website: Added GCN3 page
Change-Id: I448d69a6b2df8e4db1202256f9a47509fe85866c
Reviewed-on: https://gem5-review.googlesource.com/c/public/gem5-website/+/29332
Reviewed-by: Jason Lowe-Power <power.jg@gmail.com>
Maintainer: Jason Lowe-Power <power.jg@gmail.com>
Tested-by: Jason Lowe-Power <power.jg@gmail.com>
diff --git a/_data/documentation.yml b/_data/documentation.yml
index 5098c3c..3f7ee46 100644
--- a/_data/documentation.yml
+++ b/_data/documentation.yml
@@ -115,6 +115,12 @@
- page: Visualization
url: /documentation/general_docs/cpu_models/visualization
+ - title: GPU Models
+ id: gpu_models
+ subitems:
+ - page: GCN3
+ url: /documentation/general_docs/gpu_models/GCN3
+
- title: M5ops
id: m5ops
url: /documentation/general_docs/m5ops
diff --git a/_pages/documentation/general_docs/gpu_models/gcn3.md b/_pages/documentation/general_docs/gpu_models/gcn3.md
new file mode 100644
index 0000000..f7a2811
--- /dev/null
+++ b/_pages/documentation/general_docs/gpu_models/gcn3.md
@@ -0,0 +1,106 @@
+---
+layout: documentation
+title: GCN3 GPU model
+doc: gem5 documentation
+parent: gpu_models
+permalink: /documentation/general_docs/gpu_models/GCN3
+---
+
+# **GCN3**
+
+Table of Contents
+
+1. [Using the model](##Using-the-model)
+2. [ROCm](##ROCm)
+3. [Documentation and Tutorials](##Documentation-and-Tutorials)
+
+The GCN3 GPU is a model that simulates a GPU at the ISA level, as opposed to the intermediate language level. This page will give you a general overview of how to use this model, the software stack the model uses, and provide resources that detail the model and how it is implemented.
+
+## **Using the model**
+
+The [gem5 repository](https://gem5.goooglesource.com/public/gem5) comes with a dockerfile located in `util/dockerfiles/gcn-gpu/`. This dockerfile contains the drivers and libraries needed to run the GPU model
+
+The [gem5-resources repository](https://gem5.googlesource.com/public/gem5-resources/) also comes with a sample application (square) that can be used to verify that the model runs correctly.
+
+#### Building the image
+```
+docker build -t <image_name> .
+```
+
+#### Building gem5 using the image
+The following command assumes the gem5 directory is a subdirectory of your current directory
+```
+docker run --rm -v $PWD/gem5:/gem5 -w /gem5 <image_name> scons -sQ -j$(nproc) build/GCN3_X86/gem5.opt
+```
+
+#### Building a GPU application using the image
+The following command assumes the gem5-resources directory is a subdirectory of your current directory
+```
+docker run --rm -v $PWD/gem5-resources:/gem5-resources -w /gem5-resources <image_name> make square
+```
+
+#### Running the sample application
+The following command assumes that gem5 and gem5-resources are subdirectories of your current directory
+```
+docker run --rm -v $PWD/gem5:/gem5 -v $PWD/gem5-resources:/gem5-resources \
+ -w /gem5 <image_name> \
+ build/GCN3_X86/gem5.opt configs/example/apu_se.py -n2 \
+ --benchmark-root=/gem5-resources/output/test-progs/square \
+ -c square
+```
+
+## **ROCm**
+
+The GCN3 model was designed with enough fidelity to not require an emulated runtime. Instead, the GCN3 model uses the Radeon Open Compute platform (ROCm). ROCm is an open platform from AMD that implements [Heterogeneous Systems Architecture (HSA)](http://www.hsafoundation.com/) principles. More information about the HSA standard can be found on the HSA Foundation's website. More information about ROCm can be found on the [ROCm website](https://rocmdocs.amd.com/en/latest/)
+
+#### Simulation support for ROCm
+The model currently only works with system-call emulation (SE) mode, therefore all kernel level driver functionality is modeled entirely within the SE mode layer of gem5. In particular, the emulated GPU driver supports the necessary `ioctl()` commands it receives from the userspace code. The source for the emulated GPU driver can be found in:
+
+* The GPU compute driver: `src/gpu-compute/gpu_compute_driver.[hh|cc]`
+
+* The HSA device driver: `src/dev/hsa/hsa_driver.[hh|cc]`
+
+The HSA driver code models the basic functionality for an HSA agent, which is any device that can be targeted by the HSA runtime and accepts Architected Query Language (AQL) packets. AQL packets are a standard format for all HSA agents, and are used primarily to initiate kernel launches on the GPU. The base `HSADriver` class holds a pointer to the HSA packet processor for the device, and defines the interface for any HSA device. An HSA agent does not have to be a GPU, it could be a generic accelerator, CPU, NIC, etc.
+
+The `GPUComputeDriver` derives from `HSADriver` and is a device-specific implementation of an `HSADriver`. It provides the implementation for GPU-specific `ioctl()` calls.
+
+The `src/dev/hsa/kfd_ioctl.h` header must match the `kfd_ioctl.h` header that comes with ROCt. The emulated driver relies on that file to interpret the `ioctl()` codes the thunk uses.
+
+#### ROCm toolchain and software stack
+The GCN3 model supports ROCm version 1.6
+
+The following ROCm components are required:
+* [Heterogeneous Compute Compiler (HCC)](https://github.com/RadeonOpenCompute/hcc)
+* [Radeon Open Compute runtime (ROCr)](https://github.com/RadeonOpenCompute/ROCR-Runtime)
+* [Radeon Open Compute thunk (ROCt)](https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface)
+* [HIP](https://github.com/ROCm-Developer-Tools/HIP)
+
+The following additional components are used to build and run machine learning programs:
+* [hipBLAS](https://github.com/ROCmSoftwarePlatform/hipBLAS/)
+* [rocBLAS](https://github.com/ROCmSoftwarePlatform/rocBLAS/)
+* [MIOpenGEMM](https://github.com/ROCmSoftwarePlatform/MIOpenGEMM/)
+* [MIOpen](https://github.com/ROCmSoftwarePlatform/MIOpen/)
+* [rocm-cmake](https://github.com/RadeonOpenCompute/rocm-cmake/)
+
+For information about installing these components locally, the commands in the GCN3 dockerfile (`util/dockerfiles/gcn-gpu/`) can be followed on an ubuntu 16 machine.
+
+## **Documentation and Tutorials**
+
+#### GCN3 Model
+Describes the GCN3 model
+* [HPCA 2018](https://ieeexplore.ieee.org/document/8327041)
+
+#### gem5 GCN3 ISCA tutorial
+Covers information about the GPU architecture, GCN3 ISA and HW-SW interfaces in gem5. Also provides an introduction to ROCm.
+* [gem5 GCN3 ISCA webpage](http://www.gem5.org/events/isca-2018)
+* [gem5 GCN3 ISCA slides](http://old.gem5.org/wiki/images/1/19/AMD_gem5_APU_simulator_isca_2018_gem5_wiki.pdf)
+
+#### GCN3 ISA
+* [GCN3 ISA](https://gpuopen.com/documentation/amd-isa-documentation/)
+
+#### ROCm Documentation
+Contains further documentation about the ROCm stack, as well as programming guides for using ROCm.
+* [ROCm webpage](https://rocmdocs.amd.com/en/latest/)
+
+#### AMDGPU LLVM Information
+* [LLVM AMDGPU](https://llvm.org/docs/AMDGPUUsage.html)
