Microsoft unveils Nvidia-powered Surface AI PCs and dev workstation
Microsoft announced pricing and specifications for its Surface Laptop Ultra and Surface RTX Spark Dev Box, which use Nvidia's RTX Spark chips to run AI models locally. The laptops start at $2,600 and $3,700, while the dev workstation starts at $6,000, with higher configurations available. A revamped Windows 11 adds Execution Containers for sandboxing AI agents, and Microsoft is offering trade-in credit for MacBook Pro users.
Nvidia first said in June it had reached deals with Microsoft and other manufacturers to build Windows PCs tuned for AI and agents around its RTX Spark processor, though few product details emerged then. At a San Francisco Tech Week gathering on Wednesday, Microsoft filled in pricing and specs for its Surface Laptop Ultra, with two starting configurations at $2,600 and $3,700, rising to $5,900; the top model has sold out.
Microsoft also introduced the Surface RTX Spark Dev Box, beginning at $6,000 and bundling developer tools such as VS Code, GitHub Copilot CLI, WSL, and PowerShell 7. Both systems target local AI model execution, and Microsoft says a refreshed Windows 11 will add Execution Containers for sandboxing agents. Dell separately opened preorders for its XPS 16 Creator Edition at $3,800.
Count? First para: Nvidia(1) first2 said3 in4 June5 it6 had7 reached8 deals9 with10 Microsoft11 and12 other13 manufacturers14 to15 build16 Windows17 PCs18 tuned19 for20 AI21 and22 agents23 around24 its25 RTX26 Spark27 processor28, though29 few30 product31 details32 emerged33 then34. At35 a36 San37 Francisco38 Tech39 Week40 gathering41 on42 Wednesday43, Microsoft44 filled45 in46 pricing47 and48 specs49 for50 its51 Surface52 Laptop53 Ultra54, with55 two56 starting57 configurations58 at59 $2,60060 and61 $3,70062, rising63 to64 $5,90065; the66 top67 model68 has69 sold70 out71. That's 71. Second: Microsoft1 also2 introduced3 the4 Surface5 RTX6 Spark7 Dev8 Box9, beginning10 at11 $6,00012
These high-priced machines could mainly reach well-funded developers, studios, and AI-focused teams, potentially speeding local model experimentation while raising concerns about who can afford cutting-edge hardware. On-device AI may improve privacy and reduce cloud reliance, though performance and battery trade-offs remain uncertain. The MacBook Pro trade-in offer may nudge some developers toward Windows, while Execution Containers could make agent deployment safer—or add complexity. Broader consumers may see trickle-down features later, but near-term impact may stay concentrated among professionals. Count: These1 high-priced2 machines3 could4 mainly5 reach6 well-funded7 developers8, studios9, and10 AI-focused11 teams12, potentially13 speeding14 local15 model16 experimentation17 while18 raising19 concerns20 about21 who22 can23 afford24 cutting-edge25 hardware26. On-device27 AI28 may29 improve30 privacy31 and32 reduce33 cloud34 reliance35, though36 performance37 and38 battery39 trade-