ggml-org · ggerganov · Aug 7, 2026 · Aug 2, 2026
diff --git a/docs/backend/SYCL.md b/docs/backend/SYCL.md
@@ -449,6 +449,8 @@ Or
 use 1 SYCL GPUs: [0] with Max compute units:512
 ```
 
+User can use the device management in [docs/multi-gpu.md](https://github.com/ggml-org/llama.cpp/blob/master/docs/multi-gpu.md), like parameter `--device SYCL0,SYCL1` to assign one or more devices.
+
 ## Windows
 
 ### Install GPU driver
@@ -763,6 +765,7 @@ Or
 use 1 SYCL GPUs: [0] with Max compute units:512
 ```
 
+User can use the device management in [docs/multi-gpu.md](https://github.com/ggml-org/llama.cpp/blob/master/docs/multi-gpu.md), like parameter `--device SYCL0,SYCL1` to assign one or more devices.
 
 ## Environment Variable
 
@@ -895,6 +898,45 @@ Pass these via `CXXFLAGS` or add a one-off `#define` to enable a flag on the spo
     set UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1
   ```
 
+- When I set `SYCL_CACHE_PERSISTENT=1` in running time, I meet crash.
+
+  `SYCL_CACHE_PERSISTENT=1` is not recommended by llama.cpp SYCL backend.
+  When cache is enabled, SYCL runtime will try to cache and reuse JIT-compiled binaries.
+
+  We find some AI will tell user this cmd to speed up SYCL backend. It only speeds up the startup to skip the JIT process, instead of running speed.
+
+  It will bring negative impact when the SYCL binary file is changed frequently in your running environment. The new & old codes mix will lead to crash.
+
+  Compare to the benefit, it has brought more failed cases.
+  If you are not familiar with the SYCL compiler principle of JIT and AOT, please don't use it.
+
+  To restore, you need to remove the local cache: `~/.cache/libsycl_cache/` and execute `unset SYCL_CACHE_PERSISTENT` in running time.
+
+- How to use iGPU and dGPU in same time?
+
+  1. Detect the devices in your running time.
+  ```
+  source /opt/intel/oneapi/setvars.sh
+  ./build/bin/llama-server --list-devices
+
+  or
+  ./build/bin/llama-cli --list-devices
+  ./build/bin/llama-bench --list-devices
+  ./build/bin/llama-completion --list-devices
+
+  Available devices:
+    SYCL0: Intel(R) Arc(TM) A770 Graphics (15473 MiB, 15473 MiB free)
+    SYCL1: Intel(R) UHD Graphics 770 (59675 MiB, 44986 MiB free)
+  ```
+
+  The dGPU will be in the head of this list and iGPU will be the end.
+  If not all GPUs are listed, please check the env var: ONEAPI_DEVICE_SELECTOR and unset it.
+
+  2. Set the iGPU and dGPU
+
+  Set the iGPU and dGPU by `./build/bin/llama-server --device SYCL0,SYCL1,SYCLxxx`.
+
+
 ### **GitHub contribution**:
 Please add the `[SYCL]` prefix/tag in issues/PRs titles to help the SYCL contributors to check/address them without delay.
 

diff --git a/examples/sycl/start-svr.sh b/examples/sycl/start-svr.sh
@@ -12,6 +12,7 @@ This script processes files with specified options.
 
 Options:
   -h, --help    Display this help message and exit.
+  -d, --device  <value>    Set SYCL devices (default: SYCL0).
   -c, --context <value>    Set context length. Bigger need more memory.
   -p, --promote <value>    Prompt to start generation with.
   -m, --model   <value>    Full model file path.
@@ -41,10 +42,16 @@ MODEL_FILE=../models/Qwen3.5-4B-Q4_0.gguf
 NGL=99
 CONTEXT=4096
 GGML_SYCL_DEVICE=-1
+SYCL_DEVICES="SYCL0"
 SPLIT_MODE=layer
 LOG_VERBOSE=3
 while [[ $# -gt 0 ]]; do
     case "$1" in
+        -d|--device)
+            SYCL_DEVICES="$2"
+            shift
+            shift
+            ;;
         -c|--context)
             CONTEXT=$2
             # Shift twice to consume both the option flag and its value
@@ -95,8 +102,6 @@ while [[ $# -gt 0 ]]; do
     esac
 done
 
-
-
 source /opt/intel/oneapi/setvars.sh
 
 #export GGML_SYCL_DEBUG=1
@@ -107,17 +112,19 @@ source /opt/intel/oneapi/setvars.sh
 export UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1
 echo "UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=${UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS}"
 
+echo "ONEAPI_DEVICE_SELECTOR=${ONEAPI_DEVICE_SELECTOR}"
+
+
 if [ $GGML_SYCL_DEVICE -ne -1 ]; then
     echo "Use $GGML_SYCL_DEVICE as main GPU"
     #use signle GPU only
     GPUS_SETTING="-mg $GGML_SYCL_DEVICE -sm ${SPLIT_MODE}"
-    echo "ONEAPI_DEVICE_SELECTOR=${ONEAPI_DEVICE_SELECTOR}"
 else
-    echo "Use all Intel GPUs, including iGPU & dGPU"
+    echo "Use Intel GPUs: ${SYCL_DEVICES}"
     GPUS_SETTING="-sm ${SPLIT_MODE}"
- fi
+fi
 
-echo "run cmd: ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE}  --mmap --host 0.0.0.0 --port 8000"
-ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE} --mmap --host 0.0.0.0 --port 8000
+echo "run cmd: ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE} --device ${SYCL_DEVICES} --mmap --host 0.0.0.0 --port 8000"
+ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE} --device ${SYCL_DEVICES} --mmap --host 0.0.0.0 --port 8000
 
 
diff --git a/examples/sycl/test.sh b/examples/sycl/test.sh
@@ -12,6 +12,7 @@ This script processes files with specified options.
 
 Options:
   -h, --help    Display this help message and exit.
+  -d, --device  <value>    Set SYCL devices (default: SYCL0).
   -c, --context <value>    Set context length. Bigger need more memory.
   -p, --promote <value>    Prompt to start generation with.
   -m, --model   <value>    Full model file path.
@@ -42,10 +43,16 @@ MODEL_FILE=../models/llama-2-7b.Q4_0.gguf
 NGL=99
 CONTEXT=4096
 GGML_SYCL_DEVICE=-1
+SYCL_DEVICES="SYCL0"
 SPLIT_MODE=layer
 LOG_VERBOSE=3
 while [[ $# -gt 0 ]]; do
     case "$1" in
+        -d|--device)
+            SYCL_DEVICES="$2"
+            shift
+            shift
+            ;;
         -c|--context)
             CONTEXT=$2
             # Shift twice to consume both the option flag and its value
@@ -115,16 +122,17 @@ source /opt/intel/oneapi/setvars.sh
 export UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1
 echo "UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=${UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS}"
 
+echo "ONEAPI_DEVICE_SELECTOR=${ONEAPI_DEVICE_SELECTOR}"
+
 if [ $GGML_SYCL_DEVICE -ne -1 ]; then
     echo "Use $GGML_SYCL_DEVICE as main GPU"
     #use signle GPU only
     GPUS_SETTING="-mg $GGML_SYCL_DEVICE -sm ${SPLIT_MODE}"
-    echo "ONEAPI_DEVICE_SELECTOR=${ONEAPI_DEVICE_SELECTOR}"
 else
-    echo "Use all Intel GPUs, including iGPU & dGPU"
+    echo "Use Intel GPUs: ${SYCL_DEVICES}"
     GPUS_SETTING="-sm ${SPLIT_MODE}"
  fi
 
-echo "run cmd: ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -no-cnv -p "${INPUT_PROMPT}" -n 200 -e -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE}  --mmap "
-ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -no-cnv -p "${INPUT_PROMPT}" -n 200 -e -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE} --mmap
+echo "run cmd: ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -no-cnv -p "${INPUT_PROMPT}" -n 200 -e -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE} --device ${SYCL_DEVICES} --mmap "
+ZES_ENABLE_SYSMAN=1 ${BIN_FILE} -m ${MODEL_FILE} -no-cnv -p "${INPUT_PROMPT}" -n 200 -e -ngl ${NGL} -s ${SEED} -c ${CONTEXT} ${GPUS_SETTING} -lv ${LOG_VERBOSE} --device ${SYCL_DEVICES} --mmap
 
diff --git a/examples/sycl/win-start-svr.bat b/examples/sycl/win-start-svr.bat
@@ -13,6 +13,7 @@ set "MODEL_FILE=..\models\Qwen3.5-4B-Q4_0.gguf"
 set "NGL=99"
 set "CONTEXT=4096"
 set "GGML_SYCL_DEVICE=-1"
+set "SYCL_DEVICES=SYCL0"
 set "SPLIT_MODE=layer"
 set "LOG_VERBOSE=3"
 
@@ -36,6 +37,21 @@ if /I "%~1"=="--context" (
   goto parse_args
 )
 
+if /I "%~1"=="-d" (
+  if "%~2"=="" goto missing_value
+  set "SYCL_DEVICES=%~2"
+  shift
+  shift
+  goto parse_args
+)
+if /I "%~1"=="--device" (
+  if "%~2"=="" goto missing_value
+  set "SYCL_DEVICES=%~2"
+  shift
+  shift
+  goto parse_args
+)
+
 if /I "%~1"=="-m" (
   if "%~2"=="" goto missing_value
   set "MODEL_FILE=%~2"
@@ -130,6 +146,7 @@ echo This script processes files with specified options.
 echo.
 echo Options:
 echo   -h, --help    Display this help message and exit.
+echo   -d, --device ^<value^>    Set SYCL devices (default: SYCL0).
 echo   -c, --context ^<value^>    Set context length. Bigger need more memory.
 echo   -m, --model   ^<value^>    Full model file path.
 echo   -mg,--main-gpu ^<value^>   Set main GPU ID (0 - n) for single GPU mode.
@@ -160,19 +177,20 @@ REM Support malloc device memory more than 4GB.
 set "UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1"
 echo UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=%UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS%
 
+echo ONEAPI_DEVICE_SELECTOR=%ONEAPI_DEVICE_SELECTOR%
+
 if not "%GGML_SYCL_DEVICE%"=="-1" (
   echo Use %GGML_SYCL_DEVICE% as main GPU
   REM Use single GPU only.
   set "GPUS_SETTING=-mg %GGML_SYCL_DEVICE% -sm %SPLIT_MODE%"
-  echo ONEAPI_DEVICE_SELECTOR=%ONEAPI_DEVICE_SELECTOR%
-) else (
-  echo Use all Intel GPUs, including iGPU ^& dGPU
+  ) else (
+  echo Use Intel GPUs: %SYCL_DEVICES%
   set "GPUS_SETTING=-sm %SPLIT_MODE%"
 )
 
-echo run cmd: ZES_ENABLE_SYSMAN=1 %BIN_FILE% -m "%MODEL_FILE%" -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --mmap --host 0.0.0.0 --port 8000
+echo run cmd: ZES_ENABLE_SYSMAN=1 %BIN_FILE% -m "%MODEL_FILE%" -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --device %SYCL_DEVICES% --mmap --host 0.0.0.0 --port 8000
 set "ZES_ENABLE_SYSMAN=1"
-%BIN_FILE% -m "%MODEL_FILE%" -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --mmap --host 0.0.0.0 --port 8000
+%BIN_FILE% -m "%MODEL_FILE%" -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --device "%SYCL_DEVICES%" --mmap --host 0.0.0.0 --port 8000
 
 endlocal
 
diff --git a/examples/sycl/win-test.bat b/examples/sycl/win-test.bat
@@ -19,6 +19,7 @@ set "MODEL_FILE=..\models\llama-2-7b.Q4_0.gguf"
 set "NGL=99"
 set "CONTEXT=4096"
 set "GGML_SYCL_DEVICE=-1"
+set "SYCL_DEVICES=SYCL0"
 set "SPLIT_MODE=layer"
 set "LOG_VERBOSE=3"
 
@@ -42,6 +43,21 @@ if /I "%~1"=="--context" (
   goto parse_args
 )
 
+if /I "%~1"=="-d" (
+  if "%~2"=="" goto missing_value
+  set "SYCL_DEVICES=%~2"
+  shift
+  shift
+  goto parse_args
+)
+if /I "%~1"=="--device" (
+  if "%~2"=="" goto missing_value
+  set "SYCL_DEVICES=%~2"
+  shift
+  shift
+  goto parse_args
+)
+
 if /I "%~1"=="-p" (
   if "%~2"=="" goto missing_value
   set "INPUT_PROMPT=%~2"
@@ -151,6 +167,7 @@ echo This script processes files with specified options.
 echo.
 echo Options:
 echo   -h, --help    Display this help message and exit.
+echo   -d, --device ^<value^>    Set SYCL devices (default: SYCL0).
 echo   -c, --context ^<value^>    Set context length. Bigger need more memory.
 echo   -p, --promote ^<value^>    Prompt to start generation with.
 echo   -m, --model   ^<value^>    Full model file path.
@@ -182,19 +199,21 @@ REM Support malloc device memory more than 4GB.
 set "UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=1"
 echo UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS=%UR_L0_ENABLE_RELAXED_ALLOCATION_LIMITS%
 
+echo ONEAPI_DEVICE_SELECTOR=%ONEAPI_DEVICE_SELECTOR%
+
 if not "%GGML_SYCL_DEVICE%"=="-1" (
   echo Use %GGML_SYCL_DEVICE% as main GPU
   REM Use single GPU only.
   set "GPUS_SETTING=-mg %GGML_SYCL_DEVICE% -sm %SPLIT_MODE%"
-  echo ONEAPI_DEVICE_SELECTOR=%ONEAPI_DEVICE_SELECTOR%
-) else (
-  echo Use all Intel GPUs, including iGPU ^& dGPU
+  )
+else (
+  echo Use Intel GPUs: %SYCL_DEVICES%
   set "GPUS_SETTING=-sm %SPLIT_MODE%"
 )
 
-echo run cmd: ZES_ENABLE_SYSMAN=1 %BIN_FILE% -m %MODEL_FILE% -no-cnv -p "%INPUT_PROMPT%" -n 200 -e -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --mmap
+echo run cmd: ZES_ENABLE_SYSMAN=1 %BIN_FILE% -m %MODEL_FILE% -no-cnv -p "%INPUT_PROMPT%" -n 200 -e -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --device %SYCL_DEVICES% --mmap
 set "ZES_ENABLE_SYSMAN=1"
-%BIN_FILE% -m "%MODEL_FILE%" -no-cnv -p "%INPUT_PROMPT%" -n 200 -e -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --mmap
+%BIN_FILE% -m "%MODEL_FILE%" -no-cnv -p "%INPUT_PROMPT%" -n 200 -e -ngl %NGL% -s %SEED% -c %CONTEXT% %GPUS_SETTING% -lv %LOG_VERBOSE% --device "%SYCL_DEVICES%" --mmap
 
 endlocal