distributed world (#1481)

* feat: world

* feat: tests

* feat: no more backwards

* feat: recv into

* feat: whoops

* feat: test in ci

* feat: some debug logging

* feat: workflow naming

* feat: need to set pythonpath

* feat: just send to same device

.github/workflows/test.yml

@@ -104,8 +104,8 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        task: [optimage, openpilot]
-    name: ${{ matrix.task=='optimage'&&'GPU OPT and IMAGE Tests'||'openpilot (OpenCL) Tests'}}
+        task: [optimage, openpilot, multigpu]
+    name: ${{ matrix.task=='optimage'&&'GPU OPT and IMAGE Tests'|| matrix.task=='openpilot'&&'openpilot (OpenCL) Tests'|| matrix.task=='multigpu'&&'MultiGPU Tests'}}
     runs-on: ubuntu-20.04
     timeout-minutes: 20
 
@@ -154,6 +154,10 @@ jobs:
       - if: ${{ matrix.task == 'openpilot' }}
         name: Test tensor core ops
         run: GPU=1 TC=2 python3 -m pytest -n=auto test/test_ops.py
+      - if: ${{ matrix.task == 'multigpu' }}
+        name: Test multigpu
+        run: |
+          PYTHONPATH="." python test/external/dist/test_world.py
 
   testmetalwebgpu:
     name: Metal and WebGPU Tests