Accelerated Computing Academy
A new age of IPU-accelerated discovery in computingApply Now
The Next Era of Accelerated Computing
Computer scientists in academia, research institutes, government labs, supercomputing centers and in business around the world are looking to solve today's most challenging problems through computationally intensive research. In order to build entirely new algorithmic applications across a range of fields, they require advanced computing systems.
The IPU is a new massively parallel processor with a unique MIMD architecture designed to accelerate the world's most compute-intensive data processing applications with a mature software stack to enable fine grained programming.
Designed for scalar processes
SIMD/SIMT architecture. Designed for large blocks of dense contiguous data
Model and data spread across off-chip and small on-chip cache and shared memory
Massively parallel MIMD. Designed for fine-grained, high-performance computing
Model and data tightly coupled large locally distributed SRAM
Solving New Problems in Computational Research
The Accelerated Computing Academy is designed to support C++ computer scientists build entirely new applications that require highly parallel and high-performance compute.
This programme aims to help researchers advance their projects in new and unlimited directions that transcend AI and machine learning.
Whether it's simulation modelling, structural analysis, video processing, or computational finance, we're looking for new and novel ways to leverage IPUs along with our Poplar Software.
We're open to working with academia across multiple fields such as Computer Science, Mathematical Sciences, Physics, Data Analytics, Finance, Energy, Chemistry, Aerodynamics, Biology, Geographical Sciences, and more.
Accelerated Computing Academy members will gain access to a variety of benefits
Access to free IPU hardware in the cloud
Support letters for grant and funding proposals
Internship programmes for undergraduate and PhD students
Support from Graphcore Engineers
Bespoke training workshops and educational materials
Project showcase and developer spotlight promotion
IPU PROGRAMMER'S GUIDE
An introduction to the IPU architecture, programming model and tools available.See more
POPLAR + POPLIBS USER GUIDE
Information on how to use the Poplar graph programming tools to write code for the IPU.See more
POPLAR + POPLIBS API REFERENCE
Details of the functions in the Poplar and PopLibs libraries provided in the Poplar SDK.See more
A set of tutorials to introduce the Poplar graph programming framework and PopLibs libraries.See more
POPLAR C++ HPC COOKBOOK
University of Bristol's guide on low-level programming for HPC workloads using Poplar C++.See more
POPLIBS SOURCE CODE
Open source library of classes and functions to implement and deploy parallel programs on the IPU.See more