Surface Laptop Ultra review: A different kind of MacBook Pro

After launching Microsoft's Surface Laptop Ultra, I spent five minutes looking at it. I wasn't surprised by the MacBook Pro-like design or the bright display. I was trying to figure out what I was actually doing with a $5900 laptop. This is not just an ordinary Windows laptop. Microsoft in collaboration with […]

This is not just any ordinary Windows laptop. Microsoft has teamed up with Nvidia to build a high-end laptop to showcase Nvidia's new Arm-based RTX Spark chips. This is a big moment for Windows on Arm, as these chips can run demanding AI models alongside the latest AAA games and creative applications. The promise is compelling, but the reality of local AI is much messier.

Getting the most out of the Surface Laptop Ultra, especially my top-of-the-line model, requires coding experience or the patience to experiment with AI models and agents until you find something that works. If that sounds like you, Microsoft may just have created the Windows laptop you've been waiting for. Otherwise, it may not be the MacBook Pro competitor Microsoft was hoping for.

[Product: Microsoft Surface Laptop Ultra | ptop-ultra-15-3k-touchscreen-nvidia-rtx-spark-n1x-18-core-2026-24gb-memory-512gb-storage-platinum%2FJJGXP3CQW2,

It looks like Microsoft has finally built hardware to compete with Apple's professional laptops, thanks to its all-aluminum design, large touchpad, and high-end finish. Even the "night" color is just like Apple's space black.

Although there are many significant differences. The 15-inch MicroLED display can handle a maximum HDR brightness of 2000 nits at 262 pixels per inch (3270 x 2180). MacBook Pro reaches 1600 nits. This makes it the brightest screen Microsoft has ever offered on a Surface Laptop. Unlike the MacBook Pro, it also has a 3:2 touchscreen, which is great for writing code, reading text, or editing photos.

The port selection is similar to the MacBook Pro, with one big change. You get three USB-C ports, one USB-A port, one HDMI 2.1b port, an SD card reader, and a headphone jack. Where it's really different is Microsoft's new magnetic connector. Just like the old Surface Connect, but it works through the USB-C port on the right side. Just slide the connector close to the port and it snaps into place. When not charging, that port works just like any other USB-C port on a laptop.

Magnetic Connect is one of my favorite things about the Surface Laptop Ultra. I love how you can reverse the USB-C magnetic charging end, plug it into a power brick, and use the regular USB-C connector on the opposite end of the cable to charge the rest of your devices.

Touch feedback on the big screen is also welcome. It works with things like clicks, drags, and aligning windows across Windows 11, and touch feedback is enabled in apps like PowerPoint when you drag and align objects or images in a deck. It's subtle enough to make it feel like you're actually moving windows or interfaces, but I wish more apps supported it.

Microsoft has placed the 1080p front camera above the screen. Quality is good for video calls, and it supports Windows Hello facial authentication and Windows Studio effects like auto-framing, eye contact, and background blur.

Not all of the Surface Laptop Ultra's hardware is impressive, though. I've had issues where I unplugged the charger at 100% for Windows only to immediately report 97% battery life. Plugging it in would bring it back to 100% and then every time I unplugged it would immediately drop back to 97%.

[Image: The USB-C magnetic charger is a notable addition.

It seemed odd, so I asked Microsoft about this behavior, and it's designed to make the battery last longer. "When the device is plugged in and the battery is full, the device stops charging instead of continuously charging and only resumes when the level drops to 95 percent," said Microsoft spokesperson Moses Nserko. "This reduces wear and tear. When plugged in, the indicator shows 100 percent to indicate full charge. "When you disconnect the power, it immediately shows the exact level of the battery, which may be a few percent lower."

I've also had issues with the display, where things either seemed a bit laggy or there was significant ghosting. Moving the window around quickly sometimes leaves a small trace. Changing the power settings from the default "Best Energy Efficiency" to "Best Performance" seems to fix most of the lag, and the backlight ghosting seems to be less noticeable the longer you run the laptop.

I also heard a loud, booming sound from the Surface Laptop Ultra's built-in speakers. They were loud enough to sound like firecrackers, and I thought the battery was dying before I realized it was just speaker interference. Rebooting the Surface Laptop Ultra fixed it, but it's happened a few times now.

As for reboots, I found the Surface Laptop Ultra to be surprisingly slow to reboot. It seems like this is something that Microsoft can easily fix with an update.

[Image: Surface Laptop Ultra can run complex local AI models.

RTX Spark, which powers the Surface Laptop Ultra, is designed with local AI models and agents in mind. The version I tested has a 20-core CPU co-designed by Nvidia and MediaTek with a 6144-core CUDA GPU – roughly comparable to a mobile RTX 5060 – plus 128GB of integrated memory and a 1TB Gen 5 SSD. This is the top of the line laptop at $5,899 and I think you have a very specific use case in mind if you're considering this particular laptop spec. The base version costs $2,599 and has an 18-core processor, a 5120-core CUDA GPU, and 24GB of integrated memory.

Out of the box, Microsoft requires Laptop Ultra owners to choose an AI agent. You can choose between Hermes, OpenClaw and Perplexity. I tried Hermes and OpenClaw last week, but at first I had problems starting OpenClaw. Reinstalling fixed a gateway issue and then I could name my agent and start using it for local AI work.

I configured Hermes and OpenClaw to use the Qwen3.8-27B model, one of the most powerful local AI models. Because it uses all the power of the Surface Laptop Ultra, it boosts its fans with throttle and 100% GPU for even the most basic queries. I asked Hermes to resize some photos and convert them from PNG to JPG, which takes seconds using the PowerToy tool. It took several minutes to complete, all while the fans whirred and the laptop heated up.

It's probably my fault for using such a large model for such a basic task, but without testing every single local AI model, it's hard to know which one to choose for every task you need. I asked OpenClaw to create a smart home dashboard for me, but it took 30 minutes of spinning the laptop, fanning and calling various tools and finally producing nothing but an error.

I don't have a basic Surface Ultra laptop to test with, but it's easy to see that the 24GB of integrated memory means fewer local AI models.

[Image: Local AI models are currently the best for coding tasks.

I tried to find compelling use cases for the power of these chips for local AI tasks. By using a free cloud model, many tasks can be done faster and without spinning laptop fans. The power here is really in making the agent work at full capacity to build something overnight, or run a task on a large data set that usually costs a lot in cloud tokens.

I liked that I could do things locally, but without worrying about the privacy aspects of cloud models. I gave Hermes some audio and video files and it went ahead and created a local llama.cpp server to enable audio transcription. It took a while to get it working, but it sure saves me from having to install transcription apps or pay for a cloud service like Otter.

What I didn't like is that these agents are far from user-friendly, and I'm surprised Microsoft or Nvidia haven't developed their own native AI interface that would simplify a lot of this. I need to spend more time with both Hermes and OpenClaw, but as a non-coder it's clear that this is all really aimed at developers and power users.

The Surface Laptop Ultra may not be an official gaming laptop, but it can certainly run the latest games much better than Apple's MacBook Pro. Due to the poor GPU in the Qualcomm Snapdragon X series chips, I usually get nervous about Windows on Arm and games. But this is the first time we've seen an Arm processor paired with a beefier Nvidia GPU, and most games run well for a laptop that isn't quite a gaming laptop.

Instead of running a bunch of benchmarks at different settings, I decided to find the optimal settings that would allow me to play every game at 60fps or more. Since this is an RTX GPU, you also get full support for DLSS, Frame Generation, and ray tracing. So if you want 120fps to match a 120Hz display, you can enable Frame Generation.

I was able to run Black Myth: Wukong at 70fps at native resolution (3270 x 2180) with medium settings and DLSS performance mode enabled. I ran the same medium settings and DLSS performance in Cyberpunk 2077 and averaged 66 fps, and Forza Horizon 6 managed an average of 57 fps.

Most games should run at medium settings and DLSS or Balanced performance to get around 60fps, but there are exceptions where you should lower the settings or resolution. Gears of War: E-Day averaged 40 fps on medium settings and DLSS performance, and I was only able to hit 48 fps on low settings. I also had to move Doom: The Dark Ages to low settings and drop the DLSS performance to an average of 60fps.

[Image: Gears of War: E-Day running on a Surface Ultra laptop.

I use the RTX 5090 every day on a 32-inch 4K display with maximum gaming settings, but I didn't mind the appearance of medium settings on a 15-inch display. It was more like falling into console settings, and in some games it's hard to tell the texture differences on the smaller screen. I'm one of those gamers that would rather run at 60fps on lower settings than max out and have to play at 30fps.

While game compatibility is good for most titles, there are some where you will run into emulation issues. Horizon Zero Dawn Remastered was limited to 30fps at 1440p regardless of whether I was using low or medium settings. I believe this is because the game is CPU heavy, so a native Arm port would probably help for this type of game.

However, there are no native Arm games. I tried the new version of Arm Marvel Rivals and the average speed was 64 fps with medium settings and DLSS performance and 75 fps with low settings. I think we'll only see Arm ports in cases where there's a lot of CPU activity that the Prism emulator - which translates CPU instructions from x86 to Arm - can't handle well enough, as it's otherwise very capable.

We may see more Arm ports for games that have anti-cheat engines. Emulating anti-cheat engines is much more complex, so developers need to create Arm-compatible versions. While games like Arc Raiders, Counter-Strike 2, Dota 2 and Fortnite all work out of the box, Nvidia and Microsoft are working with EA, Activision and Riot Games to fully support titles like Battlefield, Call of Duty, Valorant and League of Legends.

Surface Laptop Ultra is also a very capable device for creative and everyday work. I managed to export my test 4K video in two minutes and 18 seconds on the Ultra laptop using Premiere Pro. That's about a minute behind the high-end MacBook Pro, but it beats the M5's base chip and is about 30 seconds behind the M5 Pro. I had to create this using Premiere Pro's export tool, rather than sending it to Media Encoder like we normally run this test. There seems to be a bug in the beta version of Arm Media Encoder that prevents the export from running correctly.

Where the Surface Laptop Ultra really falls behind Apple's M5 chips is CPU performance. Single-core and multi-threaded CPU tests such as Geekbench or Cinebench show that the MediaTek CPU is far inferior to the $6,999 MacBook Pro 16 with the M5 Max or even the M6 Mac Mini. Even some of the GPU tests for Geekbench show the RTX Spark falling behind Apple's MacBook Pro, with Puget Bench Photoshop, Premiere Pro, and DaVinci Resolve all noticeably behind the M5 Max.

I also tested the Surface Laptop Ultra's battery life, which seems to be around seven hours if you're using local AI models and a mix of apps and games. If you're doing less intensive work, it's an average of 10 hours or more. The more I use the Laptop Ultra, the more I'll know, so check back for an updated battery display.