paper-with-me

홈 › Papers

Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI Architectures

2025-05-14 · Chenggang Zhao, Chengqi Deng, Chong Ruan, Damai Dai, Huazuo Gao, Jiashi Li, Liyue Zhang, Panpan Huang, Shangyan Zhou, Shirong Ma, Wenfeng Liang, Ying He, Yuqing Wang, Yuxuan Liu, Y. X. Wei

The rapid scaling of large language models (LLMs) has unveiled critical limitations in current hardware architectures, including constraints in memory capacity, computational efficiency, and interconnection bandwidth. DeepSeek-V3, trained on 2,048 NVIDIA H800 GPUs, demonstrates how hardware-aware model co-design can effectively address these challenges, enabling cost-efficient training and inference at scale. This paper presents an in-depth analysis of the DeepSeek-V3/R1 model architecture and its AI infrastructure, highlighting key innovations such as Multi-head Latent Attention (MLA) for enhanced memory efficiency, Mixture of Experts (MoE) architectures for optimized computation-communication trade-offs, FP8 mixed-precision training to unlock the full potential of hardware capabilities, and a Multi-Plane Network Topology to minimize cluster-level network overhead. Building on the hardware bottlenecks encountered during DeepSeek-V3's development, we engage in a broader discussion with academic and industry peers on potential future hardware directions, including precise low-precision computation units, scale-up and scale-out convergence, and innovations in low-latency communication fabrics. These insights underscore the critical role of hardware and model co-design in meeting the escalating demands of AI workloads, offering a practical blueprint for innovation in next-generation AI systems.

📄 PDF Abstract BibTeX arXiv:2505.09343

Code (0)

등록된 구현이 없습니다.

Tasks

Computational EfficiencyMixture-of-Experts

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

DeepSeek LLM: Scaling Open-Source Language Models with Longtermism

2024-01-05 · DeepSeek-AI, :, Xiao Bi, Deli Chen 외

The rapid development of open-source large language models (LLMs) has been truly remarkable. However, the scaling law described in previous literature presents varying conclusions, which casts a dark cloud over scaling L…

DeepSeek: Paradigm Shifts and Technical Evolution in Large AI Models

2025-07-14 · Luolin Xiong, Haofen Wang, Xi Chen, Lu Sheng 외 arxiv

DeepSeek, a Chinese Artificial Intelligence (AI) startup, has released their V3 and R1 series models, which attracted global attention due to their low cost, high performance, and open-source advantages. This paper begin…

DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models

2024-01-11 · Damai Dai, Chengqi Deng, Chenggang Zhao, R. X. Xu 외

In the era of large language models, Mixture-of-Experts (MoE) is a promising architecture for managing computational costs when scaling up model parameters. However, conventional MoE architectures like GShard, which acti…

Language ModellingLarge Language ModelMixture-of-Experts

Evaluating Test-Time Scaling LLMs for Legal Reasoning: OpenAI o1, DeepSeek-R1, and Beyond

2025-03-20 · Yaoyao Yu, Leilei Gan, Yinghao Hu, Bin Wei 외

Recently, Test-Time Scaling Large Language Models (LLMs), such as DeepSeek-R1 and OpenAI o1, have demonstrated exceptional capabilities across various domains and tasks, particularly in reasoning. While these models have…

Legal Reasoning

Quantifying the Capability Boundary of DeepSeek Models: An Application-Driven Performance Analysis

2025-02-16 · Kaikai Zhao, Zhaoxiang Liu, Xuejiao Lei, Jiaojiao Zhao 외

DeepSeek-R1, known for its low training cost and exceptional reasoning capabilities, has achieved state-of-the-art performance on various benchmarks. However, detailed evaluations for DeepSeek Series models from the pers…

Logical ReasoningModel SelectionText Generation