diff --git a/backend/cpp/ds4/CMakeLists.txt b/backend/cpp/ds4/CMakeLists.txt
index afc3dd43f9cf..10d999729d80 100644
--- a/backend/cpp/ds4/CMakeLists.txt
+++ b/backend/cpp/ds4/CMakeLists.txt
@@ -76,12 +76,13 @@ elseif(DS4_GPU STREQUAL "cpu")
     set(DS4_OBJS "${DS4_DIR}/ds4_cpu.o")
 endif()
 
-# ds4.c now references ds4_distributed.c (distributed inference) and ds4_ssd.c
-# (SSD expert-cache), each split into its own translation unit upstream. Both
-# are GPU-agnostic objects shared by every GPU mode, so link them in regardless
-# of DS4_GPU.
+# Upstream splits distributed inference, tensor-parallel transport, the SSD
+# expert cache, and layer placement into GPU-agnostic translation units. Link
+# them regardless of DS4_GPU.
 list(APPEND DS4_OBJS "${DS4_DIR}/ds4_distributed.o")
+list(APPEND DS4_OBJS "${DS4_DIR}/ds4_tp.o")
 list(APPEND DS4_OBJS "${DS4_DIR}/ds4_ssd.o")
+list(APPEND DS4_OBJS "${DS4_DIR}/ds4_layer_pack.o")
 
 add_executable(${TARGET}
     grpc-server.cpp
diff --git a/backend/cpp/ds4/Makefile b/backend/cpp/ds4/Makefile
index a739a7d5659a..cfc0b151bc25 100644
--- a/backend/cpp/ds4/Makefile
+++ b/backend/cpp/ds4/Makefile
@@ -1,10 +1,10 @@
 # ds4 backend Makefile.
 #
-# Upstream pin lives below as DS4_VERSION?=80ebbc396aee40eedc1d829222f3362d10fa4c6c
+# Upstream pin lives below as DS4_VERSION?=efdadd41e20134af4f3381e1ed90e96fe4faef6f
 # (.github/bump_deps.sh) can find and update it - matches the
 # llama-cpp / ik-llama-cpp / turboquant convention.
 
-DS4_VERSION?=80ebbc396aee40eedc1d829222f3362d10fa4c6c
+DS4_VERSION?=efdadd41e20134af4f3381e1ed90e96fe4faef6f
 DS4_REPO?=https://github.com/antirez/ds4
 
 CURRENT_MAKEFILE_DIR := $(dir $(abspath $(lastword $(MAKEFILE_LIST))))
@@ -18,20 +18,19 @@ UNAME_S := $(shell uname -s)
 
 CMAKE_ARGS ?= -DCMAKE_BUILD_TYPE=Release
 
-# ds4_distributed.o and ds4_ssd.o are GPU-agnostic translation units that
-# ds4.c/ds4_cpu.o now reference (upstream split distributed inference and the
-# SSD expert-cache into their own .c files). Both objects are shared by every
-# GPU mode, so they are appended unconditionally below.
+# Upstream splits distributed inference, tensor-parallel transport, the SSD
+# expert cache, and layer placement into GPU-agnostic translation units. They
+# are shared by every GPU mode, so append them unconditionally below.
 ifeq ($(BUILD_TYPE),cublas)
     CMAKE_ARGS += -DDS4_GPU=cuda
-    DS4_OBJ_TARGET := ds4.o ds4_cuda.o ds4_distributed.o ds4_ssd.o
+    DS4_OBJ_TARGET := ds4.o ds4_cuda.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 else ifeq ($(UNAME_S),Darwin)
     CMAKE_ARGS += -DDS4_GPU=metal
-    DS4_OBJ_TARGET := ds4.o ds4_metal.o ds4_distributed.o ds4_ssd.o
+    DS4_OBJ_TARGET := ds4.o ds4_metal.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 else
     # CPU reference path (Linux only - macOS CPU path is broken by VM bug per ds4 README).
     CMAKE_ARGS += -DDS4_GPU=cpu
-    DS4_OBJ_TARGET := ds4_cpu.o ds4_distributed.o ds4_ssd.o
+    DS4_OBJ_TARGET := ds4_cpu.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 endif
 
 ifneq ($(NATIVE),true)
@@ -56,11 +55,11 @@ ds4:
 # the right per-platform compile flags (Objective-C/Metal on Darwin, nvcc on Linux+CUDA).
 ds4/ds4.o: ds4
 ifeq ($(BUILD_TYPE),cublas)
-	+$(MAKE) -C ds4 ds4.o ds4_cuda.o ds4_distributed.o ds4_ssd.o
+	+$(MAKE) -C ds4 ds4.o ds4_cuda.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 else ifeq ($(UNAME_S),Darwin)
-	+$(MAKE) -C ds4 ds4.o ds4_metal.o ds4_distributed.o ds4_ssd.o
+	+$(MAKE) -C ds4 ds4.o ds4_metal.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 else
-	+$(MAKE) -C ds4 ds4_cpu.o ds4_distributed.o ds4_ssd.o
+	+$(MAKE) -C ds4 ds4_cpu.o ds4_distributed.o ds4_tp.o ds4_ssd.o ds4_layer_pack.o
 endif
 
 grpc-server: ds4/ds4.o