Hey there,
We are a group of researchers interested in Systems for Machine Learning at Imperial College London.
We host monthly Seminars in conjunction with the LSDS group and occasional Reading Groups.
Check our upcoming events below.
Upcoming Events
Past Events
Roger Waleffe will present his work MariusGNN: Training Graph Neural Networks over Billion-Scale Graphs on a Single Machine (EuroSys'23)
Fan Mo will present his work PPFL: Privacy-preserving Federated Learning with Trusted Execution Environments (MobiSys'21)
Michael Kuchnik will present his work on Plumber (MLSys'22)
Colin Unger will present his work on Unity (OSDI'22)
Tags
Accelerators
Adversarial Attacks
AGI
Algebraic Optimization
Algorithms
Analytical Modelling
Approximate Methods
Asynchronous
Auto-parallelization
Automatic Differentiation
Big Data
Collective Communication
Communication
Compilers
Data Parallelism
Dataflow
Distributed
DL Training
Domain Specific Languages
Embeddings
Federated Learning
Gang-Scheduling
GEMM
GNN
GNN Training
GPU
Graph Neural Network
Graphs
Input Pipelines
Intel SGX
JAX
LLVM
Memory Hierarchy
Memory Management
Memory Safety
Memory-Efficiency
ML for MLSys
Model Parallelism
MPI
Multi-Modal
Multi-Task
Networking
Non-uniform Parameter Access
Optimization
Parameter Management
Parameter Servers
Pipeline Parallelism
Pipelining
Privacy
Quantization
Sampling
Scale-up
Skew
SPMD
Stencil Computations
Strategy Search
Tensor Parallelism
TPU
Transformers
Trusted Execution Environments
Verification
XLA