Explaining new laptop's powerful techfrom @thespenceryan

Transcript

This is the jack of all trains. After a week of trying to break the Surface Laptop Ultra, I think we need to stop pretending that Rank can't handle real work and admit that Microsoft is cooking because this is a triple threat. So for starters, we've never paired ARM efficiency with a full NVIDIA Blackwell RTX GPU in a laptop before. But what does that even mean? Well, in simple terms, ARM chips are the quiet marathon runners. They sip battery all day, and the RTX GPU is the sprinter. That's the one that renders three d, runs AI, and games. This laptop just glued those two together into one single chip called the RTX Spark. You see, in a normal laptop, the CPU has its own Rank, and the GPU has VRAM. They're two separate pools, which is why your fancy gaming rig still can't load the big AI models. Here, they share one single pool, up to a 128 gigs of memory. It's like a kitchen with a giant pantry ready for whatever you need to cook. So what does this actually get you and I? Well, in my reveal video, I got a lot of critical comments, so let's kinda distinguish the myth versus reality here. So for starters, this is not going to throttle or overheat or melt your table. Although it's 18 millimeters thin, the thermal architecture has been designed from the ground up. It's built for sustained power, not just a burst and throttle. And speaking of throttle, this does not throttle on battery power. But then it's like, who cares? Right? It's armed. It can't run most apps or games. Well, Windows has Prism emulation. Prism used to fit the emulation is slow and trash stereotype, but now we have triple a games hitting high settings, and the performance penalty is as low as 2%. And, of course, native ARM games and apps run at full speed. All of this is paired with DLSS and the power of CUDA. Now as a creator, I'm running my real worse, DaVinci Resolve, Fusion, heavy timelines, three d scenes, and it's doing great. The GPU is doing all the heavy lifting, and the CPU is keeping everything responsive. This is a workstation in an 18 millimeter chassis with a unified memory pool. And that's exactly why I was able to move my entire Claude workflow locally offline. And that's a mental shift because most people assume AI means sending your data to a server and hoping it doesn't leak. This runs on machine. This is not a casual buy, and my benchmarks even put it in the neighborhood of an m three, m three Max for CPU scores. GPU is much closer to a fifty seventy. So it's safe to say that this is not a laptop for everybody. You probably shouldn't even buy it. It's a laptop for people with serious workstation needs. It is a jack of all trades. And if that's not you, that's fine. You don't need to get this. But the Rank can't handle real work chapter, that chapter is officially closed.