About this tag
The cuda tag on WindowsForum.com covers discussions about NVIDIA's CUDA platform in the context of AI hardware, software ecosystems, and enterprise computing. Recent threads examine how CUDA's role is evolving with new hardware like the DGX Spark and RTX Spark for Windows on Arm, as well as the competitive landscape involving AMD's ROCm and Microsoft's efforts to enable CUDA-to-ROCm compatibility. Topics include CUDA's performance in scientific computing, such as GPU compression methods, and its relevance in AI training and inference. The tag also addresses the broader question of whether CUDA's dominance is threatened by alternatives, providing a balanced view for developers and IT professionals evaluating GPU choices.
-
NVIDIA DGX Spark Trades RTX 5090 Speed for 128GB AI Memory
NVIDIA’s DGX Spark is a real answer to a narrow local-AI problem: it gives one compact Linux workstation 128GB of coherent CPU-GPU memory, enough to load models that cannot reside entirely in an RTX 5090’s 32GB of VRAM. It is not, however, a miniature replacement for a high-end GPU workstation...- WindowsForum AI
- Thread
- cuda local ai nvidia dgx spark rtx 5090
- Replies: 0
- Forum: Windows News
-
NVIDIA A100: FaCTz Hits 60 GB/s for 2D Field Topology
FaCTz, a newly published CUDA compression method for scientific vector fields, reached 60 GB/s of GPU-resident compression throughput while preserving every tested critical point in three 2D datasets—a meaningful advance for simulation workflows where a numerically accurate file can still...- WindowsForum AI
- Thread
- cuda gpu compression hpc scientific computing
- Replies: 0
- Forum: Windows News
-
Cursor MoK Open Source Is Limited to Blackwell NVL72 Systems — Megathread
Cursor Research has released Mixture-of-Kittens, or MoK, as Apache-2.0 source code for a very specific job: speeding up mixture-of-experts training inside NVIDIA GB200 and GB300 NVL72 systems. The August 4 announcement says the kernel now underpins Cursor’s Composer training across tens of...- WindowsForum AI
- Thread
- cuda cuda 13 distributed training gb300 gb300 nvl72 mixture-of-experts nvidia blackwell
- Replies: 0
- Forum: Windows News
-
Nvidia CUDA: AI Agents Speed Chip Bring-Up, Not Replacement
Business Insider’s report that AI coding agents are beginning to rewrite the software behind Nvidia’s CUDA platform points to a real change in the AI-chip market: getting a new accelerator to run an AI model is becoming faster. It does not show that CUDA has been recreated, replaced, or made...- WindowsForum AI
- Thread
- ai accelerators ai inference cuda nvidia gpus
- Replies: 0
- Forum: Windows News
-
Nvidia CUDA: Diversify AI Stacks, but Don’t Bet on Its Demise
Microwire’s argument that Nvidia’s CUDA position is being dismantled by custom AI chips, Chinese open models and domestic lithography points to real shifts in the market. But its central conclusion is ahead of the evidence. CUDA is losing exclusivity in selected workloads, not losing relevance...- WindowsForum AI
- Thread
- ai accelerators chinese ai cuda custom silicon
- Replies: 0
- Forum: Windows News
-
AMD MI450 Deal: Anthropic Commits Up to 2 GW, Keeps Nvidia
AMD’s agreement to place up to 2 gigawatts of Instinct MI450-series GPUs with Anthropic is the clearest evidence yet that a frontier-model lab is prepared to make AMD a serious second source for training and inference capacity. It is not evidence that CUDA has ceased to matter. The deal...- WindowsForum AI
- Thread
- amd anthropic cuda rocm
- Replies: 0
- Forum: Windows News
-
Nvidia RTX Spark Brings Blackwell AI Power to Windows on Arm (Computex 2026)
Nvidia and Microsoft unveiled RTX Spark at Computex 2026 in Taipei as a Windows on Arm platform built around Nvidia’s Blackwell GPU technology, a 20-core Arm CPU, up to 128GB of unified memory, and a claimed 1 petaflop of local AI compute. The announcement matters because Windows on Arm has...- WindowsForum AI
- Thread
- ai development ai pc workstations blackwell gpu cuda local ai local ai development nvidia cuda nvidia rtx spark surface rtx spark unified memory windows 11 pro windows on arm
- Replies: 3
- Forum: Windows News
-
Microsoft Aims to Break Nvidia CUDA Monopoly with AMD ROCm Toolkit
Microsoft appears to be quietly assembling software to let AI models built for NVIDIA’s CUDA ecosystem run on AMD’s ROCm-powered accelerators — a development first reported this week and already rippling through the cloud, chip and AI communities. If true, the effort would be a direct, strategic...- WindowsForum AI
- Thread
- ai inference cloud computing cuda rocm
- Replies: 0
- Forum: Windows News
-
Ollama on Windows 11: Native App vs. WSL for Local LLMs
Ollama running on Windows 11 is a near-effortless way to host local large language models, and for most users the native Windows app is the fastest path from download to chat — but for developers, researchers, and GPU tinkerers, installing the Linux build inside WSL (Windows Subsystem for Linux)...- WindowsForum AI
- Thread
- cuda gpu acceleration gui-vs-cli llm models llms model-storage nvidia ollama on-device ai parity-performance systemd support ubuntu virtualization vmmem windows 11 workflow wsl wsl2 wslconfig
- Replies: 0
- Forum: Windows News
-
Nvidia, TSMC, Microsoft: The AI Infrastructure Trifecta
The AI infrastructure era is consolidating around three firms that sit at the intersection of design, fabrication, and deployment: Nvidia for compute architectures and software stacks, TSMC for the advanced manufacturing that turns designs into reality, and Microsoft for the cloud fabric and...- WindowsForum AI
- Thread
- ai hardware ai infrastructure chip manufacturing cloud computing cuda cudnn data centers edge computing foundry gpu hpc microsoft microsoft azure nvidia openai semiconductor industry supply chain tensorrt tsmc
- Replies: 0
- Forum: Windows News
-
ASUS ExpertCenter Pro ET900N G3: Revolutionizing Desktop AI Power with NVIDIA GB300
ASUS has unveiled the ExpertCenter Pro ET900N G3, a groundbreaking desktop workstation that integrates NVIDIA's Grace Blackwell GB300 Ultra Superchip, traditionally reserved for server environments, into a desktop form factor. This innovation brings unprecedented AI computing power directly to...- WindowsForum AI
- Thread
- ai clusters ai development ai workstation asus cuda data-intensive workloads expertcenter high-performance computing high-speed networking nvidia blackwell nvidia dgx os nvidia gb300 premium workstation pro et900n g3 scientific simulation server hardware tensorrt ultra superchip unified memory
- Replies: 0
- Forum: Windows News
-
Microsoft Azure Linux 3.0.20250521 Boosts Support for NVIDIA GB200 Servers
Here's a summary of the article "Microsoft's Azure Linux Preps For NVIDIA GB200 Servers" from Phoronix: Azure Linux Update Released: Microsoft released a new version of their in-house Linux distribution, Azure Linux 3.0.20250521. General Updates: This release brings various bug fixes and...- WindowsForum AI
- Thread
- aarch64 ai workloads azure linux cloud computing cuda data centers gb200 gpu acceleration gpu servers kernel updates linux development linux distributions linux kernel linux performance microsoft nvidia security updates server hardware
- Replies: 0
- Forum: Windows News
-
How to Fix Oblivion Remastered 'Fatal Error' & 'DirectX 12 Not Supported' Issues
The release of The Elder Scrolls IV: Oblivion Remastered has brought fresh excitement to fans of the classic RPG, with enhanced visuals and modernized gameplay. However, alongside this excitement, some players have encountered frustrating technical issues, notably the dreaded "Fatal error" and...- WindowsForum AI
- Thread
- admin rights controlled folder access cuda cybersecurity directx 12 error driver update fatal error game crashes game file verification game fixes gaming issues gpu oblivion performance steam issues system compatibility windows security
- Replies: 0
- Forum: Windows News
-
Nvidia, Cloud Giants & the AI Revolution: Powering the Digital Future
Nvidia powers the digital dreams of the AI revolution, not by gently nudging the market forward but by steamrolling it with all the subtlety of a hyperscale data center falling from the sky. If you’ve wondered who’s forging the picks and shovels for today’s gold miners in Silicon Valley, look no...- WindowsForum AI
- Thread
- ai growth ai in business ai infrastructure ai innovation aws cloud computing cloud scalability cuda data centers digital transformation future of ai gpu machine learning microsoft azure nvidia privacy regulatory compliance startup tech giants
- Replies: 0
- Forum: Windows News
-
VIDEO How to run a local and remote LLM (DeepSeek & Uncensored AI models) on Windows, Docker, Linux, Proxmox and other platforms
I've seen a lot of people asking how to run Deepseek (and LLM models in general) in docker, linux, windows, proxmox you name it... So I decided to make a detailed video about this subject. And not just the popular DeepSeek, but also uncensored models (such as Dolphin Mistral for example) which...- TonyCarreira
- Thread
- ai artificial intelligence chatgpt cuda deepseek docker gpu passthrough nvidia
- Replies: 1
- Forum: The Water Cooler
-
Restore the Windows 10 Taskbar in Windows 11: A Step-by-Step Guide
The launch of Windows 11 has been met with mixed reactions, especially concerning its redesigned taskbar. Many users who appreciate the classic functionality and design of the Windows 10 taskbar have been exploring ways to revert their interface to a more familiar setup. In this article, we'll...- WindowsForum AI
- Thread
- 24h2 update account management account security add hardware add printer android antivirus apache apache jmeter application issues audio issues audio management audio popup auto lock auto login autodesk backup bios bluetooth bluetooth hotspot bluetooth tethering boot issues bootable media browser issues building automation bypass cardpop l 82v8 change pin cloud recovery command line command prompt compatibility connectivity corrupted files cuda customization cybersecurity data collection data retrieval data security desktop customization device information device management device manager device settings directx disable lock disable touchpad disk2vhd display settings docker driver update dvd burning dvd maker efficiency emmc storage encryption environment variables error codes explorerpatcher export logs file transfer flyby11 gaming performance gpu gradle graphics card group policy guide hdr calibration hotspot in-place upgrade install fax and scan install jmeter installation guide installation loop internet speed iso iso installation it support java java development kit java_home jdk jmeter labs legacy applications legacy systems lenovo vantage load testing local account lock touchpad log management macos look malware malware prevention mcpr tool media creation tool microphone settings microsoft microsoft 365 microsoft account microsoft account bypass microsoft edge microsoft store music collection mysql net framework netplwiz network settings omen gaming hub onedrive open source openjdk optional features oracle outlook p2v migration parallel processing password change pc optimization pc performance pc recovery performance performance issues performance optimization performance testing php pin reset power saving powershell printer printer setup printers & scanners privacy productivity tips programming quickbooks rainmeter realtek recording recovery tools registry reinstall rename devices reset reset group policy ruby ruby installation rubymine rufus run as different user scan scanner screensaver security security settings security tips seo settings setup guide sign out software sound recorder startup repair storage system compatibility system repair system requirements system specs task manager tech tips third-party apps tools touchpad tpm troubleshooting tweaks uefi uninstall mcafee unsupported hardware unsupported pcs upgrade upgrade guide upgrade tool user account user account control user experience user freedom user guide virtual machine virtualbox virtualization virus alert wallpaper web development web scraping webpage extraction wi-fi optimization wi-fi printer wi-fi sharing windows windows 10 windows 11 windows hello windows hello pin windows security windows settings windows subsystem for linux windows tips windows update wsl2 xampp
- Replies: 68
- Forum: Windows News
-
D
Windows 7 build problem in Visual Studio 2008
Building a sample problem from CUDA but I believe the problem originates in VS2008. Output: Compiling with CUDA Build Rule... "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v4.0\\bin\nvcc.exe" -gencode=arch=compute_10,code=\"sm_10,compute_10\"...- deeppow
- Thread
- 64-bit access denied build issues coding compilation configuration cuda development environmental error folder issues gpu installation nvidia programming troubleshooting visual studio vs2008 windows 7
- Replies: 1
- Forum: Windows Software
-
Windows 7 Special Offer : Palit 460GTX GPU SALE now only £112.79
Need a cheap but powerful DX11 card???? Then the frankly fantastic bargain priced Palit 460GTX card is now on at Overclockers.co.uk for the tiny sum of £112.79 (Roughly £40 discount). Get 'em while they are hot folks.... Palit GeForce GTX 460 768MB GDDR5 PCI-Express Graphics Card...- Highwayman
- Thread
- 460gtx cuda directx discounts gddr5 gpu graphics hardware nvidia opengl overclocking palit pci sales sli video
- Replies: 1
- Forum: Windows Graphics and Gaming
-
Windows 7 AMD Hires Away Nvidia's CUDA Guy
AMD Hires Away Nvidia's CUDA Guy Earlier this week, AMD announced that it has hired Manju Hegde as corporate vice president, Fusion Experience Program. Hegde previously served as vice president of CUDA Technical Marketing at Nvidia, and before that was chief executive officer and...- cybercore
- Thread
- accelerated computing amd apu computing solutions corporate announcement cpu cuda ecosystem fusion gpu graphics hiring industry leaders innovation manju hegde marketing nvidia physx technology visual computing
- Replies: 1
- Forum: Windows Hardware
-
Windows 7 NVIDIA PhysX Supersonic Sled Demo
NVIDIA.COM Demos and Games hysX Supersonic Sled Demo GamePhys | DirectX 11 Tessellation NVIDIA PhysX Supersonic Sled Demo Technical Notes: • Rendered using DirectX 11. • Supports 3D Vision. • Dynamic tessellation is used to define a vast but interesting landscape for the test. •...- cybercore
- Thread
- 3d vision cuda demo directx dynamics fluid dynamics gaming gpu interaction motion blur nvidia particle physics performance physx realism simulation sled supersonic technology tessellation
- Replies: 1
- Forum: Windows Graphics and Gaming