# Latest **URL:** https://forums.developer.nvidia.com/latest.md [Latest](https://forums.developer.nvidia.com/latest.md) · [Categories](https://forums.developer.nvidia.com/categories.md) · [Tags](https://forums.developer.nvidia.com/tags.md) --- ## [PyNvVideoCodec 2.2 is released](https://forums.developer.nvidia.com/t/pynvvideocodec-2-2-is-released/380856)
**Author:** [@mandar\_godse](https://forums.developer.nvidia.com/u/mandar_godse)\ **Replies:** 4\ **Last updated:** [September 16, 2026, 10:24am UTC](https://forums.developer.nvidia.com/t/pynvvideocodec-2-2-is-released/380856 "2026-09-16T10:24:00Z")
We are pleased to announce the release of PyNvVideoCodec 2.2, the latest version of our GPU-accelerated Python library for video decoding, encoding, and transcoding , all with the simplicity and flexibility of Python. P… --- ## [Introducing NVIDIA NemoClaw](https://forums.developer.nvidia.com/t/introducing-nvidia-nemoclaw/363701)
**Author:** [@rjensen](https://forums.developer.nvidia.com/u/rjensen)\ **Replies:** 0\ **Last updated:** [March 16, 2026, 8:22pm UTC](https://forums.developer.nvidia.com/t/introducing-nvidia-nemoclaw/363701 "2026-03-16T20:22:22Z")
NVIDIA NemoClaw is an open source stack that simplifies running OpenClaw always-on assistants more safely, with a single command. It installs the NVIDIA OpenShell runtime, part of the NVIDIA Agent Toolkit, a secure envir… --- ## [DRIVEOS 7.2.5 AGX Thor P3960: OX03H10-EXAE-BA0C DriveWorks camera.gmsl fails in parseCSIPortMapInfo due to missing des\_tx\_port while NvSIPL works](https://forums.developer.nvidia.com/t/driveos-7-2-5-agx-thor-p3960-ox03h10-exae-ba0c-driveworks-camera-gmsl-fails-in-parsecsiportmapinfo-due-to-missing-des-tx-port-while-nvsipl-works/388009)
**Author:** [@keshav.poygaipakkamkhanda](https://forums.developer.nvidia.com/u/keshav.poygaipakkamkhanda)\ **Replies:** 19\ **Last updated:** [November 5, 2026, 7:27pm UTC](https://forums.developer.nvidia.com/t/driveos-7-2-5-agx-thor-p3960-ox03h10-exae-ba0c-driveworks-camera-gmsl-fails-in-parsecsiportmapinfo-due-to-missing-des-tx-port-while-nvsipl-works/388009 "2026-11-05T19:27:18Z")
DRIVE OS Version: Platform NVIDIA DRIVE AGX Thor P3960 DRIVE OS 7.2.5 Target OS: Ubuntu 24.04.4 LTS Kernel: 6.8.12-rt-tegra Camera: OX03H10-EXAE-BA0C, 3 MP Deserializer: MAX96724 DriveWorks: 7.2 (libdw\_sensors\_std.so.… --- ## [Qwen4](https://forums.developer.nvidia.com/t/qwen4/388002)
**Author:** [@Digital\_David](https://forums.developer.nvidia.com/u/Digital_David)\ **Replies:** 11\ **Last updated:** [November 5, 2026, 7:24pm UTC](https://forums.developer.nvidia.com/t/qwen4/388002 "2026-11-05T19:24:03Z")
There’s a pattern worth naming that nobody seems to be tracking: the Qwen4 era is arriving in serving stacks before any Qwen4 weights exist. Not speculation any longer, and I’m excited! 1. The engine landed before the w… --- ## [620 release feedback & discussion](https://forums.developer.nvidia.com/t/620-release-feedback-discussion/388030)
**Author:** [@aplattner](https://forums.developer.nvidia.com/u/aplattner)\ **Replies:** 5\ **Last updated:** [November 5, 2026, 7:14pm UTC](https://forums.developer.nvidia.com/t/620-release-feedback-discussion/388030 "2026-11-05T19:14:11Z")
620.70.06 (x86\_64 / aarch64) Improved performance of explicit synchronization in high-framerate benchmarks on Wayland. Fixed a regression, introduced in the 595.xx driver series, which caused Assassin’s Creed Shadows t… --- ## [Some GPIO Pins on the Nvidia Jetson are not working](https://forums.developer.nvidia.com/t/some-gpio-pins-on-the-nvidia-jetson-are-not-working/383446)
**Author:** [@bigbor14coe](https://forums.developer.nvidia.com/u/bigbor14coe)\ **Replies:** 9\ **Last updated:** [November 5, 2026, 7:05pm UTC](https://forums.developer.nvidia.com/t/some-gpio-pins-on-the-nvidia-jetson-are-not-working/383446 "2026-11-05T19:05:51Z")
Hi, i have the following Problem with my Nvidia Jetson AGX Orin Dev Kit: Some of my GPIO Pins are constantly set to High or Low. I have already used the GPIO configuration tool to make sure that the problematic pins ar… --- ## [OTP to Verify Phone Number Error](https://forums.developer.nvidia.com/t/otp-to-verify-phone-number-error/388055)
**Author:** [@daboldflex](https://forums.developer.nvidia.com/u/daboldflex)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 7:00pm UTC](https://forums.developer.nvidia.com/t/otp-to-verify-phone-number-error/388055 "2026-11-05T19:00:57Z")
Looking to verify account to use API. Have used multiple phone numbers, and have double and triple checked them. Still get the same error: “The phone number you provided is invalid, please try again.” Please fix! --- ## [Installing Linux Drivers on Ubuntu](https://forums.developer.nvidia.com/t/installing-linux-drivers-on-ubuntu/388105)
**Author:** [@atlnycdude23](https://forums.developer.nvidia.com/u/atlnycdude23)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 7:00pm UTC](https://forums.developer.nvidia.com/t/installing-linux-drivers-on-ubuntu/388105 "2026-11-05T19:00:48Z")
Hello, I’m new here. I need help on installing drivers on pc that i’m building. I need help installing drivers on Ubuntu 26.04.1 LTS for (2x) RTX 3090 GPUs and CUDA ready for GPU, please assist, thanks! --- ## [Alpamayo 2 Edge (2B): expected release date for weights,?](https://forums.developer.nvidia.com/t/alpamayo-2-edge-2b-expected-release-date-for-weights/388068)
**Author:** [@alejandro.bastida](https://forums.developer.nvidia.com/u/alejandro.bastida)\ **Replies:** 2\ **Last updated:** [November 5, 2026, 6:57pm UTC](https://forums.developer.nvidia.com/t/alpamayo-2-edge-2b-expected-release-date-for-weights/388068 "2026-11-05T18:57:27Z")
Hi everyone, I’m a student at Universidad Politécnica de Cartagena (Spain) working on my final degree project (TFG). The project is about deploying Alpamayo-based VLA models on real hardware: closed-loop validation in A… --- ## [Bit-reproducible MXFP4 GEMV on an RTX 4060 (sm\_89): how to make it faster without changing the summation order?](https://forums.developer.nvidia.com/t/bit-reproducible-mxfp4-gemv-on-an-rtx-4060-sm-89-how-to-make-it-faster-without-changing-the-summation-order/387388)
**Author:** [@arielstone883](https://forums.developer.nvidia.com/u/arielstone883)\ **Replies:** 8\ **Last updated:** [November 5, 2026, 6:44pm UTC](https://forums.developer.nvidia.com/t/bit-reproducible-mxfp4-gemv-on-an-rtx-4060-sm-89-how-to-make-it-faster-without-changing-the-summation-order/387388 "2026-11-05T18:44:23Z")
Hello, I’m a student working on a personal project: serving the MXFP4 expert weights of gpt-oss-120b on a laptop with an 8 GB RTX 4060. I directed it with Claude Code: I set the constraints and the checks, and agents wr… --- ## [Running nvidia/DeepSeek-V4.1-Flash-1017-NVFP4 (552B + 196B Engram) on a single 128 GB GB10 (ASUS Ascent GX10 / DGX Spark) in Pure Rust (1.1 MB binary,](https://forums.developer.nvidia.com/t/running-nvidia-deepseek-v4-1-flash-1017-nvfp4-552b-196b-engram-on-a-single-128-gb-gb10-asus-ascent-gx10-dgx-spark-in-pure-rust-1-1-mb-binary/388102)
**Author:** [@tommy.k.a.a](https://forums.developer.nvidia.com/u/tommy.k.a.a)\ **Replies:** 3\ **Last updated:** [November 5, 2026, 6:33pm UTC](https://forums.developer.nvidia.com/t/running-nvidia-deepseek-v4-1-flash-1017-nvfp4-552b-196b-engram-on-a-single-128-gb-gb10-asus-ascent-gx10-dgx-spark-in-pure-rust-1-1-mb-binary/388102 "2026-11-05T18:33:57Z")
Hi everyone, We built a standalone, zero-dependency Rust + sm\_121 PTX inference engine (Flash1017-Engine / flash1017-gx10) specifically locked and tuned for the NVIDIA Grace Blackwell GB10 (128 GB Unified Memory — ASUS … --- ## [Apı key error](https://forums.developer.nvidia.com/t/api-key-error/388122)
**Author:** [@ai.velmora.ai](https://forums.developer.nvidia.com/u/ai.velmora.ai)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 6:30pm UTC](https://forums.developer.nvidia.com/t/api-key-error/388122 "2026-11-05T18:30:06Z")
Title: NVIDIA Cosmos3 Nano API Returns HTTP 404 – Image-to-Video Generation Not Working Hello NVIDIA Developer Support Team, I am trying to integrate the NVIDIA Cosmos3 Nano model into a custom Python application on Wi… --- ## [RTX PRO 6000 Blackwell Max-Q: Xid 79, CUDA failure and Gen3 link under load](https://forums.developer.nvidia.com/t/rtx-pro-6000-blackwell-max-q-xid-79-cuda-failure-and-gen3-link-under-load/388120)
**Author:** [@mabarkdoll](https://forums.developer.nvidia.com/u/mabarkdoll)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 6:21pm UTC](https://forums.developer.nvidia.com/t/rtx-pro-6000-blackwell-max-q-xid-79-cuda-failure-and-gen3-link-under-load/388120 "2026-11-05T18:21:53Z")
System: Lenovo ThinkStation P8, Threadripper PRO 7965WX, 1400 W PSU Three RTX PRO 6000 Blackwell Max-Q Workstation Edition GPUs, each at its stock 300 W power limit AlmaLinux 9, kernel 5.14.0-687.65.1.el9\_8.x86\_64 NVI… --- ## [Is it worth buying a second DGX Spark now?](https://forums.developer.nvidia.com/t/is-it-worth-buying-a-second-dgx-spark-now/384081)
**Author:** [@alef204](https://forums.developer.nvidia.com/u/alef204)\ **Replies:** 45\ **Last updated:** [November 5, 2026, 6:16pm UTC](https://forums.developer.nvidia.com/t/is-it-worth-buying-a-second-dgx-spark-now/384081 "2026-11-05T18:16:52Z")
I have a second DGX Spark reserved (ASUS, 1 TB SSD) for €5,100 including VAT. I bought my first one in March for €3,200. I’m currently running the Qwen3.8-Flash-Next model. I’d also like to run faster models with around… --- ## [Timeout issue when executing hololink.start()](https://forums.developer.nvidia.com/t/timeout-issue-when-executing-hololink-start/388047)
**Author:** [@tom153](https://forums.developer.nvidia.com/u/tom153)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 6:05pm UTC](https://forums.developer.nvidia.com/t/timeout-issue-when-executing-hololink-start/388047 "2026-11-05T18:05:32Z")
I have encountered timeout issues while testing HSB on the Jetson AGX Thor. hololink-enumerate is able to retrieve device information correctly. However, when I run my own program, it gets stuck at hololink.start() … --- ## [The system hangs while running the fTPM stress test script on the Orin Nano](https://forums.developer.nvidia.com/t/the-system-hangs-while-running-the-ftpm-stress-test-script-on-the-orin-nano/388089)
**Author:** [@mack.shi](https://forums.developer.nvidia.com/u/mack.shi)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 6:02pm UTC](https://forums.developer.nvidia.com/t/the-system-hangs-while-running-the-ftpm-stress-test-script-on-the-orin-nano/388089 "2026-11-05T18:02:11Z")
Hardware: Orin Nano L4T / JetPack version: cat /etc/nv\_tegra\_release # R39 (release), REVISION: 2.0, GCID: 45755727, BOARD: generic, EABI: aarch64, DATE: Mon Jun 1 09:28:48 PM UTC 2026 # KERNEL\_VARIANT: oot TARGET\_USE… --- ## [Manual Account Verification Request – API Access Unavailable](https://forums.developer.nvidia.com/t/manual-account-verification-request-api-access-unavailable/388119)
**Author:** [@svenstark](https://forums.developer.nvidia.com/u/svenstark)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 5:55pm UTC](https://forums.developer.nvidia.com/t/manual-account-verification-request-api-access-unavailable/388119 "2026-11-05T17:55:43Z")
Hello NVIDIA Team, I am unable to generate an API key on build.nvidia.com. The API Keys page shows “API Access Unavailable – Please contact support to verify your account access.” My personal organization appears to be … --- ## [Incorrect resizing of Fortran array in loop](https://forums.developer.nvidia.com/t/incorrect-resizing-of-fortran-array-in-loop/388059)
**Author:** [@edoy](https://forums.developer.nvidia.com/u/edoy)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 5:51pm UTC](https://forums.developer.nvidia.com/t/incorrect-resizing-of-fortran-array-in-loop/388059 "2026-11-05T17:51:15Z")
Hi, I found this bug while trying to get our coupled ocean model working with NVHPC. At -O2, when an unallocated allocatable array is sized at the first iteration of a loop, and then resized at a later iteration, the si… --- ## [DIY DGX Spark Cooling System for 2x Cluster](https://forums.developer.nvidia.com/t/diy-dgx-spark-cooling-system-for-2x-cluster/371943)
**Author:** [@porespellar](https://forums.developer.nvidia.com/u/porespellar)\ **Replies:** 2\ **Last updated:** [November 5, 2026, 5:11pm UTC](https://forums.developer.nvidia.com/t/diy-dgx-spark-cooling-system-for-2x-cluster/371943 "2026-11-05T17:11:49Z")
I’ve found that DGX Sparks can get pretty warm when you cluster them together. You are forced to keep them close together because the ConnectX-7 cable made for these is extremely short )like less than a foot). I have… --- ## [cusolverDnXsyevBatched occasionally reports non-convergence (info \> 0) although cusolverDnXsyevd succeeds and residual is small](https://forums.developer.nvidia.com/t/cusolverdnxsyevbatched-occasionally-reports-non-convergence-info-0-although-cusolverdnxsyevd-succeeds-and-residual-is-small/387343)
**Author:** [@gpenazzi](https://forums.developer.nvidia.com/u/gpenazzi)\ **Replies:** 2\ **Last updated:** [November 5, 2026, 5:03pm UTC](https://forums.developer.nvidia.com/t/cusolverdnxsyevbatched-occasionally-reports-non-convergence-info-0-although-cusolverdnxsyevd-succeeds-and-residual-is-small/387343 "2026-11-05T17:03:40Z")
Hi all, We have observed what looks like potentially undesired behavior in cusolverDnXsyevBatched. Occasionally the routine returns info \> 0 for matrices arising in our quantum chemistry workloads. The behavior appears… --- ## [Configure, Certify, and Operate Kubernetes GPU Clusters with NVIDIA DSX OS](https://forums.developer.nvidia.com/t/configure-certify-and-operate-kubernetes-gpu-clusters-with-nvidia-dsx-os/388118)
**Author:** [@jwitsoe](https://forums.developer.nvidia.com/u/jwitsoe)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 5:00pm UTC](https://forums.developer.nvidia.com/t/configure-certify-and-operate-kubernetes-gpu-clusters-with-nvidia-dsx-os/388118 "2026-11-05T17:00:13Z")
Originally published at: Configure, Certify, and Operate Kubernetes GPU Clusters with NVIDIA DSX OS | NVIDIA Technical Blog One underperforming GPU can slow an entire distributed training job. The node may pass basic… --- ## [Issue where HSB multi-camera transmission terminates unexpectedly with an error](https://forums.developer.nvidia.com/t/issue-where-hsb-multi-camera-transmission-terminates-unexpectedly-with-an-error/388048)
**Author:** [@tom153](https://forums.developer.nvidia.com/u/tom153)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 4:38pm UTC](https://forums.developer.nvidia.com/t/issue-where-hsb-multi-camera-transmission-terminates-unexpectedly-with-an-error/388048 "2026-11-05T16:38:50Z")
Hi, When testing HSB multi-camera streaming on the Jetson AGX Thor, I observed that the program suddenly throws an error and exits after running for a period of time. What could be the potential causes for this? Os … --- ## [Can not access Self Paced Course RDMA over Converged Ethernet (RoCE) from A to Z](https://forums.developer.nvidia.com/t/can-not-access-self-paced-course-rdma-over-converged-ethernet-roce-from-a-to-z/387880)
**Author:** [@mshellik](https://forums.developer.nvidia.com/u/mshellik)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 4:26pm UTC](https://forums.developer.nvidia.com/t/can-not-access-self-paced-course-rdma-over-converged-ethernet-roce-from-a-to-z/387880 "2026-11-05T16:26:26Z")
Can not access Free Self Paced Course RDMA over Converged Ethernet (RoCE) from A to Z --- ## [MSS: Failed to initialize SDRAM](https://forums.developer.nvidia.com/t/mss-failed-to-initialize-sdram/387993)
**Author:** [@gebruiker639](https://forums.developer.nvidia.com/u/gebruiker639)\ **Replies:** 2\ **Last updated:** [November 5, 2026, 4:10pm UTC](https://forums.developer.nvidia.com/t/mss-failed-to-initialize-sdram/387993 "2026-11-05T16:10:29Z")
Dear all, After temperature testing the Orin AGX module on a custom carrier board, the Orin AGX module does no longer boot. The boot log of the serial port mentions: \[0000.990\] E\> MSS: Failed to initialize SDRAM. \[000… --- ## [Five Ways NVIDIA Co-Design Maximizes the Value of Every Byte in the AI Factory](https://forums.developer.nvidia.com/t/five-ways-nvidia-co-design-maximizes-the-value-of-every-byte-in-the-ai-factory/388116)
**Author:** [@jwitsoe](https://forums.developer.nvidia.com/u/jwitsoe)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 4:00pm UTC](https://forums.developer.nvidia.com/t/five-ways-nvidia-co-design-maximizes-the-value-of-every-byte-in-the-ai-factory/388116 "2026-11-05T16:00:13Z")
Originally published at: Five Ways NVIDIA Co-Design Maximizes the Value of Every Byte in the AI Factory | NVIDIA Technical Blog In an AI factory, high bandwidth memory (HBM)is the fastest, most expensive, and most co… --- ## [Holoscan 25G and Dual/Quad Network Examples for Thor AGX](https://forums.developer.nvidia.com/t/holoscan-25g-and-dual-quad-network-examples-for-thor-agx/388013)
**Author:** [@joe\_schmo](https://forums.developer.nvidia.com/u/joe_schmo)\ **Replies:** 1\ **Last updated:** [November 5, 2026, 3:59pm UTC](https://forums.developer.nvidia.com/t/holoscan-25g-and-dual-quad-network-examples-for-thor-agx/388013 "2026-11-05T15:59:13Z")
I noticed today that there is a new release (3.0.0) for the Holoscan Sensor Bridge. Looking at the provided examples, I still do not see any 25G or any Dual/Quad Network examples. Are there any such examples that exist? … --- ## [Manual account verification request (Russia +7)](https://forums.developer.nvidia.com/t/manual-account-verification-request-russia-7/388115)
**Author:** [@RomanOneAndFun](https://forums.developer.nvidia.com/u/RomanOneAndFun)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 3:58pm UTC](https://forums.developer.nvidia.com/t/manual-account-verification-request-russia-7/388115 "2026-11-05T15:58:55Z")
Hello! I am trying to verify my NVIDIA Developer account to get access to NVIDIA NIM / APIs, but I am unable to receive the SMS verification code. The phone country code for Russia (+7) is missing from the dropdown list… --- ## [Manual Account Verification from Belarus](https://forums.developer.nvidia.com/t/manual-account-verification-from-belarus/388114)
**Author:** [@night1429werdi](https://forums.developer.nvidia.com/u/night1429werdi)\ **Replies:** 0\ **Last updated:** [November 5, 2026, 3:50pm UTC](https://forums.developer.nvidia.com/t/manual-account-verification-from-belarus/388114 "2026-11-05T15:50:25Z")
Hello-hello! I wanted to verify my account on Nvidia to use its API, but… I cant do that, because of Belarus missing in the list of accesible countrys. If there any way to verify it manualy? Email: night1429werdi@gmail.… --- ## [OpenMP \`target nowait\` has large overhead at start of new stream in nvfortran 26.3/26.9](https://forums.developer.nvidia.com/t/openmp-target-nowait-has-large-overhead-at-start-of-new-stream-in-nvfortran-26-3-26-9/388005)
**Author:** [@utheriwagura](https://forums.developer.nvidia.com/u/utheriwagura)\ **Replies:** 2\ **Last updated:** [November 5, 2026, 3:47pm UTC](https://forums.developer.nvidia.com/t/openmp-target-nowait-has-large-overhead-at-start-of-new-stream-in-nvfortran-26-3-26-9/388005 "2026-11-05T15:47:21Z")
I’ve spent some time over the past week comparing the performance of OpenMP’s nowait clause to OpenACC’s async in nvfortran 26.3. It initially seemed that as though the nowait clause was slower, but upon further investi… --- ## [590 release feedback & discussion](https://forums.developer.nvidia.com/t/590-release-feedback-discussion/353310)
**Author:** [@aplattner](https://forums.developer.nvidia.com/u/aplattner)\ **Replies:** 461\ **Last updated:** [November 5, 2026, 3:33pm UTC](https://forums.developer.nvidia.com/t/590-release-feedback-discussion/353310 "2026-11-05T15:33:17Z")
590.44.01 (x86\_64 / aarch64) Raised the minimum supported Wayland version to 1.20. Fixed a bug that prevented the PowerMizer preferred mode dropdown menu in the nvidia-settings control panel from functioning correctly … [Next page](https://forums.developer.nvidia.com/latest.md?page=1)