Modern enterprises run diverse workloads—from ETL pipelines and business intelligence dashboards to machine learning training jobs on Slurm/HPC clusters. Across all these environments, one principle ...

Overview: High-Performance Computing (HPC) training spans foundational parallel programming, optimization techniques, ...

We are still mulling over all of the new HPC-AI supercomputer systems that were announced in recent months before and during the SC25 supercomputing conference in St Louis, particularly how the slew ...

Abstract: Cloud Computing allows users to access large computing infrastructures quickly. In the High Performance Computing (HPC) context, public cloud resources emerge as an economical alternative, ...

在 GPU 计算领域，CUDA 曾是无可替代的 "武林秘籍"—— 掌握它，就意味着手握 GPU 加速计算的钥匙。但 2025 年末，英伟达用 CUDA Toolkit 13.1 掀起了一场颠覆性变革，Tile 编程模型横空出世，让 GPU 编程从专业开发者的 "专属特权"，变成了普通开发者触手可及的工具，堪称自 2006 年 CUDA 诞生以来最彻底的范式升级。

As AI adoption continues to surge, companies that provide the infrastructure powering large-scale models are increasingly in ...

12月23日，加利西亚超级计算中心（CESGA）与IQM Quantum Computers及Telefónica签署协议，计划在西班牙安装两台全栈量子计算机，预计2026年6月交付。

甩一个暴论，DSL大战本身就是一场闹剧。 gpu编程或者更广泛的HPC，和以往的CPU编程有一个巨大的区别，GPU编程就一个要求，特么得快。不像CPU，速度，可读维护性，扩展性，跨平台，一切得兼顾。所以即便CPU孕育出了从javescript到C++，rust各种风格迥异的语言，也不表示GPU/HPC需要这么玩儿。

Hello folks，我是 Luga，今天我们来聊一下人工智能应用场景中大语言模型（LLM）底层算力资源支撑设施 - AMD ROCm。 在过去十多年里，GPU 的竞争往往被简化为制程、算力峰值和显存带宽的对比。但随着 AI、HPC ...

Abstract: Exponential growth in the number of parameters used to train deep neural network (DNN)/machine learning (ML) models for artificial intelligence (AI) training/ inference applications requires ...