diff --git a/README_GAUDI.md b/README_GAUDI.md
index 5109f7ddf9927..9e289658fd5c2 100644
--- a/README_GAUDI.md
+++ b/README_GAUDI.md
@@ -315,9 +315,9 @@ mark 90% of free device memory at that point as usable. Next, KV cache
 gets allocated, model is warmed up, and HPU Graphs are captured.
 Environment variable `VLLM_GRAPH_RESERVED_MEM` defines the ratio of
 memory reserved for HPU Graphs capture. With its default value
-(`VLLM_GRAPH_RESERVED_MEM=0.4`), 40% of usable memory will be reserved
+(`VLLM_GRAPH_RESERVED_MEM=0.1`), 10% of usable memory will be reserved
 for graph capture (later referred to as \"usable graph memory\"), and
-the remaining 60% will be utilized for KV cache. Environment variable
+the remaining 90% will be utilized for KV cache. Environment variable
 `VLLM_GRAPH_PROMPT_RATIO` determines the ratio of usable graph memory
 reserved for prefill and decode graphs. By default
 (`VLLM_GRAPH_PROMPT_RATIO=0.5`), both stages have equal memory
@@ -445,7 +445,7 @@ Environment variables
 -   `VLLM_SKIP_WARMUP`: if `true`, warmup will be skipped, `false` by
     default
 -   `VLLM_GRAPH_RESERVED_MEM`: percentage of memory dedicated for
-    HPUGraph capture, `0.4` by default
+    HPUGraph capture, `0.1` by default
 -   `VLLM_GRAPH_PROMPT_RATIO`: percentage of reserved graph memory
     dedicated for prompt graphs, `0.5` by default
 -   `VLLM_GRAPH_PROMPT_STRATEGY`: strategy determining order of prompt
diff --git a/docs/source/getting_started/gaudi-installation.rst b/docs/source/getting_started/gaudi-installation.rst
index 27bc0186675f1..5af81210c4159 100644
--- a/docs/source/getting_started/gaudi-installation.rst
+++ b/docs/source/getting_started/gaudi-installation.rst
@@ -243,7 +243,7 @@ Before KV cache gets allocated, model weights are loaded onto the device, and a
 Only after that, ``gpu_memory_utilization`` flag is utilized - at its default value,  will mark 90% of free device memory at that point as usable.
 Next, KV cache gets allocated, model is warmed up, and HPU Graphs are captured. 
 Environment variable ``VLLM_GRAPH_RESERVED_MEM`` defines the ratio of memory reserved for HPU Graphs capture. 
-With its default value (``VLLM_GRAPH_RESERVED_MEM=0.4``), 40% of usable memory will be reserved for graph capture (later referred to as "usable graph memory"), and the remaining 60% will be utilized for KV cache. 
+With its default value (``VLLM_GRAPH_RESERVED_MEM=0.1``), 10% of usable memory will be reserved for graph capture (later referred to as "usable graph memory"), and the remaining 90% will be utilized for KV cache. 
 Environment variable ``VLLM_GRAPH_PROMPT_RATIO`` determines the ratio of usable graph memory reserved for prefill and decode graphs. By default (``VLLM_GRAPH_PROMPT_RATIO=0.5``), both stages have equal memory constraints. 
 Lower value corresponds to less usable graph memory reserved for prefill stage, e.g. ``VLLM_GRAPH_PROMPT_RATIO=0.2`` will reserve 20% of usable graph memory for prefill graphs, and 80% of usable graph memory for decode graphs. 
 
@@ -322,7 +322,7 @@ Environment variables
 **Performance tuning knobs:**
 
 -   ``VLLM_SKIP_WARMUP``: if ``true``, warmup will be skipped, ``false`` by default
--   ``VLLM_GRAPH_RESERVED_MEM``: percentage of memory dedicated for HPUGraph capture, ``0.4`` by default
+-   ``VLLM_GRAPH_RESERVED_MEM``: percentage of memory dedicated for HPUGraph capture, ``0.1`` by default
 -   ``VLLM_GRAPH_PROMPT_RATIO``: percentage of reserved graph memory dedicated for prompt graphs, ``0.5`` by default
 -   ``VLLM_GRAPH_PROMPT_STRATEGY``: strategy determining order of prompt graph capture, ``min_tokens`` or ``max_bs``, ``min_tokens`` by default
 -   ``VLLM_GRAPH_DECODE_STRATEGY``: strategy determining order of decode graph capture, ``min_tokens`` or ``max_bs``, ``max_bs`` by default
diff --git a/vllm/worker/habana_worker.py b/vllm/worker/habana_worker.py
index 9d083915041fe..291a7fc0d4489 100644
--- a/vllm/worker/habana_worker.py
+++ b/vllm/worker/habana_worker.py
@@ -150,7 +150,7 @@ def determine_num_available_blocks(self) -> Tuple[int, int]:
 
         cache_block_size = self.get_cache_block_size_bytes()
         graph_reserved_mem = (float(
-            os.environ.get('VLLM_GRAPH_RESERVED_MEM', '0.4'))
+            os.environ.get('VLLM_GRAPH_RESERVED_MEM', '0.1'))
                               if not self.model_config.enforce_eager else 0)
         graph_headroom = 1 - graph_reserved_mem
         available_hpu_memory = free_hpu_memory * \