biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Hardware][Powerpc]Enable prefix caching and chunked prefill for ppc64le (#35081)

Browse Source 
Signed-off-by: Akash kaothalkar <akash.kaothalkar@ibm.com>
Co-authored-by: Akash kaothalkar <akash.kaothalkar@ibm.com>
This commit is contained in:
Akash kaothalkar
2026-02-26 15:51:24 +05:30
committed by GitHub
parent 02acd16861
commit e03ddcfbd4
1 changed files with 5 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
vllm/engine/arg_utils.py
View File

@@ -2076,20 +2076,19 @@ class EngineArgs:
)
# Disable chunked prefill and prefix caching for:
# POWER (ppc64le)/RISCV CPUs in V1
# RISCV CPUs in V1
if current_platform.is_cpu() and current_platform.get_cpu_architecture() in (
CpuArchEnum.POWERPC,
CpuArchEnum.RISCV,
):
logger.info(
"Chunked prefill is not supported for POWER, "
"and RISC-V CPUs; "
"Chunked prefill is not supported for"
"RISC-V CPUs; "
"disabling it for V1 backend."
)
self.enable_chunked_prefill = False
logger.info(
"Prefix caching is not supported for POWER, "
"and RISC-V CPUs; "
"Prefix caching is not supported for "
"RISC-V CPUs; "
"disabling it for V1 backend."
)
self.enable_prefix_caching = False