diff --git a/CMakeLists.txt b/CMakeLists.txt
index 6167dd662..83756c9cb 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -81,7 +81,7 @@ endif(CUDA_STATIC_RUNTIME)
 
 target_link_libraries(rmm INTERFACE rmm::Thrust)
 target_link_libraries(rmm INTERFACE spdlog::spdlog_header_only)
-target_compile_features(rmm INTERFACE cxx_std_14 $<BUILD_INTERFACE:cuda_std_14>)
+target_compile_features(rmm INTERFACE cxx_std_17 $<BUILD_INTERFACE:cuda_std_17>)
 
 # Set logging level. Must go before including gtests and benchmarks.
 
diff --git a/README.md b/README.md
index 8923f7acc..4d9d2261c 100644
--- a/README.md
+++ b/README.md
@@ -28,15 +28,15 @@ RMM can be installed with Conda ([miniconda](https://conda.io/miniconda.html), o
 [Anaconda distribution](https://www.anaconda.com/download)) from the `rapidsai` channel:
 
 ```bash
-# for CUDA 10.2
+# for CUDA 11.2
 conda install -c nvidia -c rapidsai -c conda-forge -c defaults \
-    rmm cudatoolkit=10.2
-# for CUDA 10.1
+    rmm cudatoolkit=11.2
+# for CUDA 11.1
 conda install -c nvidia -c rapidsai -c conda-forge -c defaults \
-    rmm cudatoolkit=10.1
-# for CUDA 10.0
+    rmm cudatoolkit=11.1
+# for CUDA 11.0
 conda install -c nvidia -c rapidsai -c conda-forge -c defaults \
-    rmm cudatoolkit=10.0
+    rmm cudatoolkit=11.0
 ```
 
 We also provide [nightly Conda packages](https://anaconda.org/rapidsai-nightly) built from the HEAD
@@ -44,7 +44,7 @@ of our latest development branch.
 
 Note: RMM is supported only on Linux, and with Python versions 3.7 and later.
 
-Note: The RMM package from Conda requires building with GCC 7 or later. Otherwise, your application may fail to build.
+Note: The RMM package from Conda requires building with GCC 9 or later. Otherwise, your application may fail to build.
 
 See the [Get RAPIDS version picker](https://rapids.ai/start.html) for more OS and version info.
 
@@ -54,14 +54,14 @@ See the [Get RAPIDS version picker](https://rapids.ai/start.html) for more OS an
 
 Compiler requirements:
 
-* `gcc`     version 7.0 or higher required
-* `nvcc`    version 9.0 or higher recommended
-* `cmake`   version 3.18 or higher
+* `gcc`     version 9.3+
+* `nvcc`    version 11.0+
+* `cmake`   version 3.18+
 
 CUDA/GPU requirements:
 
-* CUDA 9.0+
-* NVIDIA driver 396.44+
+* CUDA 11.0+
+* NVIDIA driver 450.51+
 * Pascal architecture or better
 
 You can obtain CUDA from [https://developer.nvidia.com/cuda-downloads](https://developer.nvidia.com/cuda-downloads)
diff --git a/cmake/Modules/SetGPUArchs.cmake b/cmake/Modules/SetGPUArchs.cmake
index c37385a4c..83449488f 100644
--- a/cmake/Modules/SetGPUArchs.cmake
+++ b/cmake/Modules/SetGPUArchs.cmake
@@ -38,16 +38,6 @@ if(NOT DEFINED CUDAToolkit_VERSION AND CMAKE_CUDA_COMPILER)
   unset(NVCC_OUT)
 endif()
 
-if(CUDAToolkit_VERSION_MAJOR LESS 11)
-  list(REMOVE_ITEM SUPPORTED_CUDA_ARCHITECTURES "80")
-endif()
-if(CUDAToolkit_VERSION_MAJOR LESS 10)
-  list(REMOVE_ITEM SUPPORTED_CUDA_ARCHITECTURES "75")
-endif()
-if(CUDAToolkit_VERSION_MAJOR LESS 9)
-  list(REMOVE_ITEM SUPPORTED_CUDA_ARCHITECTURES "70")
-endif()
-
 if(${PROJECT_NAME}_BUILD_FOR_ALL_ARCHS)
   set(CMAKE_CUDA_ARCHITECTURES ${SUPPORTED_CUDA_ARCHITECTURES})