diff --git a/.github/workflows/docker-bases.yml b/.github/workflows/docker-bases.yml
index 7eb281d89d..ba1582ac2a 100644
--- a/.github/workflows/docker-bases.yml
+++ b/.github/workflows/docker-bases.yml
@@ -1,5 +1,9 @@
 name: Build base compilers docker images
 
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
 on:
   push:
     paths:
@@ -16,11 +20,11 @@ on:
 
 jobs:
 #######################################################
-################### CPU ###############################
+############## Basic gcc CPU ##########################
 #######################################################
   deploy-cpu-bases:
-    name: ${{ matrix.tag }}
-    runs-on: ${{ matrix.runner }}
+    name: "cpu-base"
+    runs-on: ubuntu-latest
     env:
       DOCKER_BUILDKIT: "1"
 
@@ -46,34 +50,65 @@ jobs:
       - name: cleanup
         run: docker system prune -a -f
 
-      - name: Docker image
+      - name: GCC image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.cpu'
           push: true
-          build-args: |
-            'arch=gcc'
+          target: 'gcc'
+          build-args: 'arch=gcc'
           tags: 'devitocodes/bases:cpu-gcc'
 
-      - name: Docker image
+#######################################################
+############## Intel OneApi CPU #######################
+#######################################################
+  deploy-oneapi-bases:
+    name: "oneapi-base"
+    runs-on: ubuntu-latest
+    env:
+      DOCKER_BUILDKIT: "1"
+
+    steps:
+      - name: Checkout devito
+        uses: actions/checkout@v3
+
+      - name: Check event name
+        run: echo ${{ github.event_name }}
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v2
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v2
+
+      - name: Login to DockerHub
+        uses: docker/login-action@v2
+        with:
+          username: ${{ secrets.DOCKER_USERNAME }}
+          password: ${{ secrets.DOCKER_PASSWORD }}
+
+      - name: cleanup
+        run: docker system prune -a -f
+
+      - name: ICX image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.cpu'
           push: true
-          build-args: |
-            'arch=icx'
+          target: 'icx'
+          build-args: 'arch=icx'
           tags: 'devitocodes/bases:cpu-icx'
 
-      - name: Docker image
+      - name: ICC image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.cpu'
           push: true
-          build-args: |
-            'arch=icc'
+          target: 'icc'
+          build-args: 'arch=icc'
           tags: 'devitocodes/bases:cpu-icc'
 
 #######################################################
@@ -107,42 +142,38 @@ jobs:
       - name: cleanup
         run: docker system prune -a -f
 
-      - name: Docker image
+      - name: NVC image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.nvidia'
           push: true
-          build-args: |
-            'arch=nvc'
+          target: 'nvc'
+          build-args: 'arch=nvc'
           tags: 'devitocodes/bases:nvidia-nvc'
 
-      - name: Docker image
+      - name: NVCC image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.nvidia'
           push: true
-          build-args: |
-            'arch=nvcc'
+          target: 'nvcc'
+          build-args: 'arch=nvcc'
           tags: 'devitocodes/bases:nvidia-nvcc'
 
-      - name: Docker image
+      - name: NVC host image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.nvidia'
           push: true
-          build-args: |
-            'arch=nvc-host'
+          target: 'nvc-host'
+          build-args: 'arch=nvc-host'
           tags: 'devitocodes/bases:cpu-nvc'
 
 #######################################################
 ################### Nvidia clang ######################
-# # clang is only officialy supporting cuda<11.5. The latest nvidia sdk with cuda 11.5 is 21.11 (last of 2021)
-# # We cannot use newer ones such as 22-1 because the package manager automatically installs the latest 2022
-# # that re-install the latest (11.7) version on top of it.
-# # For more info check https://llvm.org/docs/CompileCudaWithLLVM.html and the Prerequisites 
 #######################################################
   deploy-nvidia-clang-base:
     name: "nvidia-clang-base"
@@ -172,15 +203,14 @@ jobs:
       - name: cleanup
         run: docker system prune -a -f
 
-      - name: Docker image
+      - name: Nvidia clang image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.nvidia'
           push: true
-          build-args: |
-            'arch=clang'
-            'ver=nvhpc-21-11'
+          target: 'clang'
+          build-args: 'arch=clang'
           tags: 'devitocodes/bases:nvidia-clang'
 
 #######################################################
@@ -214,20 +244,22 @@ jobs:
       - name: cleanup
         run: docker system prune -a -f
 
-      - name: Docker image
+      - name: AMD image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.amd'
           push: true
+          target: 'aomp'
           tags: devitocodes/bases:amd
 
-      - name: Docker image
+      - name: AMD HIP image
         uses: docker/build-push-action@v3
         with:
           context: .
           file: './docker/Dockerfile.amd'
           push: true
+          target: 'hip'
           build-args: |
             arch=hip
           tags: devitocodes/bases:amd-hip
\ No newline at end of file
diff --git a/.github/workflows/docker-devito.yml b/.github/workflows/docker-devito.yml
index a5f127b8d4..59aeae8b79 100644
--- a/.github/workflows/docker-devito.yml
+++ b/.github/workflows/docker-devito.yml
@@ -18,38 +18,38 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          - base: 'bases:nvidia-nvc'
+          - base: 'test-bases:nvidia-nvc'
             tag: 'nvidia-nvc'
             flag: '--gpus all'
             test: 'tests/test_gpu_openacc.py tests/test_gpu_common.py'
             runner: ["self-hosted", "nvidiagpu"]
 
-          - base: 'bases:nvidia-clang'
+          - base: 'test-bases:nvidia-clang'
             tag: 'nvidia-clang'
             flag: '--gpus all'
             test: 'tests/test_gpu_openmp.py tests/test_gpu_common.py'
             runner: ["self-hosted", "nvidiagpu"]
 
           # Runtime gpu flags from https://hub.docker.com/r/rocm/tensorflow/
-          - base: 'bases:amd'
+          - base: 'test-bases:amd'
             tag: 'amd'
             flag: '--network=host --device=/dev/kfd --device=/dev/dri --ipc=host --group-add video --group-add 109 --cap-add=SYS_PTRACE --security-opt seccomp=unconfined'
             test: 'tests/test_gpu_openmp.py'
             runner: ["self-hosted", "amdgpu"]
 
-          - base: 'bases:cpu-gcc'
+          - base: 'test-bases:cpu-gcc'
             tag: "gcc"
             flag: ''
             test: 'tests/test_operator.py'
             runner: ubuntu-latest
 
-          - base: 'bases:cpu-icc'
+          - base: 'test-bases:cpu-icc'
             tag: "icc"
             flag: ''
             test: 'tests/test_operator.py'
             runner: ubuntu-latest
 
-          - base: 'bases:cpu-icx'
+          - base: 'test-bases:cpu-icx'
             tag: "icx"
             flag: ''
             test: 'tests/test_operator.py'
@@ -90,16 +90,16 @@ jobs:
           tags: |
             type=raw,value=${{ matrix.tag }}-dev
             type=raw,value=${{ matrix.tag }}-latest,enable=${{ github.event_name == 'release' }}
-            type=raw,value=latest,enable=${{ matrix.base == 'bases:cpu-gcc' }}
+            type=raw,value=latest,enable=${{ matrix.base == 'test-bases:cpu-gcc' }}
             type=semver,pattern={{raw}},prefix=${{ matrix.tag }}-,enable=${{ github.event_name == 'release' }}
             # Legacy "gpu" tag
-            type=raw,value=gpu-dev,enable=${{ matrix.base == 'bases:nvidia-nvc' }}
-            type=semver,pattern={{raw}},prefix=gpu-,enable=${{ github.event_name == 'release' &&  matrix.base == 'bases:nvidia-nvc' }}
-            type=semver,pattern={{raw}},value=gpu-latest,enable=${{ github.event_name == 'release' &&  matrix.base == 'bases:nvidia-nvc' }}
+            type=raw,value=gpu-dev,enable=${{ matrix.base == 'test-bases:nvidia-nvc' }}
+            type=semver,pattern={{raw}},prefix=gpu-,enable=${{ github.event_name == 'release' &&  matrix.base == 'test-bases:nvidia-nvc' }}
+            type=semver,pattern={{raw}},value=gpu-latest,enable=${{ github.event_name == 'release' &&  matrix.base == 'test-bases:nvidia-nvc' }}
             # Legacy "cpu" tag
-            type=raw,value=cpu-dev,enable=${{ matrix.base == 'bases:cpu-gcc' }}
-            type=semver,pattern={{raw}},prefix=cpu-,enable=${{ github.event_name == 'release' && matrix.base == 'bases:cpu-gcc' }}
-            type=semver,pattern={{raw}},value=cpu-latest,enable=${{ github.event_name == 'release' && matrix.base == 'bases:cpu-gcc' }}
+            type=raw,value=cpu-dev,enable=${{ matrix.base == 'test-bases:cpu-gcc' }}
+            type=semver,pattern={{raw}},prefix=cpu-,enable=${{ github.event_name == 'release' && matrix.base == 'test-bases:cpu-gcc' }}
+            type=semver,pattern={{raw}},value=cpu-latest,enable=${{ github.event_name == 'release' && matrix.base == 'test-bases:cpu-gcc' }}
 
       - name: Check tags
         run: echo "${{ steps.meta.outputs.tags }}"
diff --git a/.github/workflows/pytest-core-mpi.yml b/.github/workflows/pytest-core-mpi.yml
index 57d2c5b2fc..d6ca8e80e8 100644
--- a/.github/workflows/pytest-core-mpi.yml
+++ b/.github/workflows/pytest-core-mpi.yml
@@ -15,7 +15,7 @@ on:
       - master
 
 jobs:
-  build:
+  test-mpi-basic:
     name: pytest-mpi
     runs-on: ubuntu-20.04
     strategy:
@@ -25,6 +25,7 @@ jobs:
     env:
       DEVITO_LANGUAGE: "openmp"
       DEVITO_ARCH: "gcc-9"
+      OMP_NUM_THREADS: "1"
       CC: "gcc-9"
       CXX: "g++-9"
 
@@ -52,3 +53,22 @@ jobs:
       with:
         token: ${{ secrets.CODECOV_TOKEN }}
         name: pytest-mpi
+
+  test-mpi-docker:
+      name: pytest-mpi
+      runs-on: ubuntu-latest
+      strategy:
+        matrix:
+          arch: [gcc, icc, icx]
+      
+      steps:
+      - name: Checkout devito
+        uses: actions/checkout@v3
+      
+      - name: Build docker image
+        run: |
+          docker build . --file docker/Dockerfile.devito --tag devito_img --build-arg base=devitocodes/bases:cpu-${{ matrix.arch }}
+      
+      - name: Test with pytest
+        run: |
+          docker run --rm -e CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }} -e OMP_NUM_THREADS=1 --name testrun devito_img pytest tests/test_mpi.py
\ No newline at end of file
diff --git a/.github/workflows/pytest-core-nompi.yml b/.github/workflows/pytest-core-nompi.yml
index 738daf79ed..11063b6b36 100644
--- a/.github/workflows/pytest-core-nompi.yml
+++ b/.github/workflows/pytest-core-nompi.yml
@@ -141,7 +141,7 @@ jobs:
     - name: Set run prefix
       run: |
           if [[ "${{ matrix.name }}" =~ "docker" ]]; then
-              echo "RUN_CMD=docker run --rm -e CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }} -e DEVITO_ARCH=${{ matrix.arch }} --name testrun devito_img"  >> $GITHUB_ENV
+              echo "RUN_CMD=docker run --rm -e CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }} --name testrun devito_img"  >> $GITHUB_ENV
           else
               echo "RUN_CMD=" >> $GITHUB_ENV
           fi
diff --git a/devito/arch/compiler.py b/devito/arch/compiler.py
index 01a0c71525..1e7a91a167 100644
--- a/devito/arch/compiler.py
+++ b/devito/arch/compiler.py
@@ -10,7 +10,7 @@
 
 import numpy.ctypeslib as npct
 from codepy.jit import compile_from_string
-from codepy.toolchain import GCCToolchain
+from codepy.toolchain import GCCToolchain, call_capture_output
 
 from devito.arch import (AMDGPUX, Cpu64, M1, NVIDIAX, POWER8, POWER9, GRAVITON,
                          INTELGPUX, IntelSkylake, get_nvidia_cc, check_cuda_runtime,
@@ -54,6 +54,8 @@ def sniff_compiler_version(cc):
         compiler = "clang"
     elif ver.startswith("icc"):
         compiler = "icc"
+    elif ver.startswith("icx"):
+        compiler = "icx"
     elif ver.startswith("pgcc"):
         compiler = "pgcc"
     elif ver.startswith("cray"):
@@ -62,7 +64,7 @@ def sniff_compiler_version(cc):
         compiler = "unknown"
 
     ver = Version("0")
-    if compiler in ["gcc", "icc"]:
+    if compiler in ["gcc", "icc", "icx"]:
         try:
             # gcc-7 series only spits out patch level on dumpfullversion.
             res = run([cc, "-dumpfullversion"], stdout=PIPE, stderr=DEVNULL)
@@ -707,7 +709,17 @@ def __init__(self, *args, **kwargs):
             if mpi_distro != 'IntelMPI':
                 warning("Expected Intel MPI distribution with `%s`, but found `%s`"
                         % (self.__class__.__name__, mpi_distro))
-            self.cflags.append("-cc=%s" % self.CC)
+            self.cflags.insert(0, '-cc=%s' % self.CC)
+
+    def get_version(self):
+        if configuration['mpi']:
+            cmd = [self.cc, "-cc=%s" % self.CC, "--version"]
+        else:
+            cmd = [self.cc, "--version"]
+        result, stdout, stderr = call_capture_output(cmd)
+        if result != 0:
+            raise RuntimeError(f"version query failed: {stderr}")
+        return stdout
 
     def __lookup_cmds__(self):
         self.CC = 'icc'
@@ -720,9 +732,9 @@ def __lookup_cmds__(self):
         # we try to use `mpiicc` first, while `mpicc` is our fallback, which may
         # or may not be an Intel distribution
         try:
-            check_output(["mpiicc", "--version"]).decode("utf-8")
+            check_output(["mpiicc", "-cc=%s" % self.CC, "--version"]).decode("utf-8")
             self.MPICC = 'mpiicc'
-            self.MPICXX = 'mpiicpc'
+            self.MPICXX = 'mpicxx'
         except FileNotFoundError:
             self.MPICC = 'mpicc'
             self.MPICXX = 'mpicxx'
@@ -777,7 +789,7 @@ def __lookup_cmds__(self):
         self.CC = 'icx'
         self.CXX = 'icpx'
         self.MPICC = 'mpicc'
-        self.MPICX = 'mpicx'
+        self.MPICXX = 'mpicxx'
 
 
 class CustomCompiler(Compiler):
diff --git a/devito/core/autotuning.py b/devito/core/autotuning.py
index 23dac69d61..603a78efd6 100644
--- a/devito/core/autotuning.py
+++ b/devito/core/autotuning.py
@@ -122,7 +122,8 @@ def autotune(operator, args, level, mode):
             at_args.update(dict(run))
 
             # Drop run if not at least one block per thread
-            if not configuration['develop-mode'] and nblocks_per_thread.subs(at_args) < 1:
+            if not configuration['develop-mode'] and \
+                    nblocks_per_thread.subs(normalize_args(at_args)) < 1:
                 continue
 
             # Run the Operator
diff --git a/devito/types/dense.py b/devito/types/dense.py
index 5fe38ebc09..5d50e22d9d 100644
--- a/devito/types/dense.py
+++ b/devito/types/dense.py
@@ -16,7 +16,7 @@
 from devito.logger import debug, warning
 from devito.mpi import MPI
 from devito.parameters import configuration
-from devito.symbolics import FieldFromPointer
+from devito.symbolics import FieldFromPointer, normalize_args
 from devito.finite_differences import Differentiable, generate_fd_shortcuts
 from devito.tools import (ReducerMap, as_tuple, c_restrict_void_p, flatten, is_integer,
                           memoized_meth, dtype_to_ctype, humanbytes)
@@ -1580,8 +1580,9 @@ def make(self, shape=None, initializer=None, allocator=None, **kwargs):
                 raise ValueError("Either `shape` or `kwargs` (Operator overrides) "
                                  "must be provided.")
             shape = []
+            args = normalize_args(kwargs)
             for n, i in enumerate(self.shape):
-                v = i.subs(kwargs)
+                v = i.subs(args)
                 if not v.is_Integer:
                     raise ValueError("Couldn't resolve `shape[%d]=%s` with the given "
                                      "kwargs (obtained: `%s`)" % (n, i, v))
diff --git a/devito/types/dimension.py b/devito/types/dimension.py
index 42f6d26b08..19b0026628 100644
--- a/devito/types/dimension.py
+++ b/devito/types/dimension.py
@@ -321,7 +321,8 @@ def _arg_check(self, args, size, interval):
                 upper = interval.upper
             else:
                 # Autopadding causes non-integer upper limit
-                upper = interval.upper.subs(args)
+                from devito.symbolics import normalize_args
+                upper = interval.upper.subs(normalize_args(args))
             if args[self.max_name] + upper >= size:
                 raise InvalidArgument("OOB detected due to %s=%d" % (self.max_name,
                                                                      args[self.max_name]))
diff --git a/docker/Dockerfile.amd b/docker/Dockerfile.amd
index 4b392a09c1..507bccab75 100644
--- a/docker/Dockerfile.amd
+++ b/docker/Dockerfile.amd
@@ -3,7 +3,7 @@
 # Based on  https://github.com/amd/InfinityHub-CI/tree/main/base-gpu-mpi-rocm-docker
 ##############################################################
 
-ARG ROCM_VERSION=5.4.2
+ARG ROCM_VERSION=5.5.1
 ARG arch="aomp"
 
 FROM rocm/dev-ubuntu-22.04:${ROCM_VERSION}-complete as sdk-base
@@ -82,26 +82,8 @@ RUN cd /tmp/ \
     && make -j ${nproc} \
     && make install
 
-# Install OpenMPI
-RUN cd /tmp \
-    && git clone --recursive https://github.com/open-mpi/ompi.git -b ${OMPI_BRANCH} \
-    && cd ompi \
-    && ./autogen.pl \
-    && mkdir build \
-    && cd build \
-    && ../configure --prefix=$OMPI_HOME --with-ucx=$UCX_HOME \
-        --enable-mca-no-build=btl-uct  \
-        --without-verbs \
-        --with-pmix  \
-        --enable-mpi \
-        --enable-mpi-fortran=yes \
-        --disable-debug \
-    && make -j ${nproc} \
-    && make install
-
 # Cleanup
-RUN rm -rf /tmp/ompi && rm -rf /tmp/ucx
-
+RUN rm -rf /tmp/ucx
 
 # Adding OpenMPI and UCX to Environment
 ENV PATH=$OMPI_HOME/bin:$UCX_HOME/bin:$PATH \
@@ -113,7 +95,6 @@ ENV PATH=$OMPI_HOME/bin:$UCX_HOME/bin:$PATH \
     INCLUDE=$OMPI_HOME/include:$UCX_HOME/include:$INCLUDE \
     PKG_CONFIG_PATH=$OMPI_HOME/lib/pkgconfig:$UCX_HOME/lib/pkgconfig:$PKG_CONFIG_PATH
 
-
 # Adding environment variable for Running as ROOT
 ENV OMPI_ALLOW_RUN_AS_ROOT=1
 ENV OMPI_ALLOW_RUN_AS_ROOT_CONFIRM=1
@@ -132,12 +113,37 @@ RUN python3 -m venv /venv && \
     /venv/bin/pip install --no-cache-dir mpi4py && \
     rm -rf ~/.cache/pip
 
+RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
+    rm -rf /var/lib/apt/lists/*
+
+EXPOSE 8888
+CMD ["/bin/bash"]
+
 ########################################################################
 # AOMP for GPUs (OpenMP offloading)
 ########################################################################
 # This will only trigger if arch is aomp since the final stage depends on it
 FROM sdk-base as aomp
 
+# Install OpenMPI
+RUN cd /tmp \
+    && git clone --recursive https://github.com/open-mpi/ompi.git -b ${OMPI_BRANCH} \
+    && cd ompi \
+    && ./autogen.pl \
+    && mkdir build \
+    && cd build \
+    && ../configure --prefix=$OMPI_HOME --with-ucx=$UCX_HOME \
+        CC=amdclang CXX=amdclang++ FC=amdflang F90=amdflang \
+        --enable-mca-no-build=btl-uct  \
+        --without-verbs \
+        --with-pmix  \
+        --enable-mpi \
+        --enable-mpi-fortran=yes \
+        --disable-debug \
+    && make -j ${nproc} \
+    && make install
+RUN RUN rm -rf /tmp/ompi
+
 # MPI env
 ENV OMPI_CC="amdclang"
 ENV OMPI_CXX="amdclang++"
@@ -147,13 +153,31 @@ ENV DEVITO_ARCH="aomp"
 ENV DEVITO_PLATFORM="amdgpuX"
 ENV DEVITO_LANGUAGE="openmp"
 
-
 ########################################################################
 # HIPCC for GPUs (HIP)
 ########################################################################
 # This will only trigger if arch is hip since the final stage depends on it
 FROM sdk-base as hip
 
+# Install OpenMPI
+RUN cd /tmp \
+    && git clone --recursive https://github.com/open-mpi/ompi.git -b ${OMPI_BRANCH} \
+    && cd ompi \
+    && ./autogen.pl \
+    && mkdir build \
+    && cd build \
+    && ../configure --prefix=$OMPI_HOME --with-ucx=$UCX_HOME \
+        CC=hipcc CXX=hipcc FC=hipfort F90=hipfort \
+        --enable-mca-no-build=btl-uct  \
+        --without-verbs \
+        --with-pmix  \
+        --enable-mpi \
+        --enable-mpi-fortran=yes \
+        --disable-debug \
+    && make -j ${nproc} \
+    && make install
+RUN RUN rm -rf /tmp/ompi
+
 # MPI env
 ENV OMPI_CC="hipcc"
 ENV OMPI_CXX="hipcc"
@@ -162,15 +186,3 @@ ENV OMPI_CXX="hipcc"
 ENV DEVITO_ARCH="hip"
 ENV DEVITO_PLATFORM="amdgpuX"
 ENV DEVITO_LANGUAGE="hip"
-
-
-########################################################################
-# Final image
-########################################################################
-FROM ${arch} as final
-
-RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
-    rm -rf /var/lib/apt/lists/*
-
-EXPOSE 8888
-CMD ["/bin/bash"]
diff --git a/docker/Dockerfile.cpu b/docker/Dockerfile.cpu
index 7dfc00423d..02d7e91552 100644
--- a/docker/Dockerfile.cpu
+++ b/docker/Dockerfile.cpu
@@ -1,17 +1,21 @@
+# syntax=docker/dockerfile:1
 ##############################################################
 # This Dockerfile contains the Devito codes and can be built using different base images.
 ##############################################################
 
-ARG pyversion=python:3.7
+ARG pyversion=python:3.9
 ARG arch=gcc
 
 # Base image 
-FROM ${pyversion} as base
+FROM ${pyversion}-slim-bullseye as base
 
 ENV DEBIAN_FRONTEND noninteractive
 
 # Install for basic base not containing it
-RUN apt-get update && apt-get install -y vim wget git flex libnuma-dev tmux numactl hwloc
+RUN apt-get update && \
+    apt-get install -y vim wget git flex libnuma-dev tmux \
+        numactl hwloc curl \
+        autoconf libtool build-essential
 
 # Install tmpi
 RUN curl https://raw.githubusercontent.com/Azrael3000/tmpi/master/tmpi -o /usr/local/bin/tmpi
@@ -19,6 +23,12 @@ RUN curl https://raw.githubusercontent.com/Azrael3000/tmpi/master/tmpi -o /usr/l
 # Install OpenGL library, necessary for the installation of GemPy
 RUN apt-get install -y libgl1-mesa-glx
 
+RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
+    rm -rf /var/lib/apt/lists/*
+
+EXPOSE 8888
+CMD ["/bin/bash"]
+
 ##############################################################
 # GCC standard image
 ##############################################################
@@ -30,7 +40,7 @@ RUN mkdir -p /deps && mkdir -p /opt/openmpi && cd /deps && \
     mkdir build &&  cd build && \
     ../configure --prefix=/opt/openmpi/ \
                  --enable-mca-no-build=btl-uct --enable-mpi1-compatibility && \
-    make -j $(( $(lscpu | awk '/^Socket\(s\)/{ print $2 }') * $(lscpu | awk '/^Core\(s\) per socket/{ print $4 }') )) && \
+    make -j ${nproc} && \
     make install && \
     rm -rf /deps/openmpi
 
@@ -43,46 +53,50 @@ ENV DEVITO_ARCH="gcc"
 ENV DEVITO_LANGUAGE="openmp"
 
 ##############################################################
-# ICC image
+# Intel Oneapi base
 ##############################################################
-FROM base as icc
+FROM base as oneapi
 
- # Download the key to system keyring
- # https://www.intel.com/content/www/us/en/develop/documentation/installation-guide-for-intel-oneapi-toolkits-linux/top/installation/install-using-package-managers/apt.html#apt
+# Download the key to system keyring
+# https://www.intel.com/content/www/us/en/develop/documentation/installation-guide-for-intel-oneapi-toolkits-linux/top/installation/install-using-package-managers/apt.html#apt
 RUN wget -O- https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB | gpg --dearmor > /usr/share/keyrings/oneapi-archive-keyring.gpg
 RUN echo "deb [signed-by=/usr/share/keyrings/oneapi-archive-keyring.gpg] https://apt.repos.intel.com/oneapi all main" > /etc/apt/sources.list.d/oneAPI.list
 
-# Install wanted components only (icc and mpiicc)
-RUN apt-get update -y && apt-get install -y intel-oneapi-compiler-dpcpp-cpp-and-cpp-classic intel-oneapi-mpi-devel
-
 # Set en vars by hand since we can't use entrypoint for intermediate image
 ENV I_ICC_ROOT=/opt/intel/oneapi/compiler/latest/
 ENV I_MPI_ROOT=/opt/intel/oneapi/mpi/latest/
-ENV I_MPI_CFLAGS="-cc=icc"
 ENV PATH $I_MPI_ROOT/libfabric/bin:$I_MPI_ROOT/bin:$I_ICC_ROOT/linux/bin/intel64:$I_ICC_ROOT/linux/bin:${PATH}
 ENV LD_LIBRARY_PATH $I_MPI_ROOT/libfabric/lib:$I_MPI_ROOT/lib/release:$I_MPI_ROOT/lib:$I_ICC_ROOT/linux/lib:$I_ICC_ROOT/linux/lib/x64:$I_ICC_ROOT/linux/compiler/lib/intel64_lin:${LD_LIBRARY_PATH}
 ENV FI_PROVIDER_PATH $I_MPI_ROOT/libfabric/lib/prov:/usr/lib64/libfabric:${LD_LIBRARY_PATH}
 
+##############################################################
+# ICC image
+##############################################################
+FROM oneapi as icc
+
+RUN apt-get update -y && apt-get install -y intel-oneapi-compiler-dpcpp-cpp-and-cpp-classic intel-oneapi-mpi-devel && \
+    apt-get clean && apt-get autoclean && apt-get autoremove && \
+    rm -rf /var/lib/apt/lists/*
+
 # Devito config
 ENV DEVITO_ARCH="icc"
 ENV DEVITO_LANGUAGE="openmp"
 # MPICC compiler for mpi4py
 ENV MPICC=$I_MPI_ROOT/bin/mpiicc
-
-FROM icc as icx
-# Devito config
-ENV DEVITO_ARCH="icx"
-ENV DEVITO_LANGUAGE="openmp"
-# MPICC compiler for mpi4py
-ENV I_MPI_CFLAGS="-cc=icx"
+ENV MPI4PY_FLAGS='CFLAGS="-cc=icc"'
 
 ##############################################################
-# Published image
+# ICX image
 ##############################################################
-FROM ${arch} as final
+FROM oneapi as icx
 
-RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
+RUN apt-get update -y && apt-get install -y intel-oneapi-compiler-dpcpp-cpp intel-oneapi-mpi-devel && \
+    apt-get clean && apt-get autoclean && apt-get autoremove && \
     rm -rf /var/lib/apt/lists/*
 
-EXPOSE 8888
-CMD ["/bin/bash"]
+# Devito config
+ENV DEVITO_ARCH="icx"
+ENV DEVITO_LANGUAGE="openmp"
+# MPICC compiler for mpi4py
+ENV MPICC=$I_MPI_ROOT/bin/mpiicc
+ENV MPI4PY_FLAGS='CFLAGS="-cc=icx"'
diff --git a/docker/Dockerfile.devito b/docker/Dockerfile.devito
index 08bea3a0c3..089ec5b2f8 100644
--- a/docker/Dockerfile.devito
+++ b/docker/Dockerfile.devito
@@ -7,9 +7,14 @@ ARG base=devitocodes/bases:cpu-gcc
 
 FROM $base as builder
 
+# User/Group Ids
+ARG USER_ID=1000
+ARG GROUP_ID=1000
+
 ##################  Install devito ############################################
 # Copy Devito
 ADD . /app/devito
+
 # Remove git files
 RUN rm -rf /app/devito/.git
 
@@ -18,6 +23,7 @@ RUN python3 -m venv /venv && \
     /venv/bin/pip install --no-cache-dir --upgrade pip && \
     /venv/bin/pip install --no-cache-dir jupyter && \
     /venv/bin/pip install --no-cache-dir wheel && \
+    eval "$MPI4PY_FLAGS /venv/bin/pip install --no-cache-dir mpi4py" && \
     /venv/bin/pip install --no-cache-dir -e /app/devito[extras,mpi,tests] && \
     rm -rf ~/.cache/pip
 
@@ -29,6 +35,10 @@ FROM $base as user
 # COPY is much faster than RUN chown by order of magnitude so we have a final step that
 # just copies the built image into the user.
 
+# User/Group Ids
+ARG USER_ID=1000
+ARG GROUP_ID=1000
+
 ## Create App user
 # Set the home directory to our app user's home.
 ENV HOME=/app
diff --git a/docker/Dockerfile.nvidia b/docker/Dockerfile.nvidia
index a8811d5519..b2bfb62fde 100644
--- a/docker/Dockerfile.nvidia
+++ b/docker/Dockerfile.nvidia
@@ -31,8 +31,8 @@ RUN if [ "$ver" = "nvhpc" ]; then \
         apt-get install -y -q --allow-unauthenticated ${ver}; \
     else \
         export year=$(echo $ver | cut -d "-" -f 2) && export minor=$(echo $ver | cut -d "-" -f 3) && \
-        wget https://developer.download.nvidia.com/hpc-sdk/ubuntu/amd64/nvhpc-20${year}_${year}.${minor}_amd64.deb && \
-        apt-get install --allow-unauthenticated -y -q ./nvhpc-20${year}_${year}.${minor}_amd64.deb; \
+        wget https://developer.download.nvidia.com/hpc-sdk/ubuntu/amd64/nvhpc_${year}.${minor}_amd64.deb && \
+        apt-get install --allow-unauthenticated -y -q ./nvhpc_${year}.${minor}_amd64.deb; \
     fi;
 
 RUN curl -sL https://deb.nodesource.com/setup_18.x | bash - && \
@@ -47,7 +47,6 @@ RUN curl -sL https://deb.nodesource.com/setup_18.x | bash - && \
 ARG MPI_VER=HPCX
 ENV MPIVER=${MPI_VER}
 
-
 # nvidia-container-runtime
 ENV NVIDIA_VISIBLE_DEVICES all
 ENV NVIDIA_DRIVER_CAPABILITIES compute,utility
@@ -74,11 +73,21 @@ ENV UCX_TLS=cuda,cuda_copy,cuda_ipc,sm,shm,self
 #ENV UCX_TLS=cuda,cuda_copy,cuda_ipc,sm,shm,self,rc_x,gdr_copy
 
 # Make simlink for path setup since ENV doesn't accept shell commands.
-RUN export NVARCH=$(ls -1 opt/nvidia/hpc_sdk/Linux_x86_64/ | grep '\.' | head -n 1) && \
-    export CUDA_V=$(ls opt/nvidia/hpc_sdk/Linux_x86_64/${NVARCH}/cuda/ | grep '\.') && \
+RUN export NVARCH=$(ls -1 /opt/nvidia/hpc_sdk/Linux_x86_64/ | grep '\.' | head -n 1) && \
+    export CUDA_V=$(ls /opt/nvidia/hpc_sdk/Linux_x86_64/${NVARCH}/cuda/ | grep '\.') && \
     ln -sf /opt/nvidia/hpc_sdk/Linux_x86_64/${NVARCH} /opt/nvhpc && \
-    ln -sf /opt/nvidia/hpc_sdk/Linux_x86_64/${NVARCH}/cuda/${CUDA_V}/extras/CUPTI /opt/CUPTI
-
+    ln -sf /opt/nvidia/hpc_sdk/Linux_x86_64/${NVARCH}/cuda/${CUDA_V}/extras/CUPTI /opt/CUPTI && \
+    ln -sf /opt/nvidia/hpc_sdk/Linux_x86_64/comm_libs/${CUDA_V}/nvshmem /opt/nvhpc/comm_libs/nvshmem && \
+    ln -sf /opt/nvidia/hpc_sdk/Linux_x86_64/comm_libs/${CUDA_V}/nccl /opt/nvhpc/comm_libs/nccl
+
+# Starting nvhpc 23.5 and cuda 12.1, hpcx and openmpi are inside the cuda version folder, only the bin is in the comm_libs path
+RUN export CUDA_V=$(ls /opt/nvhpc/${NVARCH}/cuda/ | grep '\.') && \
+    ls /opt/nvhpc/comm_libs/${CUDA_V}/hpcx/ &&\
+    if [ -d /opt/nvhpc/comm_libs/${CUDA_V}/hpcx ]; then \
+        rm -rf /opt/nvhpc/comm_libs/hpcx && rm -rf /opt/nvhpc/comm_libs/openmpi4 && \
+        ln -sf /opt/nvhpc/comm_libs/${CUDA_V}/hpcx /opt/nvhpc/comm_libs/hpcx && \
+        ln -sf /opt/nvhpc/comm_libs/${CUDA_V}/openmpi4 /opt/nvhpc/comm_libs/openmpi4;\
+    fi;
 # Set base path based on version
 ENV HPCSDK_HOME=/opt/nvhpc
 ENV HPCSDK_CUPTI=/opt/CUPTI
@@ -98,15 +107,14 @@ ENV NVHPC_CUDA_HOME $HPCSDK_HOME/cuda
 ENV CUDA_ROOT $HPCSDK_HOME/cuda/bin
 ENV PATH $HPCSDK_HOME/compilers/bin:$HPCSDK_HOME/cuda/bin:$HPCSDK_HOME/comm_libs/mpi/bin:${PATH}
 ENV LD_LIBRARY_PATH $HPCSDK_HOME/cuda/lib:$HPCSDK_HOME/cuda/lib64:$HPCSDK_HOME/compilers/lib:$HPCSDK_HOME/math_libs/lib64:$HPCSDK_HOME/comm_libs/mpi/lib:$HPCSDK_CUPTI/lib64:bitcomp_DIR:${LD_LIBRARY_PATH}
-ENV CPATH $HPCSDK_HOME/comm_libs/mpi/include:${CPATH}
+ENV CPATH $HPCSDK_HOME/comm_libs/mpi/include:$HPCSDK_HOME/comm_libs/nvshmem/include:$HPCSDK_HOME/comm_libs/nccl/include:${CPATH}
 
 # MPI
 RUN if [ "x$MPI_VER" = "x4" ]; then \
         rm -f  $HPCSDK_HOME/comm_libs/mpi && \
-        ln -sf $HPCSDK_HOME/comm_libs/openmpi4/openmpi-4.0.5 \
+        ln -sf $HPCSDK_HOME/comm_libs/openmpi4/latest \
                $HPCSDK_HOME/comm_libs/mpi ; \
-    fi;  \
-    if [ "$MPI_VER" = "HPCX" ]; then \
+    elif [ "$MPI_VER" = "HPCX" ]; then \
         rm -f  $HPCSDK_HOME/comm_libs/mpi && \
         ln -sf $HPCSDK_HOME/comm_libs/hpcx/latest/ompi \
                $HPCSDK_HOME/comm_libs/mpi ; \
@@ -121,6 +129,12 @@ RUN python3 -m venv /venv && \
     /venv/bin/jupyter serverextension enable dask_labextension && \
     rm -rf ~/.cache/pip
 
+RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
+    rm -rf /var/lib/apt/lists/*
+
+EXPOSE 8888
+CMD ["/bin/bash"]
+
 ########################################################################
 # NVC for GPUs via OpenACC config
 ########################################################################
@@ -130,7 +144,7 @@ FROM sdk-base as nvc
 ADD docker/nvdashboard.json /app/nvdashboard.json
 
 # mpi4py
-RUN CC=nvc CFLAGS="-noswitcherror -tp=px" /venv/bin/pip install --no-cache-dir mpi4py && rm -rf ~/.cache/pip
+ENV MPI4PY_FLAGS='CC=nvc CFLAGS="-noswitcherror -tp=px"'
 
 ENV DEVITO_ARCH="nvc"
 ENV DEVITO_PLATFORM="nvidiaX"
@@ -139,13 +153,7 @@ ENV DEVITO_LANGUAGE="openacc"
 ########################################################################
 # NVC for GPUs via CUDA config
 ########################################################################
-FROM sdk-base as nvcc
-
-# Make devito env vars file and extras
-ADD docker/nvdashboard.json /app/nvdashboard.json
-
-# mpi4py
-RUN CC=nvc CFLAGS="-noswitcherror -tp=px" /venv/bin/pip install --no-cache-dir mpi4py && rm -rf ~/.cache/pip
+FROM nvc as nvcc
 
 ENV DEVITO_ARCH="cuda"
 ENV DEVITO_PLATFORM="nvidiaX"
@@ -154,10 +162,7 @@ ENV DEVITO_LANGUAGE="cuda"
 ########################################################################
 # NVC for CPUs config
 ########################################################################
-FROM sdk-base as nvc-host
-
-# mpi4py
-RUN CC=nvc CFLAGS="-noswitcherror -tp=px" /venv/bin/pip install --no-cache-dir mpi4py && rm -rf ~/.cache/pip
+FROM nvc as nvc-host
 
 ENV DEVITO_ARCH="nvc"
 ENV DEVITO_PLATFORM="cpu64"
@@ -193,8 +198,11 @@ RUN cd /llvm-project/build && \
 # Set path
 ENV PATH /llvm/bin:${PATH}
 ENV LD_LIBRARY_PATH /llvm/lib:${LD_LIBRARY_PATH}
+ENV CPATH /llvm/include:${CPATH}
 
 RUN rm -rf llvm-project
+RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
+    rm -rf /var/lib/apt/lists/*
 
 # Recompile mpi4py with clang
 ENV OMPI_CC="clang"
@@ -204,14 +212,3 @@ RUN /venv/bin/pip install --no-cache-dir mpi4py && rm -rf ~/.cache/pip
 ENV DEVITO_ARCH="clang"
 ENV DEVITO_PLATFORM="nvidiaX"
 ENV DEVITO_LANGUAGE="openmp"
-
-########################################################################
-# Final image
-########################################################################
-FROM ${arch} as final
-
-RUN apt-get clean && apt-get autoclean && apt-get autoremove && \
-    rm -rf /var/lib/apt/lists/*
-
-EXPOSE 8888
-CMD ["/bin/bash"]