DeepSeek v3.2

Search documents
加速了!刚刚,重大突破!发生了什么?
Zheng Quan Shi Bao Wang· 2025-10-09 05:03
大盘加速突破! 10月9日早盘, 沪指突破3900点整数关口,为10年来首次,日内涨幅0.58%。深成指涨超1%,创业板指涨超1%。科创50更是 大涨超5%,半导体产业链持续大爆发。在A股带动之下,香港市场亦由跌转涨。A50大涨超1%。 | 上山 指示 | 深证成指 | 深证综指 | 沪深300 | 创业板指 科创50 | ┣证F50 | 中uF500 中iF1000 | : Y.uF50 | 恒生指数 | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | 品种代码 | 品种名称 | | 贡献点数 | 张中雷% | 昨收盘 | 参与计算股本(万) | 权重% | 名次 | | $ 601138 | 工作営联 | | 4. 75 | 5.71 | 66.010 | 1985818. 25 | 2.14 | 1 | | $ 601899 | 器等班,不 | | 2.49 | 6.49 | 29. 440 | 2058873. 50 | 0.99 | 2 | | $ 688041 | 海米信息 | | 2. 40 | 6.44 | 252. 600 ...
大盘加速突破,发生了什么?
Zheng Quan Shi Bao· 2025-10-09 04:23
10月9日早盘, 沪指突破3900点整数关口,为10年来首次,日内涨幅0.58%。深成指涨超1%,创业板指涨超1%。科创50更是 大涨超5%,半导体产业链持续大爆发。在A股带动之下,香港市场亦由跌转涨。A50大涨超1%。 分析人士认为,市场加速突破,主要有两大原因:一是融资盘加速回补可能是主因,节前最后一个交易日,两市融资余额减 少近338亿元;二是节后外围市场普遍造好,人工智能爆款不断,有色金属迭创新高,给市场提供了炒作的土壤。 加速了! 开门红后如何演绎? 国庆中秋假期全球主要股指和商品价格普遍上涨,其中日本股市、国际铜价涨幅最大,港股先涨后跌,涨幅较小。美国政府 关门扰动、未来降息和衰退的预期使得全球投资者对美元信用和美国主权债务的担忧进一步上升,这推动了金银等贵金属和 比特币价格进一步走高。 与此同时,供给短缺和算力革命逻辑下,铜价近期也明显走强。Sora2、DeepSeek v3.2等AI新品发布,亮点纷呈。英伟达与 AMD构筑生态系统,推动AI基础设施从单一硬件竞争转向系统性生态竞争。那么,开门红之后,市场将如何演绎? 分析人士认为,10月份的市场大概率还是震荡向上的格局。从历史来看,国庆后市场也 ...
DeepSeek突然拥抱国产GPU语言,TileLang对标CUDA替代Triton,华为昇腾Day0官宣支持适配
3 6 Ke· 2025-09-30 02:52
DeepSeek v3.2有一个新改动,在论文里完全没提,只在官方公告中出现一次,却引起墙裂关注。 开源TileLang版本算子,其受关注程度甚至超过新稀疏注意力机制DSA,从画线转发的数量就可以看出来。 海外社区也注意到DeepSeek使用了它而不是OpenAI开发的Triton语言。 有接触过的开发者感叹TileLang是一种非常优雅的语言,只需不到100行代码就能写出比Flash Attention 2原版快30%的注意力实现。 那么什么是TileLang,又为何引人瞩目? 首先,TileLang是一种专门用来开发GPU内核的领域专用语言,性能上可以对标英伟达CUDA,DeepSeek官方推荐使用此版本做实验,在方便调试和快速 迭代上有优势。 更重要的是,TileLang与国产算力生态适配,连华为昇腾都要在第一时间公告对TileLang的支持。 在几周前的华为全联接大会2025的开发者日上,TileLang团队成员董宇骐就介绍了TileLang实现FlashAttention算子开发,代码量从500+行减少至80行,并 保持了与官方版本持平的性能。 此外TileLang团队成员王磊和沐曦集成电路的高级总 ...
DeepSeek突然拥抱国产GPU语言!TileLang对标CUDA替代Triton,华为昇腾Day0官宣支持适配
量子位· 2025-09-30 00:57
Core Viewpoint - The article highlights the significance of TileLang, a domain-specific language for GPU kernel development, which has been adopted by DeepSeek in its v3.2 update, showcasing its performance advantages over traditional methods like Flash Attention 2 [1][6][26]. Group 1: TileLang Overview - TileLang is designed to simplify the development of high-performance GPU/CPU kernels, comparable to NVIDIA's CUDA, and is recommended by DeepSeek for experiments due to its debugging and rapid iteration advantages [6][10]. - The language allows developers to write efficient code with significantly reduced lines, achieving performance parity with existing implementations [5][8]. - TileLang's development is led by a team from Peking University, including key figures such as Wang Lei and Dong Yuqi [15][19]. Group 2: DeepSeek's Adoption of TileLang - DeepSeek's choice to use TileLang was first showcased at the Beijing Zhiyuan Conference in June, where its potential for faster operator implementation was discussed [10][11]. - The integration of TileLang has been recognized by industry leaders, including Huawei, which announced support for the language [7][4]. - DeepSeek's v3.2 release demonstrates that TileLang can effectively be used for model training, validating its capabilities in real-world applications [34][26]. Group 3: Performance and Technical Aspects - TileLang provides three programming interfaces catering to different developer expertise levels, from beginners to performance-focused experts [20][21][23]. - The language's architecture allows for decoupling scheduling space from data flow, enabling more efficient optimization by the compiler [19]. - DeepSeek's implementation of TileLang has resulted in significant performance improvements, with claims of achieving a 30% speed increase over traditional methods [5][27].