6 papers
The Anatomy of a Triton Attention Kernel
Burkhard Ringlein, Jan van Lunteren, Radu Stoica +1
A long-standing goal in both industry and academia is to develop an LLM inference platform that is portable across hardware architectures, eliminates the need for low-level hand-tu…
Efficient and Reuseable Cloud Configuration Search Using Discovery Spaces
Michael Johnston, Burkhard Ringlein, Christoph Hagleitner +4
Finding the optimal set of cloud resources to deploy a given workload at minimal cost while meeting a defined service level agreement is an active area of research. Combining tens…
GPU Performance Portability needs Autotuning
Burkhard Ringlein, Thomas Parnell, Radu Stoica
As LLMs grow in complexity, achieving state-of-the-art performance requires tight co-design across algorithms, software, and hardware. Today's reliance on a single dominant platfor…
A System Development Kit for Big Data Applications on FPGA-based Clusters: The EVEREST Approach
Christian Pilato, Subhadeep Banik, Jakub Beranek +28
Modern big data workflows are characterized by computationally intensive kernels. The simulated results are often combined with knowledge extracted from AI models to ultimately sup…
Acceleration-as-a-μService: A Cloud-native Monte-Carlo Option Pricing Engine on CPUs, GPUs and Disaggregated FPGAs
Dionysios Diamantopoulos, Raphael Polig, Burkhard Ringlein +5
The evolution of cloud applications into loosely-coupled microservices opens new opportunities for hardware accelerators to improve workload performance. Existing accelerator techn…
Agile Autotuning of a Transprecision Tensor Accelerator Overlay for TVM Compiler Stack
Dionysios Diamantopoulos, Burkhard Ringlein, Mitra Purandare +2
Specialized accelerators for tensor-operations, such as blocked-matrix operations and multi-dimensional convolutions, have been emerged as powerful architecture choices for high-pe…