[NPUW]Add MOE caching properties to NPU plugin (#34648)

### Details:
Add the following NPUW MOE-related properties to _cachingProperties:
- NPUW_MOE_TOKEN_CHUNK_SIZE
- NPUW_MOE_POOL_SIZE
- NPUW_LLM_PREFILL_MOE_HINT
- NPUW_LLM_GENERATE_MOE_HINT

These properties affect model compilation and need to be part of the
cache key.

### Tickets:
 - *ticket-id*

### AI Assistance:
 - *AI assistance used: no / yes*
- *If yes, summarize how AI was used and what human validation was
performed (build/tests/manual checks).*

---------

Signed-off-by: intelgaoxiong <xiong.gao@intel.com>

Author: intelgaoxiong

src/plugins/intel_npu/src/plugin/include/properties.hpp

@@ -182,6 +182,8 @@ class Properties final {
         ov::intel_npu::npuw::partitioning::dcoff_type.name(),
         ov::intel_npu::npuw::partitioning::dcoff_with_scale.name(),
         ov::intel_npu::npuw::partitioning::funcall_for_all.name(),
+        ov::intel_npu::npuw::partitioning::moe_token_chunk_size.name(),
+        ov::intel_npu::npuw::partitioning::moe_pool_size.name(),
         ov::intel_npu::npuw::funcall_async.name(),
         ov::intel_npu::npuw::unfold_ireqs.name(),
         ov::intel_npu::npuw::fallback_exec.name(),
@@ -211,7 +213,9 @@ class Properties final {
         ov::intel_npu::npuw::llm::shared_lm_head_config.name(),
         ov::intel_npu::npuw::llm::additional_shared_lm_head_config.name(),
         ov::intel_npu::npuw::llm::optimize_fp8.name(),
-        ov::intel_npu::npuw::eagle::enabled.name()};
+        ov::intel_npu::npuw::eagle::enabled.name(),
+        ov::intel_npu::npuw::llm::prefill_moe_hint.name(),
+        ov::intel_npu::npuw::llm::generate_moe_hint.name()};
 
     const std::vector<ov::PropertyName> _internalSupportedProperties = {ov::internal::caching_properties.name(),
                                                                         ov::internal::caching_with_mmap.name(),

src/plugins/intel_npu/src/plugin/src/plugin.cpp

@@ -340,6 +340,8 @@ void init_config(const IEngineBackend* backend, OptionsDesc& options, FilteredCo
     REGISTER_OPTION(NPUW_KOKORO);
     REGISTER_OPTION(NPUW_KOKORO_BLOCK_SIZE);
     REGISTER_OPTION(NPUW_KOKORO_OVERLAP_SIZE);
+    REGISTER_OPTION(NPUW_MOE_TOKEN_CHUNK_SIZE);
+    REGISTER_OPTION(NPUW_MOE_POOL_SIZE);
 
     config.enableRuntimeOptions();
 

src/plugins/intel_npu/src/plugin/src/properties.cpp

@@ -659,6 +659,8 @@ void Properties::registerPluginProperties() {
     TRY_REGISTER_SIMPLE_PROPERTY(ov::intel_npu::npuw::kokoro::enabled, NPUW_KOKORO);
     TRY_REGISTER_SIMPLE_PROPERTY(ov::intel_npu::npuw::kokoro::block_size, NPUW_KOKORO_BLOCK_SIZE);
     TRY_REGISTER_SIMPLE_PROPERTY(ov::intel_npu::npuw::kokoro::overlap_size, NPUW_KOKORO_OVERLAP_SIZE);
+    TRY_REGISTER_SIMPLE_PROPERTY(ov::intel_npu::npuw::partitioning::moe_token_chunk_size, NPUW_MOE_TOKEN_CHUNK_SIZE);
+    TRY_REGISTER_SIMPLE_PROPERTY(ov::intel_npu::npuw::partitioning::moe_pool_size, NPUW_MOE_POOL_SIZE);
 
     // 2. Metrics (static device and enviroment properties)
     // ========

src/plugins/intel_npu/tests/functional/internal/plugin/test_properties.hpp

@@ -247,6 +247,8 @@ class PropertiesManagerTests : public ov::test::behavior::OVPluginTestBase,
         REGISTER_OPTION(NPUW_KOKORO);
         REGISTER_OPTION(NPUW_KOKORO_BLOCK_SIZE);
         REGISTER_OPTION(NPUW_KOKORO_OVERLAP_SIZE);
+        REGISTER_OPTION(NPUW_MOE_TOKEN_CHUNK_SIZE);
+        REGISTER_OPTION(NPUW_MOE_POOL_SIZE);
 
         npu_config.enableRuntimeOptions();