… Specifically, this is Microsoft’s answer to the recent spate of Linux-based AI development boxes that have been released by both NVIDIA and AMD partners over the last year, where both NVIDIA’s GB10 and AMD’s Ryzen AI Max Strix Halo have been showing up in all sorts of developers kits. …
… Today at Microsoft Build, we are introducing Surface RTX Spark Dev Box, a compact developer PC engineered with NVIDIA RTX Spark superchip and built on the Windows developer platform, designed for local-first AI development. …
… The Surface RTX Spark Dev Box is built on the Windows developer platform and powered by NVIDIA's RTX Spark superchip. …
Microsoft has two new Surface devices arriving later this year, both powered by Nvidia’s RTX Spark chips. …
… On the hardware front, we didn’t get any updates for existing Surface devices not counting yesterday’s Surface Laptop Ultra announcement , but we did get something new: the Surface RTX Spark Dev Box is “a compact developer PC” built around Nvidia’s new RTX Spark chip with up to 128GB of built-in me… …
… While the latter does remain the first choice for a lot of AI developers and the only choice for rival NVIDIA’s DGX Spark boxes , AMD believes there are enough AI developers and users running Windows as their daily driver to warrant supporting both at parity. …
… The new machine is powered by NVIDIA's RTX Spark superchip, combining a Grace CPU and Blackwell-based RTX GPU into a unified platform aimed at accelerating modern AI workflows. …
… In an interview with my colleague Alaina Yee, Andrew Hill, corporate vice president of Surface for Microsoft, confirmed that consumers will be able to buy the Surface RTX Spark Dev Box, the mini PC with Nvidia's latest RTX Spark chip inside. …
… The primary audience for this device would be local AI developers, as AMD has placed its most powerful APU in a compact box to provide an alternative to NVIDIA's DGX Spark platform. This should be sufficient for good compute performance and local AI development. …
There is a lot of disdain for DGX Sparks here on the sub. And I get it. A lot of people say “It could have been great if it had been better memory bandwidth”, “SM-121 is a fake /second-class Blackwell chip” yadda, yadda.…
I'm a big fan of on-device AI inference for a million reasons, especially its potential to significantly reduce or even potentially eliminate the need for massive AI data center projects in the United States. But so far,…