TL;DR
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
MindStudio reported that an MSI Prestige RTX Spark laptop running native Linux generated 13.9 tokens per second on GPT-OSS-120B, compared with 65.4 on Windows and 65.1 in WSL. The report attributes the gap to Linux’s inability to access encrypted power-management firmware; Nvidia has not provided a response in the supplied material.
A MindStudio test found that an MSI Prestige RTX Spark laptop running native Linux produced 13.9 tokens per second on GPT-OSS-120B, far below the 65.4 tokens per second recorded on Windows and 65.1 in Windows Subsystem for Linux (WSL). The report links the difference to power-management firmware that Linux cannot read, leaving the GPU at a much lower clock speed while plugged in.
In MindStudio’s benchmark, the laptop’s GPU ran at about 790 MHz under native Linux, compared with roughly 2.2 GHz on Windows. The report says the Linux result was 13.9 tokens per second, while the Windows result was 65.4 and WSL reached 65.1 on the same model. Nvidia’s DGX Spark desktop, tested with its own Linux build, recorded 61 tokens per second. These are results reported by MindStudio, not an independent assessment of every RTX Spark laptop or Linux setup.
The report attributes the speed gap to Nvidia’s Platform Controllers and Framework, a Windows-only component that manages power delivery between the CPU and GPU. MindStudio says the relevant power table is encrypted and that only the Windows driver can decrypt it, leaving Linux without the same power-management controls. It also reports that temperatures stayed between 40 and 47°C during testing, which the author says points away from thermal throttling as the explanation.
Other tests in the report show a similar difference: with vLLM and eight concurrent clients, WSL reached 127 tokens per second and native Linux 39.5. A longer run reportedly took about eight minutes on Windows and 37 minutes on Linux. The article also describes Linux installation and memory-configuration hurdles, including a default firmware allocation that left Linux able to see only 28GB of the laptop’s 128GB unified memory until the split was changed in Windows.
Linux’s GPU Power Limit
The reported gap matters to developers considering RTX Spark laptops for Linux-first AI workloads. Tools such as vLLM and SGLang are commonly used in Linux environments, but the MindStudio tests suggest that installing Linux directly may substantially reduce inference throughput on this machine, even when GPU acceleration is available.
For now, the report identifies WSL as a practical workaround: it runs a Linux environment while retaining the Windows driver stack, with benchmark results close to native Windows. That option may not meet every developer’s requirements, and the reported figures do not establish that all workloads or laptop configurations will behave the same way. Still, the comparison highlights how platform firmware and driver support can affect performance as much as the GPU hardware itself.
Popular right now
Samsung The Frame Pro LS03HW 65-inch Neo QLED TV with 27-Month Axiom Care
Samsung 65-Inch The Frame Pro Neo QLED 4K TV (2026) Bundle with Modern Brown Bezel, HDMI Cables, and Home Theater Guide
Ninja DualBrew Pro Specialty Coffee Maker, Black, CFP301
As an affiliate, we earn on qualifying purchases.
RTX Spark Laptop Setup
The RTX Spark name covers laptops such as the MSI Prestige, Asus ProArt and Dell XPS, which the source says use the same GPU design as Nvidia’s DGX Spark desktop. The laptops ship with Windows on ARM, while the desktop DGX Spark ships with Nvidia’s Linux build. The platform’s unified memory, configurable up to 128GB, is shared between CPU and GPU and is a key feature for running large models locally.
MindStudio’s account describes a Linux installation that required hardware-specific workarounds. A Debian attempt lacked several laptop drivers, while the author later used a distribution with Nvidia’s open-source driver support. The default memory reservation also had to be adjusted from Windows: after reducing the GPU reservation to 4GB, the report says Linux could access 120GB. These setup details are specific to the machine and process described, rather than a guarantee of the same experience on other models.
“Windows produced 65.4 tokens per second. WSL … hit 65.1 tokens per second. Native Linux on the laptop … managed only 13.9.”
— MindStudio report
Limits of the Reported Tests
The supplied material does not state the report’s publication date, provide a full benchmark methodology, or establish whether the same performance gap occurs across every RTX Spark laptop, Linux distribution, driver version or AI workload. The results should be read as one reported test setup, not a universal measurement. The source also says Linux performance changes with the power source, but the supplied text cuts off before giving the unplugged result, so that comparison cannot be reported in full.
No Nvidia response is included in the source material, and it does not confirm whether the company plans a Linux driver change or when one might arrive. The precise technical reason for the firmware restriction is presented through MindStudio’s account; independent verification is not provided here.
Nvidia Linux Support Ahead
The immediate option described by MindStudio is to run Linux tools through WSL on Windows, which retains the Windows GPU driver and delivered near-Windows throughput in the reported test. Developers choosing native Linux can use the reported installation workarounds, but should account for the lower plugged-in performance and memory configuration described in the article.
The longer-term question is whether Nvidia will add equivalent platform power-management support to its Linux driver. The source presents that as a potential fix but gives no announcement, commitment or schedule from Nvidia. Until the company provides further details or additional tests establish how the issue behaves across systems, the reported figures remain a warning about this specific laptop configuration rather than a settled benchmark for the RTX Spark range.
Key Questions
How fast was native Linux in the reported test?
MindStudio measured 13.9 tokens per second on GPT-OSS-120B on the MSI Prestige RTX Spark laptop running native Linux.
How did Windows and WSL compare?
The report recorded 65.4 tokens per second on Windows and 65.1 in WSL for the same model. MindStudio also reported 61 tokens per second on a DGX Spark desktop.
What does the report say causes the slowdown?
It attributes the lower Linux clock speed to a power-management component and encrypted firmware table that, according to the report, are accessible through the Windows driver but not Linux.
Is there a workaround for Linux AI tools?
MindStudio points to WSL, which runs Linux tools while using the Windows driver stack. Its test results were close to native Windows performance.
Has Nvidia said it will fix the issue?
No Nvidia statement or fix schedule is included in the supplied source material. The report says a fix would require Linux power-management support, but does not say Nvidia has committed to one.
Source: mindstudio.ai
