1 citations
Counting papers with at most 25 authors, so collaboration author lists do not stand in for the people here.
- Argonne National LaboratoryUS1 paper
- Barcelona Supercomputing CenterES1 paper
- Brookhaven National LaboratoryUS1 paper
- Centre for Development of Advanced ComputingIN1 paper
- Centre Inria de l'Université de RennesFR1 paper
- DePaul UniversityUS1 paper
- FernUniversität in HagenDE1 paper
- Forschungszentrum JülichDE1 paper
- General Electric (Norway)NO1 paper
- German Cancer Research CenterDE1 paper
- Hewlett Packard Enterprise (United States)US1 paper
- Hewlett-Packard (United States)US1 paper
3 papers
Minima-KV: Retention-Preserving KV Cache Compression with Mixed-Format Paged Attention
Sergii Kozyrev, Davyd Maiboroda
The key-value (KV) cache is a primary capacity and bandwidth bottleneck in long-context LLM serving. We present Minima-KV, a retention-preserving hierarchy for mixed-format paged a…
A Practical Tensor-Network Compression Pipeline for Production-Scale Large Language Models
Sergii Kozyrev, Davyd Maiboroda
Large language models are limited in deployment by GPU memory and inference latency. We present Minima, a production compression pipeline that learns where and how to structurally…
Workflows Community Summit 2024: Future Trends and Challenges in Scientific Workflows
Rafael Ferreira da Silva, Deborah Bard, Kyle Chard +108
The Workflows Community Summit gathered 111 participants from 18 countries to discuss emerging trends and challenges in scientific workflows, focusing on six key areas: time-sensit…