Nvidia’s RTX Spark Arm Chip Takes Aim at Apple’s M5 at Computex

Jensen Huang just turned the PC chip war into a four-way knife fight.

Nvidia unveiled its RTX Spark Arm-based laptop chip at Computex 2026, openly targeting Apple’s M5, Intel, AMD, and Qualcomm. The superchip packs 20 CPU cores plus 6,144 Blackwell CUDA cores, with integrated graphics that MSI reps say rival a desktop-class RTX 5070 GPU.

Picture this: it’s a humid June morning in Taipei, the Computex floor is packed, and Jensen Huang strolls out in his trademark leather jacket to announce that Nvidia, the company everyone associates with AI training clusters and gaming GPUs, is now coming for the laptop processor itself. Apple, Intel, AMD, Qualcomm. All of them. At once.

What Happened

Nvidia announced the RTX Spark, a custom Arm-based laptop chip co-designed with MediaTek, set to power a new wave of Windows 11 laptops. The company is positioning the launch alongside Microsoft and Arm as “a new era of PC.” It’s the first time Nvidia has stepped directly into laptop CPU territory, and it lands at the exact moment Apple’s M-series momentum has the rest of the industry sweating.

The Backstory

Apple ditched Intel in 2020 and rebuilt the Mac around its own Arm-based M-series silicon. The result: thinner machines, better battery life, and a $2,199 MacBook Pro with the M5 Max that became the benchmark to beat. Microsoft tried to copy the move with Qualcomm-powered Arm Windows laptops, and the results have been, well, polite. Nobody is mistaking them for Apple’s success story. So Microsoft and Arm went looking for a bigger partner. Enter Nvidia, which already owns the AI hardware conversation and now wants the laptop conversation too.

The Plan

Here’s the spec sheet that matters:

• 20 CPU cores, custom-designed by Nvidia and MediaTek
• 6,144 CUDA cores on Blackwell architecture
• Integrated GPU performance roughly equivalent to an RTX 5070
• Support for up to 128GB of LPDDR5X memory
• Rumored TDP of 45 to 80 watts
• DLSS upscaling, ray tracing, and G-Sync supported

The first wave of RTX Spark laptops will come from Dell, MSI, Lenovo, HP, Asus, and Microsoft itself, including the Dell XPS 16, MSI Prestige N16 Flip AI+, Lenovo Yoga Pro 9n, Asus ProArt P14 and P16, HP OmniBook X 14, HP OmniBook Ultra 16, and Microsoft Surface Laptop Ultra. Expect them to land “in the fall,” so September at the earliest. Jensen Huang takes the Build 2026 stage with Satya Nadella on June 2 to dig deeper. Notably absent: the rumored lower-powered N1 chip (12-core and 10-core variants with up to 64GB LPDDR5X) that everyone expected. No N1 means no mass-market push yet.

One more wrinkle: these aren’t gaming laptops. Nvidia is aiming RTX Spark at content creators first, because Arm-based Windows still needs an emulation layer like Prism to run native PC games, and emulation costs you performance.

The Business Model Angle

This is a textbook platform expansion play. Read more about how dominant tech companies leverage their moats to enter adjacent markets on our blog, because that’s exactly what’s happening here.

Nvidia owns AI compute. It owns the GPU. It owns the developer toolchain (CUDA). Once you have that kind of category lock, the smart move is to push sideways into hardware categories where your existing advantages compound. Apple did it with M-series, taking GPU and CPU back in-house to control the whole stack. Nvidia is borrowing the playbook and adding a twist: instead of building alone, it partnered with MediaTek to share fab knowledge and cost. That’s a startup-style move from a $3 trillion company, and it tells you Nvidia takes execution risk seriously enough to bring a partner.

The lesson for founders: when your core product has true platform power, the right time to expand into a neighboring category is when your moat is at its widest, not when growth slows. Nvidia isn’t doing this because GPUs are weak. It’s doing this because the CUDA flywheel is strong enough to fund a four-front war.

The Risk

Now the honest part. There are three real ways this falls flat.

First, price. The M5 Max MacBook Pro starts at $2,199 / £2,199 / AU$3,499. If RTX Spark laptops launch above that, they’re toast in the mainstream market. Windows on Arm has already burned consumers once with overpriced, underwhelming devices. A repeat would be brutal.

Second, software. Arm on Windows still leans on emulation for most PC games. That cripples the gaming pitch and forces Nvidia to lean hard on the content creator angle, which is a smaller, more fragmented market.

Third, power efficiency is unknown. A 45 to 80W TDP that combines CPU and integrated GPU is a wide range, and nobody knows yet whether RTX Spark will sip battery like the M5 or chug it like a workstation chip pretending to be a laptop part.

Quick Questions

The Bottom Line

The lesson here isn’t really about Apple or Nvidia. It’s about what happens when a company sitting on a category-defining moat decides to expand horizontally. The original TechRadar coverage of the announcement frames this as a chip story, but for founders and operators, the takeaway is sharper: dominance in one layer of the stack is a launchpad, not a destination. Nvidia could have stayed in its AI-and-GPU lane and printed money forever. Instead, it’s spending political capital and engineering hours to plant a flag in laptops. Either the move pays off and Nvidia owns three categories instead of two, or it gets a bloody nose and reminds the rest of us that vertical integration is harder than it looks. Either way, the playbook is the one to watch. When you have the moat, don’t sit on it. Build the next one.