Daily Digest
Daily brief
Loading brief…
Issues & signals
- Cannot start IsaacSim 5.1.0 because Ros2 Bridgenvidia-forum:simulation5/11/2026▲ 2deploymentv5.1.0deploymentisaac-sim
User cannot start IsaacSim 5.1.0 because of the ROS2 Bridge. This prevents launching the simulator with ROS integration enabled.
⚡ ROS 2 bridge startup failures block the primary integration path between simulation and real robot stacks.
- Can't install ros-jazzy-isaac-ros-foundationpose in Isaac ROS environmentnvidia-forum:isaac5/15/2026docs-onboardingisaacisaac-rosros2-jazzyfoundationposeinstallationpackaging
Same install failure is reposted in another NVIDIA forum category: `ros-jazzy-isaac-ros-foundationpose` cannot be installed in the Isaac ROS environment. This blocks use of FoundationPose.
⚡ Duplicated reports across NVIDIA channels suggest friction in Isaac ROS packaging that can spill over into Isaac Sim/Isaac platform perception narratives.
- Can't install ros-jazzy-isaac-ros-foundationpose in Isaac ROS environmentnvidia-forum:isaac-ros5/15/2026docs-onboardingisaac-rosros2-jazzyfoundationposeinstallationpackaginglinux
User reports they cannot install the ROS2 Jazzy package `ros-jazzy-isaac-ros-foundationpose` in an Isaac ROS environment. Without the package, they can’t proceed with FoundationPose workflows.
⚡ Installability of key Isaac ROS packages is a gating factor for adoption of NVIDIA’s perception stack in real ROS2 deployments.
- Isaac Sim 5.1.0 crashes shortly after startup on Windows Server 2025 with RTX Pro 6000 Blackwellnvidia-forum:simulation5/13/2026crashes-stabilityv5.1.0isaac-sim5-1-0windows-server-2025rtx-pro-6000blackwellstartup-crash
Report that Isaac Sim 5.1.0 crashes shortly after startup on Windows Server 2025 with an RTX Pro 6000 Blackwell GPU. User cannot run the simulator in this environment.
⚡ Stability on enterprise Windows Server + latest RTX workstation GPUs is essential for lab/IT deployments and can become a reputational risk if early crashes persist.
- Jetson AGX Orin fails to power on ISP1 at bootnvidia-forum:robotics-edge-computing5/13/2026hardware-integrationjetsonagx-orinispbootcamera
A Jetson AGX Orin report indicates ISP1 fails to power on at boot. This prevents expected camera/ISP functionality from being available after startup.
⚡ Edge platform reliability issues at boot-time peripherals can block sensor bring-up and derail deployment schedules.
- Two cameras on the jetson orin nano are not workingnvidia-forum:robotics-edge-computing5/13/2026hardware-integrationjetsonorin-nanocamerasbringupsensors
A user reports two cameras on Jetson Orin Nano are not working. No further diagnostic information is included.
⚡ Multi-camera support on Orin Nano is a common robotics requirement and failures directly limit real deployments.
- Jetson AGX Orin Developer Kit completely dead / no power responsenvidia-forum:robotics-edge-computing5/13/2026hardware-integrationjetsonagx-orinpowerbootdevkit
A Jetson AGX Orin Developer Kit is reported completely dead with no power response. No additional context is provided.
⚡ Perceived hardware “dead on arrival” events create outsized friction for developer adoption and field confidence.
- [BUG] dexterous robot hand failed to grasp and lift object in newtongithub:newton-physics/newton5/13/2026crashes-stabilitycrashusdrenderingmanipulationisaac-labnewton
A dexterous hand imported via URDF fails to grasp and lift a bottle; the object slides and remains unliftable. The same bottle can be lifted using a Franka example, suggesting contact/friction or grasp modeling differences for the hand.
⚡ Robust grasp friction/contact modeling across different hands is necessary to make Newton credible for dexterous manipulation workflows.
- [BUG] dexterous robot hand failed to grasp object in newtongithub:newton-physics/newton5/13/2026crashes-stabilitycrashusdrenderinghardwaremanipulationisaac-lab
A dexterous hand imported via URDF cannot grasp a bottle reliably; the bottle slides and cannot be lifted. The reporter notes the Franka example can lift the same object, implying a hand-specific contact/friction issue.
⚡ Failures in basic grasp stability erode developer confidence and can push users to alternative simulators for manipulation work.
- Jetson Orin NX Super 16GB not powering on after reverse polarity - D65/Q25 suspected (P3768-A04)nvidia-forum:robotics-edge-computing5/13/2026hardware-integrationjetsonorin-nxpowerhardware-damagebringup
Jetson Orin NX Super 16GB reportedly does not power on after reverse polarity, with suspected component damage (D65/Q25). This blocks device operation.
⚡ Electrical robustness incidents can turn into costly RMAs and reduce confidence in edge platforms for robotics.
- New sensor capture image failed with jetson linux 36.3.0nvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetson-linux36.3.0cameraimage-captureregression
A new sensor capture image reportedly fails with Jetson Linux 36.3.0. This indicates a capture pipeline problem on that release.
⚡ Sensor capture regressions tied to Jetson Linux versions can stall perception development and hardware rollouts.
- Jetson orin nano booting issuenvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonorin-nanobootstabilitybringup
A user reports a Jetson Orin Nano booting issue. No further details are included.
⚡ Boot instability on Orin Nano directly impacts developer productivity and device suitability for robotics prototypes.
- Orin Nano Super Dev Kit - MSS SDRAM init failure (err 0x48480112) - module previously functionalnvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonorin-nanosdrambootinit-failure
Orin Nano Super Dev Kit shows an MSS SDRAM init failure (err 0x48480112) though the module was previously functional. This prevents the system from booting normally.
⚡ Memory init failures on previously functional devices can indicate fragile update paths and cause field downtime.
- Jetson ORIN not coming out of recoverynvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonorinrecovery-modeflashingboot
Jetson ORIN is reported as not coming out of recovery. This blocks normal boot and device provisioning.
⚡ Recovery-mode issues prevent flashing and servicing, creating a hard stop for both developers and production support.
- AGX Thor / JetPack 7.1 - DCE_MBOX_TO_DCE_RM left dirty by UEFI causing RmInitAdapter failure on specific unitsnvidia-forum:robotics-edge-computing5/12/2026crashes-stabilityjetsonthorjetpack-7-1uefirminitadaptergpu-init
On AGX Thor / JetPack 7.1, UEFI leaves DCE mailbox state dirty, causing RmInitAdapter failure on specific units. This can prevent GPU driver initialization and system usability.
⚡ Firmware/UEFI-induced GPU init failures on Thor threaten platform reliability for high-end edge compute deployments.
- Orin nx 16G boot failednvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonorin-nxboot-failurebringupreliability
User reports Orin NX 16G boot failed. This prevents use of the module/device.
⚡ Recurring boot failures across Orin NX variants highlight ongoing platform reliability and supportability risks.
- JetsonOrinNano dev kit first boot issuenvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonorin-nanodevkitfirst-bootsetup
User reports a first boot issue with Jetson Orin Nano dev kit. This blocks initial setup.
⚡ First-boot issues are high-friction moments that can drive early abandonment of Jetson dev kits.
- Xavier NX new DRAM modules won’t flash/boot with L4T 32.7.1 — request PCN overlay / recommended pathnvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonxavier-nxdraml4t-32-7-1flashboot
Xavier NX with new DRAM modules won’t flash/boot with L4T 32.7.1; user requests PCN overlay or recommended path. This blocks bringing up new modules on existing software.
⚡ Module BOM changes (DRAM) that break flashing/boot without clear overlays create major field-upgrade and OEM risk.
- Having a hard time booting from Thor IGX Mininvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonthorigx-minibootstorage-boot
User reports difficulty booting from Thor IGX Mini. This blocks platform setup and testing.
⚡ Boot-from-device friction on Thor IGX Mini can slow evaluation of Thor for industrial edge deployments.
- Isaac Sim Crashnvidia-forum:robotics-edge-computing5/11/2026crashes-stabilityisaac-simcrashstabilitysimulation
User reports an Isaac Sim crash. No additional details are provided, but the issue blocks use of the simulator.
⚡ Repeated generic crash reports suggest unresolved stability issues impacting Isaac Sim adoption.
- Jetson 4.6.0 Xavier (Micron 16Gbit) NX board MB1 boot issuenvidia-forum:robotics-edge-computing5/11/2026hardware-integrationjetsonxavier-nxjetson-4-6-0mb1bootmicron-dram
Jetson 4.6.0 Xavier NX board with Micron 16Gbit memory hits an MB1 boot issue. This prevents boot and blocks device use.
⚡ MB1/bootloader bring-up issues on Xavier NX continue to affect long-lived products in the field and supply-chain refreshes.
- Nv-tegra.nvidia.com down?nvidia-forum:robotics-edge-computing5/11/2026deploymentjetsonnv-tegraoutagedownloadsinfrastructure
User reports nv-tegra.nvidia.com may be down. If true, this blocks access to resources needed for Jetson development and deployment.
⚡ Availability of key distribution endpoints is part of platform reliability; outages can halt builds, flashes, and updates globally.
- Add benchmark to detect CPU regressiongithub:newton-physics/newton5/10/2026other
Request to add a benchmark to detect CPU performance regressions, motivated by a prior issue. This would improve detection of slowdowns in the CPU code path.
⚡ Built-in CPU regression benchmarks help keep simulation/training performance predictable across releases and hardware.
- [Proposal] Standardized Imitation Learning Pipeline for ./isaaclab.sh --newgithub:isaac-sim/IsaacLab5/9/2026synthetic-datav3.0synthetic-datarldeploymentdocsintegrationfeature-request
Proposal This proposal requests the addition of a standardized Imitation Learning (IL) pipeline template to the ./isaaclab.sh --new environment creation CLI. Currently, when users create a new project, they are not presented with an option to scaffold a projec
- [BUG] VBD Rigid Body friction regressiongithub:newton-physics/newton5/14/2026crashes-stabilitynewton-physicsphysicsfrictionregressionvbd-solvercontacts
Newton reports a VBD solver change that causes rigid bodies to stop at much higher velocities than expected, indicating a friction regression. They loosened test tolerances as a stopgap but still need to investigate and fix the solver behavior.
⚡ Physics regression risk in third-party engines underscores the need for robust validation baselines when positioning NVIDIA simulation for contact-rich tasks.
- Robot Wizard Add Colliders Buutton ont show in the windownvidia-forum:simulation5/14/2026tooling-dxisaac-simrobot-wizardcollidersuiusdauthoring
User reports the Robot Wizard “Add Colliders” button does not show in the window. This likely blocks or slows collider generation during robot import/setup.
⚡ Missing or inconsistent UI affordances in robot setup tools can slow asset bring-up and undermine the ‘easy onboarding’ story for simulation users.
- Repeated `module="unique"` kernel declarations retain discarded temporary modulesgithub:NVIDIA/warp5/13/2026crashes-stabilitywarpmemory-leakkernelspythoncompiler-cachetraining-infra
Warp may retain discarded temporary modules when repeatedly declaring equivalent `@wp.kernel(module="unique")` kernels. This can keep Python object graphs reachable and grow memory over repeated factory calls.
⚡ Long-running training/simulation workloads depend on predictable memory behavior; leaks in kernel compilation/caching paths can become a hard scaling limit.
- How do I run IsaacSim via python script with ros2 bridge enablednvidia-forum:simulation5/13/2026integrationisaac-simpythonros2-bridgeautomationheadlessintegration
User asks how to run Isaac Sim via a Python script with the ROS2 bridge enabled. This implies friction in documentation or APIs for programmatic launch/configuration of ROS2 integration.
⚡ Scriptable headless + ROS2-bridge enablement is critical for automation and scalable robotics workflows built around Isaac Sim.
- Content Browser UX Issuesgithub:isaac-sim/IsaacSim5/13/2026tooling-dxv5.1.0isaac-simcontent-browserasset-browseruxassetssearch
Isaac Sim Content Browser (planned to replace Asset Browser in Isaac Sim 6) lacks key UX features compared to Asset Browser in 5.1.0: no thumbnail preview, incomplete search (excludes non-SimReady assets), and inability to do some global searches within catego
⚡ If the new Content Browser degrades core asset discovery/preview workflows, Isaac Sim 6 adoption will be slowed by daily iteration friction for builders.
- Cannot download/install PyTorch with CUDA support on Jetson Orin NX (JetPack 6.0) due to Network issuesnvidia-forum:robotics-edge-computing5/13/2026deploymentjetsonorin-nxjetpack-6pytorchcudanetwork
A user cannot download/install PyTorch with CUDA support on Jetson Orin NX (JetPack 6.0) due to network issues. This blocks setting up a CUDA-enabled ML stack on device.
⚡ Reliable package distribution on Jetson is essential because blocked installs prevent teams from running CUDA-accelerated workloads.
- The USB microphone cannot record properly on the Jetson Thor platformnvidia-forum:robotics-edge-computing5/13/2026hardware-integrationjetsonthorusb-audiomicrophonemultimodal
A USB microphone reportedly cannot record properly on the Jetson Thor platform. No additional details are provided in the post.
⚡ Audio I/O reliability on new platforms affects teleop and multimodal robotics use cases beyond vision-only stacks.
- Some random PCIe signals between SBC and NVMenvidia-forum:robotics-edge-computing5/13/2026hardware-integrationpcienvmesignal-integrityjetsoncarrier-board
A user reports random PCIe signals between an SBC and an NVMe device. The post contains no additional diagnostic content.
⚡ Signal integrity and PCIe stability are recurring edge-system risks that can undermine storage and logging pipelines in robots.
- Tegraparser_v2 problemnvidia-forum:robotics-edge-computing5/13/2026tooling-dxtegraparserjetson-linuxflashingbringuptooling
A user reports a Tegraparser_v2 problem. No further details are included in the issue body.
⚡ Low-level flashing and parsing tool failures create high support burden and slow down board bring-up cycles.
- Nova Orin Init for Nova Carter Robotnvidia-forum:robotics-edge-computing5/13/2026docs-onboardingnovaorincarterinitializationonboarding
A user asks about Nova Orin initialization for the Nova Carter Robot. The post contains no additional details.
⚡ Robust, centralized onboarding for Nova Orin/Carter reduces duplicated support across forums and accelerates deployments.
- When using a GMSL camera, image acquisition may fail midwaynvidia-forum:robotics-edge-computing5/13/2026hardware-integrationgmslcameraimage-capturejetsonstability
When using a GMSL camera, image acquisition may fail midway. No additional details are provided.
⚡ GMSL capture reliability is foundational for autonomous robots; mid-run acquisition failures can invalidate datasets and runs.
- CAN0 data bitrate issuenvidia-forum:robotics-edge-computing5/12/2026hardware-integrationcanbuscan0bitratejetsonrobotics-io
A user reports a CAN0 data bitrate issue. No additional details are provided in the post.
⚡ CAN bitrate configuration issues can stall robotics integration because CAN is a common actuator/sensor bus.
- AGX T5000 eeprom报错nvidia-forum:robotics-edge-computing5/12/2026hardware-integrationeepromagxprovisioningjetsonhardware
A user reports an EEPROM error on AGX T5000. No additional context is provided.
⚡ EEPROM errors can block provisioning and fleet management workflows for embedded robotics hardware.
- Support hardware-coherent CPU/GPU array launchesgithub:NVIDIA/warp5/12/2026hardware-integrationhardwaredocswarp
Warp plans work to support GPU kernel launches with hardware-coherent CPU memory, pinned CPU arrays, and peer GPU arrays when directly addressable. The issue emphasizes preserving clear diagnostics for invalid cases.
⚡ Zero-copy or direct-access mixed-memory launches are a strategic lever for tighter CPU-GPU co-processing in robotics pipelines.
- MultiMeshRayCaster data (pos_w, ray_hits_w) fails to update after dynamic teleportation in _apply_actionnvidia-forum:isaac5/12/2026sensors-perceptionisaacraycastingmultimesh-raycasterteleportationsimulation
MultiMeshRayCaster outputs (pos_w, ray_hits_w) reportedly fail to update after dynamic teleportation in _apply_action. The report appears in the Isaac forum with no extra details.
⚡ Common sensor primitives failing under resets/teleports undermine confidence in simulation-driven autonomy development.
- MultiMeshRayCaster data (pos_w, ray_hits_w) fails to update after dynamic teleportation in _apply_actionnvidia-forum:isaac-ros5/12/2026sensors-perceptionisaac-rosraycastingmultimesh-raycasterteleportationperception
MultiMeshRayCaster outputs (pos_w, ray_hits_w) reportedly fail to update after dynamic teleportation in _apply_action. The report appears in the Isaac ROS forum with no extra details.
⚡ Perception stack users expect deterministic, up-to-date raycast signals for navigation and obstacle avoidance behaviors.
- MultiMeshRayCaster data (pos_w, ray_hits_w) fails to update after dynamic teleportation in _apply_actionnvidia-forum:robotics-edge-computing5/12/2026sensors-perceptionraycastingmultimesh-raycasterteleportationsensorsrobotics
MultiMeshRayCaster outputs (pos_w, ray_hits_w) reportedly fail to update after dynamic teleportation in _apply_action. The report appears in robotics-edge-computing forum with no extra details.
⚡ Sensor update correctness bugs propagate across robotics stacks when the same component is used from edge to sim.
- MultiMeshRayCaster data (pos_w, ray_hits_w) fails to update after dynamic teleportation in _apply_actionnvidia-forum:simulation5/12/2026other
MultiMeshRayCaster outputs (pos_w, ray_hits_w) reportedly fail to update after dynamic teleportation in _apply_action. This suggests stale raycast data after state changes.
⚡ Correct sensor data updates under teleportation/reset are essential for RL environment correctness and reproducibility.
- Orin NX 16GB – Request for DRAM Supplier Consistency (Hynix)nvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonorin-nxdramsupply-chainfleet
A user requests DRAM supplier consistency (Hynix) for Orin NX 16GB. This indicates concerns about BOM variability impacting deployments.
⚡ Component supply variability (e.g., DRAM) can create hard-to-debug performance/compatibility differences across fleets.
- ISSAC SIM Crash GUI errornvidia-forum:simulation5/12/2026crashes-stabilitycrash
A user reports an Isaac Sim crash with a GUI error. No further detail is provided in the post.
⚡ GUI stability remains a major adoption gate for simulation users who rely on interactive workflows.
- [BUG] Error issued when adding a D6 joint with 1 angular dof and 1 linear dof to SolverMuJoCogithub:newton-physics/newton5/12/2026othernewton
Adding a D6 joint with 1 angular DOF and 1 linear DOF to SolverMuJoCo can produce a repeated joint name error because the name-amending logic doesn't change names in this case. This causes model build failure due to name collisions.
⚡ Robust joint authoring and naming rules are essential for importing real robots without brittle solver-specific failures.
- Nightly failure (2026-05-12): Warp nightly testsgithub:newton-physics/newton5/12/2026othernewtonwarp
Newton's scheduled nightly workflow failed specifically in the Warp nightly tests suite while other suites passed. The issue references the failing GitHub Actions runs and logs.
⚡ Nightly integration failures between core components (e.g., Warp) are early warning signs for release risk and ecosystem drift.
- [Bug Report] Relative texture path does not work in IsaacLab Beta version.github:isaac-sim/IsaacLab5/12/2026asset-pipelinev2.3.2usdrenderinghardwareisaac-simisaac-lab
Relative texture paths do not work in the IsaacLab Beta according to the report, even when the image is in the same folder as the USD. Loading via IsaacLab code triggers errors.
⚡ Asset portability issues (relative paths) slow dataset/scene sharing and make Isaac Lab workflows brittle across machines.
- Mmapi encoder dequeue capture not blockednvidia-forum:robotics-edge-computing5/12/2026deploymentjetsonmultimediammapiencodervideo-pipeline
User reports that mmapi encoder dequeue capture is not blocked. This suggests unexpected non-blocking behavior in the encoder pipeline.
⚡ Media pipeline blocking/latency issues can break real-time perception and streaming workloads on Jetson.
- Usb3.0 device recognize as usb2.0 devicenvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonusb3usb2link-speedperipherals
A USB 3.0 device is recognized as a USB 2.0 device. This reduces bandwidth and can break high-throughput peripherals.
⚡ USB link-speed downgrade issues can cripple sensor/peripheral throughput and degrade end-to-end system performance.
- [DOCS] Solver Documentation Soft Bodiesgithub:newton-physics/newton5/12/2026docs-onboardingdocsnewton
User reports that the solver documentation page does not show soft body support for VBD and asks whether VBD supports soft bodies. This suggests docs may be incorrect or incomplete.
⚡ Documentation ambiguity about solver capabilities creates uncertainty for users choosing physics backends for soft-body tasks.
- Public sources folder is emptynvidia-forum:robotics-edge-computing5/12/2026docs-onboardingjetsonsourcesdistributiondocumentationcompliance
User reports that a 'Public sources' folder is empty. This blocks access to expected source materials.
⚡ Missing public source drops can block compliance and debugging workflows, impacting developer trust.
- Jetson Orin Nano Super – RTSP Stream Lag, High RAM Usage, and Slow Face Detectionnvidia-forum:robotics-edge-computing5/12/2026deploymentjetsonorin-nano-superrtspmemory-usageperformanceface-detection
Jetson Orin Nano Super shows RTSP stream lag, high RAM usage, and slow face detection. This suggests performance and memory efficiency issues in a common edge AI workload.
⚡ Real-time streaming and on-device vision performance issues can undermine Jetson’s value proposition for edge AI demos and pilots.
- Isaac Sim Crashnvidia-forum:isaac5/11/2026crashes-stabilitycrashisaac-sim
User reports Isaac Sim crash. This indicates instability affecting simulator usage.
⚡ Frequent crash reports in core Isaac channels reduce confidence in using Isaac Sim as the default training and validation environment.
- Isaac Sim Crashnvidia-forum:isaac-ros5/11/2026crashes-stabilitycrashisaac-sim
Same 'Isaac Sim Crash' report appears in Isaac ROS forum. This suggests downstream robotics workflows may be blocked by simulator instability.
⚡ Cross-posted crash issues indicate end-to-end stack fragility that can affect Isaac ROS user success.
- Why can't control gpio in jetapck6.2.1?nvidia-forum:robotics-edge-computing5/11/2026hardware-integrationjetsonjetpack-6-2-1gpiodrivershardware-io
User cannot control GPIO on JetPack 6.2.1. This blocks basic hardware I/O functionality.
⚡ GPIO regressions across JetPack versions can block hardware control loops essential for robotics deployments.
- Intermittent Flash Failure when using Cloned Images on Jetsonnvidia-forum:robotics-edge-computing5/11/2026deploymentjetsonflashingcloned-imagesmanufacturingprovisioningreliability
User reports intermittent flash failures when using cloned images on Jetson. This creates unreliable provisioning and blocks repeatable deployment.
⚡ Cloning-based manufacturing workflows need robust flashing reliability to scale deployments without manual rework.
- PKC Key revocation not working on L4T R36.5.0nvidia-forum:robotics-edge-computing5/11/2026deploymentjetsonl4t-r36-5-0securitypkckey-revocationsecure-boot
PKC key revocation reportedly does not work on L4T R36.5.0. This undermines secure provisioning and lifecycle management.
⚡ Security feature reliability (PKC revocation) is critical for trustworthy edge platforms used in safety- and mission-critical robots.
- [Bug Report] CloudXR JS / Runtime — inbound message channel and hand-tracking poses silently dropped between client and Isaac Sim's OpenXR plugingithub:isaac-sim/IsaacLab5/11/2026renderingv5.1renderinghardwaredeploymentintegrationisaac-simisaac-lab
In a CloudXR + OpenXR setup, frames stream correctly but inbound messages and hand-tracking poses are silently dropped between client and Isaac Sim’s OpenXR plugin. This blocks teleop commands and hand tracking for interactive workflows.
⚡ XR/teleop data-plane reliability is essential for human-in-the-loop data collection and remote operation within Isaac Sim.
- [bug] lift_cube_sm.py ignores --viz kit in Isaac Lab 3.0 betagithub:isaac-sim/IsaacLab5/11/2026crashes-stabilityv3.0crashrenderinghardwaredocsisaac-simisaac-lab
In Isaac Lab v3.0.0-beta, lift_cube_sm.py ignores the --viz kit option and no Kit/Isaac Sim window opens despite the process running. A one-line change to AppLauncher initialization appears to fix it locally.
⚡ Reliable visualization flags are key for debugging and onboarding; script-specific headless/viz mismatches erode confidence in examples.
- Issue creating application partition using remaining SSD space during flash/build processnvidia-forum:robotics-edge-computing5/11/2026deploymentjetsonflashingpartitioningssddeploymentstorage
User has an issue creating an application partition using remaining SSD space during the flash/build process. This blocks desired storage layout for deployment.
⚡ Flash-time partitioning flexibility is important for production provisioning and for maximizing storage for logs/datasets on robots.
- [BUG] Running the test suite leads to occasional OOBgithub:newton-physics/newton5/10/2026asset-pipelineusdhardwarenewtonwarp
Running the full Newton test suite occasionally triggers out-of-bounds errors, seen in some CUDA example tests. It is hard to reproduce in isolation, suggesting an order- or resource-dependent issue.
⚡ Rare OOB errors in full test runs indicate nondeterministic memory safety bugs that threaten reliability of the physics stack.
- Consolidate documented HashGrid query typesgithub:NVIDIA/warp5/10/2026docs-onboardingdocsfeature-requestwarp
Warp docs expose dtype-specific HashGrid query types alongside the intended wp.HashGridQuery, which users shouldn't need to handle. The request is to consolidate documentation/stubs around the single public type while keeping internal codegen behavior.
⚡ A smaller, cleaner Warp public API surface reduces confusion and lowers the barrier for custom physics/perception kernels.
- Support gradients for array-rooted composite augmented assignments and matrix row writesgithub:NVIDIA/warp5/10/2026hardware-integrationhardwarewarp
Some array-rooted composite augmented assignments and matrix row writes either fail codegen or silently drop gradients on current Warp main. This breaks differentiable programs that rely on these write patterns.
⚡ Correct gradient propagation in Warp is critical for differentiable simulation, optimization, and emerging DiffRL workflows.
- [Bug Report] I can not run Multi-GPU Training from IsaacLab 2.3.2 versiongithub:isaac-sim/IsaacLab5/9/2026training-infrav2.3.2rlrenderinghardwaredocsintegrationisaac-sim
### Describe the bug When I try to use multiple GPUs for my IsaacLab project from the IsaacLab 2.3.2 version, I get an error message. ### Steps to reproduce I used the command below. `python -m torch.distributed.run --nnodes=1 --nproc_per_node=2 scripts/skrl
- Pip dependency resolver conflict: isaacsim-core pins packaging==23.0; incompatible with wheel (requires packaging>=24)github:isaac-sim/IsaacLab5/8/2026otherisaac-simisaac-lab
### Summary During pip installation of Isaac Sim / Isaac Lab dependencies, pip prints resolver conflicts involving `packaging`. This creates noisy `ERROR:` lines even when installation completes successfully. It would help users if NVIDIA could align `isaacsim
- Articulation Initialisation Error when loading scene with two policy controller robotsgithub:isaac-sim/IsaacSim5/8/2026training-infrarlusdrenderinghardwaredeploymentlocomotion
### Description When loading scene (interactive example) with two robots both using seperate policies, an error occurs in the initialisation, whereby all articulation joints in scene are found, and not per robot. Two robots have been created with code from e
- Isaac Newton on Laputahn:isaac lab5/14/2026▲ 9otherisaac-labcommunitysignal-noisehn
A Hacker News item titled “Isaac Newton on Laputa” appears under an Isaac Lab feed but provides no actionable product feedback in the content captured.
⚡ Off-topic or low-signal social chatter under the Isaac Lab label can dilute actionable feedback and complicate signal extraction for product decisions.
- Isaac Sim 4.5 GUI Crashes with Errors when Running IsaacLab Examplesnvidia-forum:simulation5/12/2026▲ 1crashes-stabilityv4.5crashisaac-simisaac-lab
Isaac Sim 4.5 GUI crashes with errors when running IsaacLab examples. This blocks users from executing standard example pipelines.
⚡ Crashes in common IsaacLab example workflows can halt onboarding and reduce confidence in version compatibility.
- Character Animation playback in Isaac Sim doesn't work as expectednvidia-forum:simulation5/11/2026▲ 1otherisaac-sim
Character animation playback in Isaac Sim does not work as expected. This indicates issues in animation playback or sequencing.
⚡ Character animation correctness affects humanoid simulation, synthetic data generation, and interactive demo credibility.
- franka-Tutorial for cuMotion MoveIt Plugin with Isaac Simnvidia-forum:isaac5/10/2026▲ 1docs-onboardingdocsintegrationisaac-sim
User requests a Franka tutorial for the cuMotion MoveIt plugin with Isaac Sim. This suggests users struggle to connect the pieces without a guided example.
⚡ Documented reference workflows for popular robots (Franka) reduce friction and increase platform stickiness.
- franka-Tutorial for cuMotion MoveIt Plugin with Isaac Simnvidia-forum:isaac-ros5/10/2026▲ 1docs-onboardingdocsintegrationisaac-sim
User requests a Franka tutorial for the cuMotion MoveIt plugin with Isaac Sim. This indicates a documentation gap for a common manipulator workflow.
⚡ High-quality end-to-end tutorials (Franka + cuMotion + MoveIt + Isaac Sim) accelerate adoption of the planning stack.
- Built a tool that generates Isaac Sim environments from text prompts — looking for feedbacknvidia-forum:simulation5/9/2026▲ 1otherisaac-sim
- PEAKgithub:NVIDIA/warp5/14/2026otherwarpgithub-issuestriagesignal-noise
A GitHub issue titled “PEAK” in NVIDIA/warp contains unrelated, non-technical text and no actionable request or bug details. This creates triage noise.
⚡ Maintaining high-quality issue intake and moderation is important to keep developer channels usable and to preserve trust in NVIDIA tooling repos.
- Improve margin and gap of brick_stacking examplegithub:newton-physics/newton5/13/2026environment-designnewton-physicscontactsstackingmanipulationcollision-marginsphysics-tuning
Newton’s brick_stacking example generates many contacts that appear to float due to large shape margin+gap settings. They want more reasonable parameters while maintaining the example’s behavior.
⚡ Contact-parameter defaults strongly shape perceived simulator fidelity; visible artifacts in stacking scenes can reduce confidence for manipulation workloads.
- [BUG] add_joint_free() allows parent bodies other than the world. MuJoCo requires the parent body to be the world. This discrepancy needs to be managed with a warning.github:newton-physics/newton5/13/2026integrationnewton-physicsmujocoapi-compatjointsmigrationwarnings
Newton’s `add_joint_free()` allows parent bodies other than the world, but MuJoCo requires the parent to be the world. They want to manage the discrepancy with a warning to avoid confusing behavior differences.
⚡ API semantics mismatches across physics/sim ecosystems increase porting costs, reinforcing demand for clear compatibility layers and warnings.
- cloth_franka: viewer COMs scattered due to cm→m conversion gapgithub:newton-physics/newton5/13/2026renderingrenderingnewton
The cloth_franka example simulates in centimeters but only partially converts data back to meters for visualization. Debug overlays like COM markers and joint/contact arrows appear meters away due to a cm→m mismatch.
⚡ Unit-consistency across simulation and visualization layers is critical to avoid misleading debugging and user misdiagnosis.
- Isaac Sim Crashing When Using RTX Sensorsnvidia-forum:simulation5/13/2026crashes-stabilitycrashrenderingisaac-sim
Isaac Sim is reported to crash when using RTX Sensors. The post provides no additional details beyond the crash condition.
⚡ RTX sensor stability is a core dependency for synthetic data and perception validation in Isaac Sim.
- [Question] Uncorrect force data from tacsl demogithub:isaac-sim/IsaacLab5/13/2026crashes-stabilitycrashusdrenderingsensorsisaac-simisaac-lab
TacSL force-field readings in an Isaac Lab demo do not increase smoothly with stepped applied normal force and instead appear irregular. The user reports Isaac Lab main branch with Isaac Sim 5.1.
⚡ Tactile sensing fidelity and correctness are key to manipulation learning credibility and downstream policy performance.
- Nova Orin Init for Nova Carter Robotnvidia-forum:isaac5/13/2026hardware-integrationhardware
A user asks about Nova Orin initialization for the Nova Carter Robot in the Isaac forum. The post contains no additional details.
⚡ Repeated Nova init onboarding questions across channels suggest missing or hard-to-find canonical setup documentation.
- Nova Orin Init for Nova Carter Robotnvidia-forum:isaac-ros5/13/2026hardware-integrationhardware
A user asks about Nova Orin initialization for the Nova Carter Robot in the Isaac ROS forum. The post contains no additional details.
⚡ Cross-posted Nova init questions indicate a need for clearer Isaac ROS + platform setup pathways.
- Robot-Fluid Coupled Simulationnvidia-forum:simulation5/12/2026other
A user asks about robot-fluid coupled simulation. No further detail is provided.
⚡ Demand for robot–fluid coupling points to a growing need for multi-physics capabilities and tighter integration surfaces.
- Investigate reasons for macos runner slowdowngithub:newton-physics/newton5/12/2026other
Newton’s macOS runner is getting progressively slower, likely tied to cache_kernel behavior during CPU compilation. A temporary mitigation moves the macOS job to post-merge but is intended to be reverted once fixed.
⚡ Developer velocity depends on predictable CI runtimes; slow macOS runners increase iteration cost and delay releases.
- [Bug Report] Articulation vel data resetgithub:isaac-sim/IsaacLab5/12/2026otherisaac-sim
Velocity-only write paths on Isaac Lab Articulation do not invalidate cached derived body state buffers. Downstream code may read stale body velocity/state after writing velocities to sim.
⚡ State caching and invalidation bugs in the env API can silently corrupt training signals and waste compute.
- Carrier-Board PCIenvidia-forum:robotics-edge-computing5/12/2026hardware-integrationcarrier-boardpciejetsonhardware-designintegration
A user asks about carrier-board PCIe. The post has no additional details.
⚡ Carrier-board PCIe integration questions signal continued need for clearer reference designs and validation guidance.
- MGBE to ethernet RJ45nvidia-forum:robotics-edge-computing5/12/2026hardware-integrationmgbeethernetrj45jetsoncarrier-board
A user asks about connecting MGBE to an ethernet RJ45. No further information is included.
⚡ Networking interface integration remains a frequent friction point for custom robotics hardware designs.
- Rigid body contact determinismgithub:newton-physics/newton5/12/2026crashes-stabilitycrash
There is a request to ensure vbd rigid body results are deterministic when the collision pipeline provides deterministic contact pairs. This targets repeatability of contact-driven dynamics.
⚡ Deterministic contact outcomes are a prerequisite for reproducible training, debugging, and regression testing in physics simulation.
- How to display DDR memory info such as manufacturernvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonddrmemorydiagnosticsmanufacturing
User asks how to display DDR memory information such as manufacturer. This is a diagnostics/visibility request.
⚡ Hardware inventory/telemetry transparency (e.g., DDR manufacturer) is important for debugging and fleet QA on Jetson.
- Creating a Digital Twin for Quantum Energy Harvesting (Project Monolith) inside Omniversenvidia-forum:simulation5/12/2026asset-pipelineusd
User is working on creating a digital twin for 'Quantum Energy Harvesting (Project Monolith)' inside Omniverse. This appears to be exploratory and likely seeks guidance or feasibility confirmation.
⚡ Non-traditional digital twin requests indicate Omniverse is being evaluated for niche domains, suggesting opportunities for extensibility narratives.
- NVIDIA Jetson Thor Developer Kit同时使用两条SSD的可能性nvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonthorssdstoragepcie
User asks about the possibility of using two SSDs simultaneously on the NVIDIA Jetson Thor Developer Kit. This is an integration capability question.
⚡ Storage expansion questions reflect growing demand for higher local throughput/capacity on Thor dev kits.
- Character_graph = ag.get_character(str(character_skel_root.GetPrimPath())) output is Nonenvidia-forum:simulation5/12/2026other
Calling ag.get_character with a character skeleton root path returns None. This blocks workflows that rely on obtaining a character_graph handle.
⚡ Animation/character-graph API failures reduce usability for humanoids and character-driven simulation scenarios.
- Jetson设备的最小文件系统的pcie相关功能nvidia-forum:robotics-edge-computing5/12/2026hardware-integrationjetsonminimal-rootfspcieembeddedenablement
User asks about PCIe-related functionality in a minimal filesystem for Jetson devices. This indicates a need for guidance on which components are required for PCIe support.
⚡ Minimal filesystem enablement for PCIe affects embedded integrators optimizing footprint and security.
- Why were the kernel and dtb partitions removed on Thor?nvidia-forum:robotics-edge-computing5/11/2026docs-onboardingjetsonthorbootpartitionsdtbkernel
User asks why kernel and dtb partitions were removed on Thor. This indicates confusion about platform boot/partition layout changes.
⚡ Platform partitioning/boot changes need proactive communication to prevent confusion and downstream tooling breakage.
- 在jetpack6.1.2适配的系统中, 如何修改默认的boot ordernvidia-forum:robotics-edge-computing5/11/2026deploymentjetsonjetpack-6-1-2boot-orderuefideployment
User asks (in Chinese) how to modify the default boot order on a system adapted for JetPack 6.1.2. This is a boot configuration question impacting deployment.
⚡ Boot order customization is a common deployment requirement; friction here slows integration with enterprise IT and field operations.
- GPIO on the Orin AGXnvidia-forum:robotics-edge-computing5/11/2026hardware-integrationjetsonorin-agxgpiohardware-iobringup
User asks about GPIO on the Orin AGX. This indicates need for guidance on pin mapping, access, or configuration.
⚡ Clear GPIO mappings and control guidance reduce integration time for custom robot electronics on Jetson.
- Register Kamino examples in README and add screenshotsgithub:newton-physics/newton5/11/2026training-infrarlmanipulationnewton
Kamino examples were added but are missing required README entries and screenshots on the release branch. This reduces discoverability and fails documented contribution requirements.
⚡ Keeping examples discoverable and well-documented improves adoption of Newton/Kamino workflows and reduces internal support costs.
- PCIe lane reversal in AGX Orinnvidia-forum:robotics-edge-computing5/11/2026hardware-integrationjetsonagx-orinpcielane-reversalcarrier-boardhardware-design
User asks about PCIe lane reversal on AGX Orin. This indicates a hardware design/bring-up question for PCIe routing.
⚡ PCIe topology nuances (lane reversal) are common stumbling blocks for custom carriers, impacting peripheral compatibility and schedules.
- Animations are not applied correctly to the characternvidia-forum:simulation5/11/2026other
Animations are not applied correctly to a character. This suggests a rig/retarget/apply pipeline problem in Isaac Sim.
⚡ Animation retargeting/application bugs reduce usability for avatar/humanoid workflows and can block content pipelines.
- Improve mouse picking for cable examples.github:newton-physics/newton5/10/2026other
Mouse dragging in cable examples introduces overly large forces, possibly from large torques applied to segments. This makes interactive manipulation unreliable.
⚡ Interactive tooling quality (mouse picking) influences how effectively users can debug and demonstrate deformables/rope simulations.
- Assign uniform color to cable segments by default.github:newton-physics/newton5/10/2026other
Request to assign a uniform color to cable segments by default because current visualization is overly colorful. This is a usability/visual quality improvement.
⚡ Small visualization defaults matter for demos and debugging, improving perceived quality of physics examples.
- PVA SDK Developer Applicationnvidia-forum:robotics-edge-computing5/12/2026▲ 1feature-requestspvasdkjetsondeveloper-accessedge-ai
A user requests a PVA SDK developer application. No additional details are provided in the post.
⚡ Interest in PVA tooling access suggests demand for on-device acceleration pathways beyond GPU-centric pipelines.
- Native Point Cloud Visualization in Viewergithub:google-deepmind/mujoco5/14/2026renderingmujocopoint-cloudpcdplyviewerdebugging
Request to add native point cloud import and visualization (e.g., PLY/PCD) to the MuJoCo viewer as an efficient vertex rendering overlay. Current workaround renders each vertex as a sphere, which is inefficient for ground-truth verification.
⚡ Point-cloud-native inspection is becoming table-stakes for robotics debugging, raising expectations for equivalent or better tooling in NVIDIA simulation workflows.
- Complete manual RC test matrix automationgithub:newton-physics/newton5/13/2026training-infranewton-physicsrelease-engineeringrc-testingtest-matrixcudadrivers
Newton wants reusable automation to run a release-candidate validation matrix (OS, Python, CUDA/driver, etc.) for sign-off. Today this is tracked and run manually; they want a consistent launch and reporting workflow for RC branches/tags/commits.
⚡ Release-quality automation across OS/Python/CUDA matrices is increasingly expected, and gaps here can slow shipping cadence and erode confidence in GPU-accelerated stacks.
- [Proposal] Integrate differentiable Reinforcement Learning (DiffRL) support into IsaacLabgithub:isaac-sim/IsaacLab5/12/2026training-infrav3.0rlintegrationisaac-simisaac-labnewtonwarp
Proposal to integrate DiffRL into IsaacLab via an isaaclab_diffrl extension and the Mineral algorithm library. It targets the Direct workflow and Newton (Warp) backend to enable end-to-end backprop through physics for algorithms like SHAC and APG/BPTT.
⚡ Differentiable RL integration would position Isaac Lab + Newton/Warp as a first-class platform for gradient-based control research and faster policy iteration.
- Distortion correctionnvidia-forum:robotics-edge-computing5/12/2026sensors-perceptionjetsoncameracalibrationdistortion-correctionperception
User asks about distortion correction. This appears to be a request for capability or guidance on correcting lens distortion.
⚡ Distortion correction demand highlights the need for robust camera calibration tooling across Jetson perception stacks.