[CI] Update job dependency for hardware and CPU jobs (#33237)

Signed-off-by: khluu <khluu000@gmail.com>

.buildkite/hardware_tests/amd.yaml

 group: Hardware
 steps:
   - label: "AMD: :docker: build image"
+    depends_on: ~
     device: amd_cpu
     no_plugin: true
     commands:

.buildkite/hardware_tests/intel.yaml

@@ -16,6 +16,7 @@ steps:
     - bash .buildkite/scripts/hardware_ci/run-hpu-test.sh
 
   - label: "Intel GPU Test"
+    depends_on: ~
     soft_fail: true
     device: intel_gpu
     no_plugin: true

.buildkite/test_areas/misc.yaml

@@ -27,7 +27,7 @@ steps:
     - pytest -v -s entrypoints/openai/correctness/test_lmeval.py::test_lm_eval_accuracy_v1_engine
 
 - label: V1 Others (CPU)
-  depends_on: ~
+  depends_on: image-build-cpu
   source_file_dependencies:
     - vllm/
     - tests/v1
@@ -114,7 +114,7 @@ steps:
   - pytest -v -s utils_
 
 - label: Async Engine, Inputs, Utils, Worker, Config (CPU)
-  depends_on: ~
+  depends_on: image-build-cpu
   timeout_in_minutes: 30
   source_file_dependencies:
   - vllm/

.buildkite/test_areas/models_basic.yaml

@@ -39,6 +39,7 @@ steps:
     - pytest -v -s models/test_transformers.py models/test_registry.py
 
 - label: Basic Models Test (Other CPU) # 5min
+  depends_on: image-build-cpu
   timeout_in_minutes: 10
   source_file_dependencies:
   - vllm/

.buildkite/test_areas/models_multimodal.yaml

@@ -14,6 +14,7 @@ steps:
     - cd .. && VLLM_WORKER_MULTIPROC_METHOD=spawn pytest -v -s tests/models/multimodal/generation/test_whisper.py -m core_model  # Otherwise, mp_method="spawn" doesn't work
 
 - label: Multi-Modal Processor Test (CPU)
+  depends_on: image-build-cpu
   timeout_in_minutes: 60
   source_file_dependencies:
   - vllm/