Publications

17 results for Burkhard Ringlein

vllm-triton-backend: How to get state-of-the-art performance on NVIDIA and AMD with just triton
- - Burkhard Ringlein
  - Thomas Parnell
  - et al.
- 2025
- PyTorch Conference 2025
Talk
The Anatomy of a Triton Attention Backend
- - Burkhard Ringlein
  - Jan van Lunteren
  - et al.
- 2025
- Triton Developer Conference 2025
Poster
Triton in Action: Real-World Optimizations for Mamba2 and vLLM
- - Jamie Yang
  - Sara Kokkila Schumacher
  - et al.
- 2025
- Triton Developer Conference 2025
Poster
Achieving Platform Portability for vLLM by using Triton Autotuning and Remembering it
- - Burkhard Ringlein
  - Thomas Parnell
- 2024
- Ray Summit 2024
Talk
Composability of Cloud Accelerators in Virtual World Simulations
- - Dionysios Diamantopoulos
  - Burkhard Ringlein
  - et al.
- 2023
- CLOUD 2023
Conference paper
DOSA: Organic Compilation for Neural Network Inference on Distributed FPGAs
- - Burkhard Ringlein
  - Francois Abel
  - et al.
- 2023
- EDGE 2023
Conference paper
Advancing Compilation of DNNs for FPGAs Using Operation Set Architectures
- - Burkhard Ringlein
  - Francois Abel
  - et al.
- 2022
- IEEE Computer Architecture Letters
Paper
OmpSs@cloudFPGA: An FPGA Task-Based Programming Model with Message Passing
- - Juan Miguel De Haro
  - Rubén Cano
  - et al.
- 2022
- IPDPS 2022
Conference paper
A Case for Function-as-a-Service with Disaggregated FPGAs
- - Burkhard Ringlein
  - François Abel
  - et al.
- 2021
- CLOUD 2021
Conference paper
Acceleration-As-A-μService: A Cloud-native Monte-Carlo Option Pricing Engine on CPUs, GPUs and Disaggregated FPGAs
- - Dionysios Diamantopoulos
  - Raphael Polig
  - et al.
- 2021
- CLOUD 2021
Short paper