Development work began in September 1999, and the project opened to the Internet a year later. Havasi described it in her dissertation as "an attempt to ... harness some of the distributed human computing power of the Internet, an idea which was then only in its early stages." The original OMCS was influenced by the website Everything2 and its predecessor, and presents a minimalist interface that is inspired by Google. Push Singh would have become a professor at the MIT Media Lab and lead the Common Sense Computing group in 2007, but committed suicide on February 28, 2006. The project is currently run by the Digital Intuition Group at the MIT Media Lab under Catherine Havasi. == Database and website == There are many different types of knowledge in OMCS. Some statements convey relationships between objects or events, expressed as simple phrases of natural language: some examples include "A coat is used for keeping warm", "The sun is very hot", and "The last thing you do when you cook dinner is wash your dishes". The database also contains information on the emotional content of situations, in such statements as "Spending time with friends causes happiness" and "Getting into a car wreck makes one angry". OMCS contains information on people's desires and goals, both large and small, such as "People want to be respected" and "People want good coffee". Originally, these statements could be entered into the Web site as unconstrained sentences of text, which had to be parsed later. The current version of the Web site collects knowledge only using more structured fill-in-the-blank templates. OMCS also makes use of data collected by the Game With a Purpose "Verbosity". In its native form, the OMCS database is simply a collection of these short sentences that convey some common knowledge. In order to use this knowledge computationally, it has to be transformed into a more structured representation. == ConceptNet == ConceptNet is a semantic network based on the information in the OMCS database. ConceptNet is expressed as a directed graph whose nodes are concepts, and whose edges are assertions of common sense about these concepts. Concepts represent sets of closely related natural language phrases, which could be noun phrases, verb phrases, adjective phrases, or clauses. ConceptNet is created from the natural-language assertions in OMCS by matching them against patterns using a shallow parser. Assertions are expressed as relations between two concepts, selected from a limited set of possible relations. The various relations represent common sentence patterns found in the OMCS corpus, and in particular, every "fill-in-the-blanks" template used on the knowledge-collection Web site is associated with a particular relation. The data structures that make up ConceptNet were significantly reorganized in 2007, and published as ConceptNet 3. The Software Agents group currently distributes a database and API for the new version 4.0. In 2010, OMCS co-founder and director Catherine Havasi, with Robyn Speer, Dennis Clark and Jason Alonso, created Luminoso, a text analytics software company that builds on ConceptNet. It uses ConceptNet as its primary lexical resource in order to help businesses make sense of and derive insight from vast amounts of qualitative data, including surveys, product reviews and social media. == Machine learning tools == The information in ConceptNet can be used as a basis for machine learning algorithms. One representation, called AnalogySpace, uses singular value decomposition to generalize and represent patterns in the knowledge in ConceptNet, in a way that can be used in AI applications. Its creators distribute a Python machine learning toolkit called Divisi for performing machine learning based on text corpora, structured knowledge bases such as ConceptNet, and combinations of the two. == Comparison to other projects == Other similar projects include Never-Ending Language Learning, Mindpixel (discontinued), Cyc, Learner, SenticNet, Freebase, YAGO, DBpedia, and Open Mind 1001 Questions, which have explored alternative approaches to collecting knowledge and providing incentive for participation. The Open Mind Common Sense project differs from Cyc because it has focused on representing the common sense knowledge it collected as English sentences, rather than using a formal logical structure. ConceptNet is described by one of its creators, Hugo Liu, as being structured more like WordNet than Cyc, due to its "emphasis on informal conceptual-connectedness over formal linguistic-rigor".