In addition to these relatively low-power platforms, Nvidia will also be producing Windows on Arm-compatible versions of its DGX Station high-end desktop PC. The DGX Station is built around the GB300 Superchip, which encompasses a 72-core Grace CPU with 496 GB of LPDDR5X memory paired with a Blackwell Ultra GPU offering 252GB of HBM3e and up to 15 PFLOPS of FP4 performance without sparsity. Developers can further expand this system with another RTX Pro GPU over PCI Express.
Nvidia also committed to future DGX Stations for high-end Windows AI workstation performance. Although those systems will doubtless be far lower volume products than RTX Spark laptops and desktop mini-PCs, it further cements the company's full commitment to creating a reliable and durable ecosystem for partners to build around and consumers to buy into.
We're expecting to learn more about Nvidia's RTX Spark systems and ecosystem this week at Computex 2026. Stay tuned for more details.
Follow Tom's Hardware on Google News , or add us as a preferred source , to get our latest news, analysis, & reviews in your feeds.
As the Senior Analyst, Graphics at Tom's Hardware, Jeff Kampman covers everything that has to do with graphics cards, gaming performance, and more. From integrated graphics processors to discrete graphics cards to the hyperscale installations powering our AI future, if it's got a GPU in it, Jeff is on it.\u00a0 ","collapsible":{"enabled":true,"maxHeight":250,"readMoreText":"Read more","readLessText":"Read less"}}), "https://slice.vanilla.futurecdn.net/13-4-24/js/authorBio.js"); } else { console.error('%c FTE ','background: #9306F9; color: #ffffff','no lazy slice hydration function available'); } Jeffrey Kampman Senior Analyst, Graphics As the Senior Analyst, Graphics at Tom's Hardware, Jeff Kampman covers everything that has to do with graphics cards, gaming performance, and more. From integrated graphics processors to discrete graphics cards to the hyperscale installations powering our AI future, if it's got a GPU in it, Jeff is on it.
das_stig Nvidia need to get right, 1. Price, they have the opportunity to decimate the market in their favour, if the make them cheap as they did in golden age of PC explosion and in volume and that includes the actual OEMs making the completed units. 2. Performance, doesn' t matter how fast they make the chips, if performance under WinArm native or Winx86 Emulated is worse than now. 3. Power, it has to sip juice and give long battery lives and gain ties in to performance. 4. No user lockout stupidity, by all means secure firmware against nasties but leave it open for the user to decided how and what to run, it's their property ! 5. Engage with communities that will popup to spur development and acceptance. Reply
Key considerations
- Investor positioning can change fast
- Volatility remains possible near catalysts
- Macro rates and liquidity can dominate flows
Reference reading
- https://www.tomshardware.com/pc-components/cpus/SPONSORED_LINK_URL
- https://www.tomshardware.com/pc-components/cpus/nvidia-unveils-dgx-sparrk-roadmap-for-laptops-and-desktop-pcs-at-computex-2026-three-generations-outlined-rubin-followed-by-rosa-feynman#main
- https://www.tomshardware.com/subscription
- A robot startup is wreaking havoc on short-term rentals in San Francisco — Airbnb hosts allege 'guests' secretly tested robots indoors, left the units completel
- Resourceful runner 'can race my own ghost' using homemade Meta Ray-Ban Display app — also adds bonus coins, mini leaderboard, and more
- AMD Ryzen 7 5800X and Nvidia RTX 5060 combo is only $439 — deal includes a free CPU cooler, totaling out to $150 in savings for a great budget setup
- Intel Xeon 7 ‘Diamond Rapids’ CPUs officially launching in 2027 on Intel 18A-P — next-gen P-core Xeon features PCIe 6.0, 50% higher core counts, and twice the m
- NVIDIA, Ineffable Intelligence Team Up to Build the Future of Reinforcement Learning Infrastructure
Informational only. No financial advice. Do your own research.