Acronym for Graphcore “Intelligence Processing Unit”. Use this tag for anything related to the IPU, including general purpose IPU usage, framework specific IPU programming, OS version compatibility, Graphcore tools (command line gc-tools and PopVision Graph Analyser). For framework specific IPU programming questions, add the corresponding tag: [poplar], [tensorflow], [popart], [pytorch].
Questions tagged [ipu]
22 questions
6
votes
1 answer
Failed to attach to any of the Graphcore IPU devices when running simple TensorFlow code example
I've tried running one of Graphcore's GitHub code examples, the Tensorflow simple replication one following the README with --replication-factor 16, and the following error was thrown:
tensorflow.python.framework.errors_impl.InternalError: Failed…
Odysseo
- 81
- 4
5
votes
1 answer
How can Poplar codelets include code from other header files?
Is it possible for codelets to reference code in other files, like header files?
If I have a codelet file
//FileA.cpp
#include "FileB.h"
class SomeCustomVertex : public Vertex {
public:
bool compute() {
int a =…
Thorben Louw
- 101
- 5
5
votes
1 answer
Difference between virtualGraph and pipelineStage Graphcore's PopART/Poplar libraries
I'm trying to implement a basic pipelined model using Graphcore's PopART framework (part of the Poplar API) to speed up my model which is split over multiple processors.
I'm following their example code, but I notice the example does not use the…
n00dle
- 5,949
- 2
- 35
- 48
4
votes
1 answer
Running a Tensorflow program on an IPU Model throws an "Illegal instruction (core dumped)" error
I’m trying to run a TensorFlow2 example from the Graphcore public examples (MNIST). I’m using the IPU model instead of IPU hardware because my machine doesn’t have access to IPU hardware, so I’ve followed the documentation (Running on the IPU Model…
LRVerkin
- 75
- 3
4
votes
1 answer
When dumping an XLA graph from a Graphcore IPU-targeted TensorFlow program, which of the dumped files contains the graph and what do the names mean?
I have a TensorFlow model which is compiled to XLA for use with some Graphcore IPUs. For debug purposes, I am trying to dump the XLA graph to a .dot file to visualise it in my browser.
For this I use the following flags:…
Niels
- 149
- 8
3
votes
1 answer
Loss tensor being pruned out of graph in PopART
I’ve written a very simple PopART program using the C++ interface, but every time I try to compile it to run on an IPU device I get the following error:
terminate called after throwing an instance of ‘popart::error’
what(): Could not find loss…
Niels
- 149
- 8
3
votes
1 answer
Why can’t I run IPU programs as non-root in Docker containers?
I’m trying to run CNN training from Graphcore’s examples repo as a non-root user from Graphcore’s TensorFlow 1.5 Docker image, but it’s throwing:
2020-04-23 11:17:32.960014: I tensorflow/compiler/jit/xla_compilation_cache.cc:250] Compiled cluster…
LRVerkin
- 75
- 3
2
votes
1 answer
Where is the V-IPU software?
I was reading the Graphcore documentation and came across the V-IPU software. The documentation says I can download it from the Graphcore software download portal, however I don't see it anywhere among the downloadable software (link…
Cy-r0
- 43
- 1
- 4
2
votes
1 answer
Porting pre-trained keras models and run them on IPU
I am trying to port two pre-trained keras models into the IPU machine. I managed to load and run them using IPUstrategy.scope but I dont know if i am doing it the right way. I have my pre-trained models in .h5 file format.
I load them this way:
def…
Apostolos Gerakaris
- 23
- 3
2
votes
1 answer
What is the meaning and purpose of the Linux /dev/ipu* device names for graphcore IPUs?
Why do I specify ipu4 and ipu4_ex both to use ipu device in docker like below command?
docker run --device=/dev/**ipu4**:/dev/ipu4 --device=/dev/**ipu4_ex**:/dev/ipu4_ex -ti graphcore/tools gc-inventory
Moon-Kee BahkMZC
- 23
- 2
2
votes
1 answer
How can I implement model parallelism on a Graphcore IPU?
I’ve managed to port a version of my TensorFlow model to a Graphcore IPU and to run with data parallelism. However the full-size model won’t fit on a single IPU and I’m looking for strategies to implement model parallelism.
I’ve not had much luck so…
Niels
- 149
- 8
1
vote
1 answer
How to use AutoML Library with IPU/TPU?
I want to use AutoML Library Autogluon with Paperspace IPU/Kaggle TPU instance for specification reasons (big RAM, big space, and fast training time). For IPU, when I try to fit the Autogluon predictor class, the library only recognizes the…
muazhari
- 77
- 5
1
vote
1 answer
Running stable-diffusion on graphcore IPU's
I have been looking for a version of Stable-Diffusion which would be able to run on the IPU's. Currently (due to the high availability) so far I can find CUDA based ones only.
Now I wonder if there is a way to run scripts/trainers/learning etc that…
Scarlett Marseille
- 13
- 2
1
vote
1 answer
Basic IPU example crashes with ValueError: Expected a parent
I wanted to test the free IPU runtime on Paperspace, as such I made a free account and selected the HuggingFace + IPU notebook.
Afterwards I created the following very simple notebook with Pytorch Lightning to perform classification on MNIST (the…
Caridorc
- 6,222
- 2
- 31
- 46
1
vote
0 answers
Does anyone know what kind of logic gates Graphcore uses?
Does anyone know what kind of logic gates Graphcore uses for their IPU's? Specifically the GC200, not the server racks (MK2000).
austinskumanich
- 11
- 1