From f566fc5b96b4fbddc72156d167f59b4c4ee39f93 Mon Sep 17 00:00:00 2001 From: Tiffany Chen Date: Wed, 1 Oct 2025 17:13:47 -0700 Subject: [PATCH 1/5] Update CPU requirements. --- docs/source/how-to/cloudxr_teleoperation.rst | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/docs/source/how-to/cloudxr_teleoperation.rst b/docs/source/how-to/cloudxr_teleoperation.rst index 6ac702575d9..55bbc3e2779 100644 --- a/docs/source/how-to/cloudxr_teleoperation.rst +++ b/docs/source/how-to/cloudxr_teleoperation.rst @@ -66,9 +66,7 @@ Prior to using CloudXR with Isaac Lab, please review the following system requir * `Docker`_ 26.0.0+, `Docker Compose`_ 2.25.0+, and the `NVIDIA Container Toolkit`_. Refer to the Isaac Lab :ref:`deployment-docker` for how to install. * For details on driver requirements, please see the `Technical Requirements `_ guide - * CPU - * Intel: 16-core Intel Core i9, X-series or higher - * AMD: Ryzen 9 5900XT, Ryzen Threadripper PRO 5955WX or higher + * CPU: 16-Cores AMD Ryzen Threadripper Pro 5955WX or higher * Memory: 64GB RAM * GPU * Minimum: 1x RTX PRO 6000 GPUs (or equivalent e.g. 1x RTX 5090) From c3edaf443155acdbe94f7da9c830402db93cd67b Mon Sep 17 00:00:00 2001 From: Tiffany Chen Date: Mon, 6 Oct 2025 15:34:03 -0700 Subject: [PATCH 2/5] Update known issues. --- docs/source/how-to/cloudxr_teleoperation.rst | 29 +++++++++++++------ .../imitation-learning/teleop_imitation.rst | 2 +- 2 files changed, 21 insertions(+), 10 deletions(-) diff --git a/docs/source/how-to/cloudxr_teleoperation.rst b/docs/source/how-to/cloudxr_teleoperation.rst index 55bbc3e2779..98e7cb54056 100644 --- a/docs/source/how-to/cloudxr_teleoperation.rst +++ b/docs/source/how-to/cloudxr_teleoperation.rst @@ -63,14 +63,13 @@ Prior to using CloudXR with Isaac Lab, please review the following system requir * Isaac Lab workstation * Ubuntu 22.04 or Ubuntu 24.04 + * Minimum hardware requirements to sustain 45 FPS with a 120Hz physics simulation: + * CPU: 16-Cores AMD Ryzen Threadripper Pro 5955WX or higher + * Memory: 64GB RAM + * GPU: 1x RTX PRO 6000 GPUs (or equivalent e.g. 1x RTX 5090) or higher + * For details on driver requirements, please see the `Technical Requirements `_ guide * `Docker`_ 26.0.0+, `Docker Compose`_ 2.25.0+, and the `NVIDIA Container Toolkit`_. Refer to the Isaac Lab :ref:`deployment-docker` for how to install. - * For details on driver requirements, please see the `Technical Requirements `_ guide - * CPU: 16-Cores AMD Ryzen Threadripper Pro 5955WX or higher - * Memory: 64GB RAM - * GPU - * Minimum: 1x RTX PRO 6000 GPUs (or equivalent e.g. 1x RTX 5090) - * Recommended: 2x RTX PRO 6000 GPUs (or equivalent e.g. 2x RTX 5090) * Apple Vision Pro @@ -280,11 +279,11 @@ On your Mac: +-------------------+---------------------+ | Isaac Lab Version | Client App Version | +-------------------+---------------------+ - | 2.3.x | v2.3.0 | + | 2.3 | v2.3.0 | +-------------------+---------------------+ - | 2.2.x | v2.2.0 | + | 2.2 | v2.2.0 | +-------------------+---------------------+ - | 2.1.x | v1.0.0 | + | 2.1 | v1.0.0 | +-------------------+---------------------+ .. code-block:: bash @@ -986,6 +985,11 @@ Known Issues This error message can be safely ignored. It is caused by a race condition in the exit handler for AR Mode. +* ``XR_ERROR_INSTANCE_LOST in xrPollEvent: Call to "xrt_session_poll_events" failed`` + + This error may occur if the CloudXR runtime exits before Isaac Lab. Restarting the CloudXR + runtime to resume teleoperation. + * ``[omni.usd] TF_PYTHON_EXCEPTION`` when starting/stopping AR Mode This error message can be safely ignored. It is caused by a race condition in the enter/exit @@ -996,6 +1000,13 @@ Known Issues This error message can be caused by shader assets authored with older versions of USD, and can typically be ignored. +* The XR device connects successfully, but no video is displayed, even though the Isaac Lab viewport responds to tracking. + + This error occurs when the GPU index differs between the host and the container, causing CUDA + to load on the wrong GPU. To fix this, set `NV_GPU_INDEX` in the runtime container to `0`, `1`, + or `2` to ensure the GPU selected by CUDA matches the host. + + Kubernetes Deployment --------------------- diff --git a/docs/source/overview/imitation-learning/teleop_imitation.rst b/docs/source/overview/imitation-learning/teleop_imitation.rst index 00037beedb4..4b691318a45 100644 --- a/docs/source/overview/imitation-learning/teleop_imitation.rst +++ b/docs/source/overview/imitation-learning/teleop_imitation.rst @@ -61,7 +61,7 @@ For tasks that benefit from the use of an extended reality (XR) device with hand .. note:: - See :ref:`cloudxr-teleoperation` to learn more about using CloudXR with Isaac Lab. + See :ref:`cloudxr-teleoperation` to learn how to use CloudXR and experience teleoperation with Isaac Lab. The script prints the teleoperation events configured. For keyboard, From 70255e4cb62a0896a4601ee22ec2559eac5d680a Mon Sep 17 00:00:00 2001 From: Tiffany Chen Date: Mon, 6 Oct 2025 16:40:52 -0700 Subject: [PATCH 3/5] Format fix. --- docs/source/how-to/cloudxr_teleoperation.rst | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/source/how-to/cloudxr_teleoperation.rst b/docs/source/how-to/cloudxr_teleoperation.rst index 98e7cb54056..71824b53023 100644 --- a/docs/source/how-to/cloudxr_teleoperation.rst +++ b/docs/source/how-to/cloudxr_teleoperation.rst @@ -985,10 +985,10 @@ Known Issues This error message can be safely ignored. It is caused by a race condition in the exit handler for AR Mode. -* ``XR_ERROR_INSTANCE_LOST in xrPollEvent: Call to "xrt_session_poll_events" failed`` +* ``XR_ERROR_INSTANCE_LOST in xrPollEvent: Call to "xrt_session_poll_events" failed`` - This error may occur if the CloudXR runtime exits before Isaac Lab. Restarting the CloudXR - runtime to resume teleoperation. + This error may occur if the CloudXR runtime exits before Isaac Lab. Restarting the CloudXR + runtime to resume teleoperation. * ``[omni.usd] TF_PYTHON_EXCEPTION`` when starting/stopping AR Mode @@ -1006,7 +1006,7 @@ Known Issues to load on the wrong GPU. To fix this, set `NV_GPU_INDEX` in the runtime container to `0`, `1`, or `2` to ensure the GPU selected by CUDA matches the host. - + Kubernetes Deployment --------------------- From e2e0e4a7ccdf87c3f4e1b79bef93b94d2c42415d Mon Sep 17 00:00:00 2001 From: Tiffany Chen Date: Mon, 6 Oct 2025 16:45:58 -0700 Subject: [PATCH 4/5] Lint fix. --- docs/source/how-to/cloudxr_teleoperation.rst | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/source/how-to/cloudxr_teleoperation.rst b/docs/source/how-to/cloudxr_teleoperation.rst index 71824b53023..87122993b7d 100644 --- a/docs/source/how-to/cloudxr_teleoperation.rst +++ b/docs/source/how-to/cloudxr_teleoperation.rst @@ -1003,8 +1003,8 @@ Known Issues * The XR device connects successfully, but no video is displayed, even though the Isaac Lab viewport responds to tracking. This error occurs when the GPU index differs between the host and the container, causing CUDA - to load on the wrong GPU. To fix this, set `NV_GPU_INDEX` in the runtime container to `0`, `1`, - or `2` to ensure the GPU selected by CUDA matches the host. + to load on the wrong GPU. To fix this, set ``NV_GPU_INDEX`` in the runtime container to ``0``, ``1``, + or ``2`` to ensure the GPU selected by CUDA matches the host. Kubernetes Deployment From 51db547fb0cd330ac0c7e08028fdfbd2386d1bf6 Mon Sep 17 00:00:00 2001 From: Tiffany Chen Date: Tue, 7 Oct 2025 13:41:49 -0700 Subject: [PATCH 5/5] Minor fixes. --- docs/source/how-to/cloudxr_teleoperation.rst | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/source/how-to/cloudxr_teleoperation.rst b/docs/source/how-to/cloudxr_teleoperation.rst index 87122993b7d..3b532853ef0 100644 --- a/docs/source/how-to/cloudxr_teleoperation.rst +++ b/docs/source/how-to/cloudxr_teleoperation.rst @@ -63,7 +63,7 @@ Prior to using CloudXR with Isaac Lab, please review the following system requir * Isaac Lab workstation * Ubuntu 22.04 or Ubuntu 24.04 - * Minimum hardware requirements to sustain 45 FPS with a 120Hz physics simulation: + * Hardware requirements to sustain 45 FPS with a 120Hz physics simulation: * CPU: 16-Cores AMD Ryzen Threadripper Pro 5955WX or higher * Memory: 64GB RAM * GPU: 1x RTX PRO 6000 GPUs (or equivalent e.g. 1x RTX 5090) or higher @@ -987,7 +987,7 @@ Known Issues * ``XR_ERROR_INSTANCE_LOST in xrPollEvent: Call to "xrt_session_poll_events" failed`` - This error may occur if the CloudXR runtime exits before Isaac Lab. Restarting the CloudXR + This error may occur if the CloudXR runtime exits before Isaac Lab. Restart the CloudXR runtime to resume teleoperation. * ``[omni.usd] TF_PYTHON_EXCEPTION`` when starting/stopping AR Mode