# 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) --- ## [The way to prepare for GTC Berlin 2026](https://forums.developer.nvidia.com/t/the-way-to-prepare-for-gtc-berlin-2026/386296)
**Author:** [@MarkusHoHo](https://forums.developer.nvidia.com/u/MarkusHoHo)\ **Replies:** 0\ **Last updated:** [October 19, 2026, 8:31am UTC](https://forums.developer.nvidia.com/t/the-way-to-prepare-for-gtc-berlin-2026/386296 "2026-10-19T08:31:53Z")
Can’t make it to #NVIDIAGTC Berlin? We’ve got you covered. Join us for 3 days of livestreams covering the biggest moments of the week, from the keynote recap to real-world AI demos and developer stories straight from B… --- ## [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… --- ## [Open a Support Case for DLI Issues](https://forums.developer.nvidia.com/t/open-a-support-case-for-dli-issues/345829)
**Author:** [@TomNVIDIA](https://forums.developer.nvidia.com/u/TomNVIDIA)\ **Replies:** 0\ **Last updated:** [September 23, 2025, 2:37pm UTC](https://forums.developer.nvidia.com/t/open-a-support-case-for-dli-issues/345829 "2025-09-23T14:37:14Z")
For DLI Course assistance, please visit the DLI Support Page to open a support case. This ensures your request is directed to the appropriate team for timely and accurate support. When submitting your case, please make … --- ## [Urgent: DLI Account Banned for 16+ Days - Pending Tickets #01286804 & #01292118](https://forums.developer.nvidia.com/t/urgent-dli-account-banned-for-16-days-pending-tickets-01286804-01292118/386492)
**Author:** [@Pb22j](https://forums.developer.nvidia.com/u/Pb22j)\ **Replies:** 1\ **Last updated:** [October 21, 2026, 2:33pm UTC](https://forums.developer.nvidia.com/t/urgent-dli-account-banned-for-16-days-pending-tickets-01286804-01292118/386492 "2026-10-21T14:33:22Z")
Hello NVIDIA DLI Team, I raised a support ticket on October 5, 2026 (more than 16 days ago) without any response so far. The problem is that my DLI account got banned/disabled, and I cannot access my account to continu… --- ## [Supported threading model for updating USD/Fabric off the main thread](https://forums.developer.nvidia.com/t/supported-threading-model-for-updating-usd-fabric-off-the-main-thread/386491)
**Author:** [@rashikoz](https://forums.developer.nvidia.com/u/rashikoz)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 2:16pm UTC](https://forums.developer.nvidia.com/t/supported-threading-model-for-updating-usd-fabric-off-the-main-thread/386491 "2026-10-21T14:16:26Z")
Operating System: Windows Linux Kit Version: 110 (Kit App Template) 109 (Kit App Template) 108 (Kit App Template) 107 (Kit App Template) 106 (Kit App Template) 105 (Launcher) Kit Template: USD Compose… --- ## [Flash to USB - Boot fails](https://forums.developer.nvidia.com/t/flash-to-usb-boot-fails/386483)
**Author:** [@sevm89](https://forums.developer.nvidia.com/u/sevm89)\ **Replies:** 2\ **Last updated:** [October 21, 2026, 2:04pm UTC](https://forums.developer.nvidia.com/t/flash-to-usb-boot-fails/386483 "2026-10-21T14:04:51Z")
Dear Nvidia Team We are trying to flash a custom carrier board directly to an attached USB device. The flashing runs through successfully with the following command: EXT\_NUM\_SECTORS=“45000000” ./l4t\_initrd\_flash.sh --… --- ## [Force recovery is hign but system into recovery mode](https://forums.developer.nvidia.com/t/force-recovery-is-hign-but-system-into-recovery-mode/385903)
**Author:** [@jerry.wu3](https://forums.developer.nvidia.com/u/jerry.wu3)\ **Replies:** 17\ **Last updated:** [October 21, 2026, 1:56pm UTC](https://forums.developer.nvidia.com/t/force-recovery-is-hign-but-system-into-recovery-mode/385903 "2026-10-21T13:56:53Z")
Subject / Issue Description: Custom Carrier Board for Jetson Thor Fails to Boot / No HDMI Output (No UART Debug Header Reserved) Current Symptoms & Troubleshooting Status: Our custom-designed carrier board fails to bo… --- ## [Two models on one DGX Spark: Qwen3.6-35B NVFP4 + Nemotron 3.5 Lightning with DSpark, numbers under shared load and a fix for cuDNN FP8 crashes](https://forums.developer.nvidia.com/t/two-models-on-one-dgx-spark-qwen3-6-35b-nvfp4-nemotron-3-5-lightning-with-dspark-numbers-under-shared-load-and-a-fix-for-cudnn-fp8-crashes/386489)
**Author:** [@michael\_allerstorfer](https://forums.developer.nvidia.com/u/michael_allerstorfer)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 1:46pm UTC](https://forums.developer.nvidia.com/t/two-models-on-one-dgx-spark-qwen3-6-35b-nvfp4-nemotron-3-5-lightning-with-dspark-numbers-under-shared-load-and-a-fix-for-cudnn-fp8-crashes/386489 "2026-10-21T13:46:51Z")
We have been running Qwen3.6-35B-A3B (NVIDIA’s NVFP4 checkpoint) and Nemotron 3.5 Lightning (NVFP4 + DSpark draft) side by side on one DGX Spark in production since September, vLLM 0.28.1 (timothystewart6/vllm-gb10 image… --- ## [Active NVIDIA API key returns 403 Authorization failed for Nemotron 3 Ultra](https://forums.developer.nvidia.com/t/active-nvidia-api-key-returns-403-authorization-failed-for-nemotron-3-ultra/386488)
**Author:** [@vaibhavjaiswar872](https://forums.developer.nvidia.com/u/vaibhavjaiswar872)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 1:46pm UTC](https://forums.developer.nvidia.com/t/active-nvidia-api-key-returns-403-authorization-failed-for-nemotron-3-ultra/386488 "2026-10-21T13:46:21Z")
Hello NVIDIA Support, I am building JAISWAR, my own neural-brain AI project, and using the hosted NVIDIA API directly. My API key is active and expires on April 21, 2027. I have verified the API configuration: Endpo… --- ## [USB Spread Spectrum Clock on Jetpack 39.2](https://forums.developer.nvidia.com/t/usb-spread-spectrum-clock-on-jetpack-39-2/385246)
**Author:** [@woods.little](https://forums.developer.nvidia.com/u/woods.little)\ **Replies:** 6\ **Last updated:** [October 21, 2026, 1:40pm UTC](https://forums.developer.nvidia.com/t/usb-spread-spectrum-clock-on-jetpack-39-2/385246 "2026-10-21T13:40:19Z")
Is there a way to enable spread spectrum clock on jetson orin nx 16GB? It currently isn’t working and we are failing USB 3.1 Gen 2 SI testing because of it. We used busybox to read the register and saw it is currently hi… --- ## [GLM-5.3-Flash: 320B total parameters / 18B active](https://forums.developer.nvidia.com/t/glm-5-3-flash-320b-total-parameters-18b-active/381350)
**Author:** [@co-le](https://forums.developer.nvidia.com/u/co-le)\ **Replies:** 881\ **Last updated:** [October 21, 2026, 1:38pm UTC](https://forums.developer.nvidia.com/t/glm-5-3-flash-320b-total-parameters-18b-active/381350 "2026-10-21T13:38:46Z")
Better than GLM-5.2 Will wait for 4bit quants, very interesting --- ## [Request to Add DeepSeek-V4.1-Flash-1017 to NVIDIA NIM](https://forums.developer.nvidia.com/t/request-to-add-deepseek-v4-1-flash-1017-to-nvidia-nim/386307)
**Author:** [@yourmomdead12](https://forums.developer.nvidia.com/u/yourmomdead12)\ **Replies:** 6\ **Last updated:** [October 21, 2026, 1:15pm UTC](https://forums.developer.nvidia.com/t/request-to-add-deepseek-v4-1-flash-1017-to-nvidia-nim/386307 "2026-10-21T13:15:29Z")
DeepSeek open-sourced DeepSeek-V4.1-Flash-1017 on October 17, 2026, under the MIT License. The weights are available on Hugging Face and ModelScope. The model permits commercial use, and the vLLM and SGLang communities a… --- ## [Flare - NVIDIA's new inference server in Rust](https://forums.developer.nvidia.com/t/flare-nvidias-new-inference-server-in-rust/386436)
**Author:** [@Digital\_David](https://forums.developer.nvidia.com/u/Digital_David)\ **Replies:** 15\ **Last updated:** [October 21, 2026, 1:06pm UTC](https://forums.developer.nvidia.com/t/flare-nvidias-new-inference-server-in-rust/386436 "2026-10-21T13:06:46Z")
NVIDIA-AI just quietly dropped its own local inference server… No announcement. Single-digit stars It’s called Flare and Apache 2.0 Built for SM120/SM121 GPUs RTX PRO 6000 Blackwell and DGX Spark PyTorch + FlashInfe… --- ## [Optimized Qwen3.8 Flash-Next on 1x RTX PRO 6000: 171 tok/s, 524K, and HiCache/NIXL persistence](https://forums.developer.nvidia.com/t/optimized-qwen3-8-flash-next-on-1x-rtx-pro-6000-171-tok-s-524k-and-hicache-nixl-persistence/381722)
**Author:** [@jpezzulli](https://forums.developer.nvidia.com/u/jpezzulli)\ **Replies:** 6\ **Last updated:** [October 21, 2026, 1:05pm UTC](https://forums.developer.nvidia.com/t/optimized-qwen3-8-flash-next-on-1x-rtx-pro-6000-171-tok-s-524k-and-hicache-nixl-persistence/381722 "2026-10-21T13:05:24Z")
For anyone interested in the local-workstation end of Qwen3.8 Flash-Next, I built and qualified an optimized SGLang runtime for the RadixArk NVFP4 checkpoint on one 96 GB RTX PRO 6000. The canonical campaign used TP=1 a… --- ## [DeepSeek-V4.1-Flash-1017 (EXL3 2.9bpw) on 2x DGX Spark with TensorFold: 57 tok/s text, 92 tok/s code, 2.1-2.3x the vLLM kit](https://forums.developer.nvidia.com/t/deepseek-v4-1-flash-1017-exl3-2-9bpw-on-2x-dgx-spark-with-tensorfold-57-tok-s-text-92-tok-s-code-2-1-2-3x-the-vllm-kit/386125)
**Author:** [@Elevatormusic](https://forums.developer.nvidia.com/u/Elevatormusic)\ **Replies:** 106\ **Last updated:** [October 21, 2026, 12:59pm UTC](https://forums.developer.nvidia.com/t/deepseek-v4-1-flash-1017-exl3-2-9bpw-on-2x-dgx-spark-with-tensorfold-57-tok-s-text-92-tok-s-code-2-1-2-3x-the-vllm-kit/386125 "2026-10-21T12:59:08Z")
DeepSeek-V4.1-Flash-1017 (EXL3 2.9 bpw, 1M context) on 2x DGX Spark, TP=2 over the CX7 cable. Recipe Github: GitHub - Elevatormusic/DeepSeek-V4.1-Flash-1017-EXL3-TensorFold-2x-DGX-Spark: DeepSeek-V4.1-Flash-1017 EXL3 2… --- ## [MiniMax M3.1 Flash comming (aka Bunny Speed)](https://forums.developer.nvidia.com/t/minimax-m3-1-flash-comming-aka-bunny-speed/386377)
**Author:** [@kafej666](https://forums.developer.nvidia.com/u/kafej666)\ **Replies:** 6\ **Last updated:** [October 21, 2026, 12:55pm UTC](https://forums.developer.nvidia.com/t/minimax-m3-1-flash-comming-aka-bunny-speed/386377 "2026-10-21T12:55:44Z")
--- ## [Windows 11 ST 2110 receive on ConnectX-6 Dx with host system clock — is BlueField-3 required?](https://forums.developer.nvidia.com/t/windows-11-st-2110-receive-on-connectx-6-dx-with-host-system-clock-is-bluefield-3-required/385961)
**Author:** [@saito-k](https://forums.developer.nvidia.com/u/saito-k)\ **Replies:** 2\ **Last updated:** [October 21, 2026, 12:39pm UTC](https://forums.developer.nvidia.com/t/windows-11-st-2110-receive-on-connectx-6-dx-with-host-system-clock-is-bluefield-3-required/385961 "2026-10-21T12:39:05Z")
Hello, We received the following note from NVIDIA Developer Program Support after our Rivermax SDK application was approved: “Windows ConnectX-6 Dx is not the right adapter. You will need to use BlueField-3.” We would… --- ## [Design Proteins with RFD4-Proteina Using Reward-Guided Inference-Time Search](https://forums.developer.nvidia.com/t/design-proteins-with-rfd4-proteina-using-reward-guided-inference-time-search/386487)
**Author:** [@jwitsoe](https://forums.developer.nvidia.com/u/jwitsoe)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 12:30pm UTC](https://forums.developer.nvidia.com/t/design-proteins-with-rfd4-proteina-using-reward-guided-inference-time-search/386487 "2026-10-21T12:30:13Z")
Originally published at: Design Proteins with RFD4-Proteina Using Reward-Guided Inference-Time Search | NVIDIA Technical Blog Protein design is a search problem. Choose amino acids for an eight-residue protein sequen… --- ## [DeepSeek v4.1 Flash](https://forums.developer.nvidia.com/t/deepseek-v4-1-flash/382725)
**Author:** [@d.piwowar](https://forums.developer.nvidia.com/u/d.piwowar)\ **Replies:** 273\ **Last updated:** [October 21, 2026, 12:23pm UTC](https://forums.developer.nvidia.com/t/deepseek-v4-1-flash/382725 "2026-10-21T12:23:37Z")
DeepSeek Flash 4.1 is just around the corner, first benchmarks available (bottom) matching GLM 5.3 flash - supposed to have new architecture, noticeably faster. Yay! Official API model ID: deepseek-v4.1-flash-expires-… --- ## [Where to obtain IMX728SL camera driver headers?](https://forums.developer.nvidia.com/t/where-to-obtain-imx728sl-camera-driver-headers/386486)
**Author:** [@bader.timalexander](https://forums.developer.nvidia.com/u/bader.timalexander)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 12:08pm UTC](https://forums.developer.nvidia.com/t/where-to-obtain-imx728sl-camera-driver-headers/386486 "2026-10-21T12:08:42Z")
DRIVE OS Version: 7.0.3 Issue Description: We are building a supplied IMX728 camera driver for DRIVE AGX Thor P3960 with DRIVE OS 7.0.3.0-41012141. The driver requires: IMX728SL\_CustomData.h IMX728SL\_CustomInterface… --- ## [Support Required – PCIe TX/RX Eye Pattern Capture on Jetson Orin Nano SOM](https://forums.developer.nvidia.com/t/support-required-pcie-tx-rx-eye-pattern-capture-on-jetson-orin-nano-som/386485)
**Author:** [@nisarg.prajapati](https://forums.developer.nvidia.com/u/nisarg.prajapati)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 12:07pm UTC](https://forums.developer.nvidia.com/t/support-required-pcie-tx-rx-eye-pattern-capture-on-jetson-orin-nano-som/386485 "2026-10-21T12:07:17Z")
Hi Support-Team, We are validating PCIe signal integrity on the Jetson Orin Nano SOM but are unable to capture proper TX and RX eye patterns. Test configurations and observations: NVMe SSD – PCIe Gen3 x4: Detected li… --- ## [Up to 70tok/s Qwen3.8-Flash-Next-Int4-AutoRound](https://forums.developer.nvidia.com/t/up-to-70tok-s-qwen3-8-flash-next-int4-autoround/382733)
**Author:** [@azampatti](https://forums.developer.nvidia.com/u/azampatti)\ **Replies:** 302\ **Last updated:** [October 21, 2026, 12:06pm UTC](https://forums.developer.nvidia.com/t/up-to-70tok-s-qwen3-8-flash-next-int4-autoround/382733 "2026-10-21T12:06:49Z")
Just like I did with my Faster Qwen3.5 122B, I’m basically doing the same with this great awesome new model from Qwen. Single-line command: git clone https://github.com/azampatti/Qwen3.8-Flash-Next-Int4-FAST.git && cd … --- ## [DeepStream 7.1 on Jetson Orin NX: Track Fragmentation and Re-ID Issues with NvDeepSORT and NvDCF](https://forums.developer.nvidia.com/t/deepstream-7-1-on-jetson-orin-nx-track-fragmentation-and-re-id-issues-with-nvdeepsort-and-nvdcf/386461)
**Author:** [@adithyaskumar235](https://forums.developer.nvidia.com/u/adithyaskumar235)\ **Replies:** 1\ **Last updated:** [October 21, 2026, 12:02pm UTC](https://forums.developer.nvidia.com/t/deepstream-7-1-on-jetson-orin-nx-track-fragmentation-and-re-id-issues-with-nvdeepsort-and-nvdcf/386461 "2026-10-21T12:02:01Z")
Dear NVIDIA Support Team, We are investigating multi-object tracking performance in NVIDIA DeepStream 7.1 on the Jetson Orin NX platform. We are experiencing track fragmentation, identity switches, and occasional instab… --- ## [ConnectX-8 MACsec Offload — Traffic drops after fixed packet count on active SA](https://forums.developer.nvidia.com/t/connectx-8-macsec-offload-traffic-drops-after-fixed-packet-count-on-active-sa/384494)
**Author:** [@tkaiser](https://forums.developer.nvidia.com/u/tkaiser)\ **Replies:** 2\ **Last updated:** [October 21, 2026, 11:48am UTC](https://forums.developer.nvidia.com/t/connectx-8-macsec-offload-traffic-drops-after-fixed-packet-count-on-active-sa/384494 "2026-10-21T11:48:22Z")
We are experiencing an issue with MACsec hardware offloading (offload mac) on ConnectX-8 where the receiver consistently drops incoming traffic after a fixed number of packets are processed on the active SA. Depending o… --- ## [WSL2 / WDDM: a vectorised (≥64-bit) global load at zero free VRAM takes the Windows host down — ~250-line pure-CUDA reproducer, one flag flips it](https://forums.developer.nvidia.com/t/wsl2-wddm-a-vectorised-64-bit-global-load-at-zero-free-vram-takes-the-windows-host-down-250-line-pure-cuda-reproducer-one-flag-flips-it/381943)
**Author:** [@stankiewiczal](https://forums.developer.nvidia.com/u/stankiewiczal)\ **Replies:** 2\ **Last updated:** [October 21, 2026, 11:48am UTC](https://forums.developer.nvidia.com/t/wsl2-wddm-a-vectorised-64-bit-global-load-at-zero-free-vram-takes-the-windows-host-down-250-line-pure-cuda-reproducer-one-flag-flips-it/381943 "2026-10-21T11:48:01Z")
WSL2 / WDDM: a vectorised (≥64-bit) global load at zero free VRAM takes the Windows host down — ~250-line pure-CUDA reproducer, one flag flips it Summary On WSL2 with GPU paravirtualisation, a CUDA workload that oversubs… --- ## [Physics instability / explosion issue when simulating a custom dual HEBI arm imported via URDF in Isaac Sim 6.2.0](https://forums.developer.nvidia.com/t/physics-instability-explosion-issue-when-simulating-a-custom-dual-hebi-arm-imported-via-urdf-in-isaac-sim-6-2-0/386382)
**Author:** [@sesetou](https://forums.developer.nvidia.com/u/sesetou)\ **Replies:** 3\ **Last updated:** [October 21, 2026, 11:47am UTC](https://forums.developer.nvidia.com/t/physics-instability-explosion-issue-when-simulating-a-custom-dual-hebi-arm-imported-via-urdf-in-isaac-sim-6-2-0/386382 "2026-10-21T11:47:25Z")
I am encountering a severe physics instability (explosion) issue when simulating a custom dual-arm robot (dual\_hebi\_arm.urdf) in Isaac Sim 6.2.0 on Ubuntu 24.04 (ROS 2 Jazzy). Issue Description: When I press the Play bu… --- ## [Thor XFI: Hardware DMA/cache synchronization issue causes stale data under concurrent traffic (r38.4.0)](https://forums.developer.nvidia.com/t/thor-xfi-hardware-dma-cache-synchronization-issue-causes-stale-data-under-concurrent-traffic-r38-4-0/386284)
**Author:** [@Ethan\_su](https://forums.developer.nvidia.com/u/Ethan_su)\ **Replies:** 7\ **Last updated:** [October 21, 2026, 11:39am UTC](https://forums.developer.nvidia.com/t/thor-xfi-hardware-dma-cache-synchronization-issue-causes-stale-data-under-concurrent-traffic-r38-4-0/386284 "2026-10-21T11:39:50Z")
During testing we observed abnormal hardware DMA and cache synchronization on Thor XFI. Under multi-stream concurrent communication, the upper layer receives stale data from the cache. Disabling hardware synchronization … --- ## [DGX Spark Software Updates - October 2026 Release](https://forums.developer.nvidia.com/t/dgx-spark-software-updates-october-2026-release/386341)
**Author:** [@aniculescu](https://forums.developer.nvidia.com/u/aniculescu)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 11:30am UTC](https://forums.developer.nvidia.com/t/dgx-spark-software-updates-october-2026-release/386341 "2026-10-21T11:30:48Z")
DGX Spark software updates began rolling out today, as a part of our October 2026 Release, with the new features added below. (Partner GB10 Grace Blackwell Superchip systems will offer the same updates in their next sof… --- ## [403 Authorization Failed Public API Endpoints access for Personal Organization](https://forums.developer.nvidia.com/t/403-authorization-failed-public-api-endpoints-access-for-personal-organization/386480)
**Author:** [@sarthakjain2006to2016](https://forums.developer.nvidia.com/u/sarthakjain2006to2016)\ **Replies:** 0\ **Last updated:** [October 21, 2026, 11:19am UTC](https://forums.developer.nvidia.com/t/403-authorization-failed-public-api-endpoints-access-for-personal-organization/386480 "2026-10-21T11:19:34Z")
Account type: Personal GET /v1/models: 200 OK POST /v1/chat/completions: 403 Authorization failed Tested models: moonshotai/kimi-k2.6 openai/gpt-oss-20b Environment: Windows / PowerShell Direct API request (not O… [Next page](https://forums.developer.nvidia.com/latest.md?page=1)