SAN FRANCISCO, April 7, 2026, 04:10 PDT
Nvidia Corp is under the microscope again after a Reuters piece spotlighted worries among AI and supercomputing pros that its acquisition of SchedMD might steer Slurm—the job-scheduling software widely used on large computing clusters—toward favoring Nvidia hardware. In response last week, Nvidia said its customers “benefit from our open source and free software” and pledged ongoing support and upgrades for Slurm for all users. Reuters
Slurm sits at the core of the system, handling the queuing, scheduling, and allocation of workloads across vast arrays of processors in AI and high-performance computing environments. That’s why the issue draws attention. After Nvidia disclosed its agreement in December, it stressed that Slurm would stay both open-source and vendor-neutral. SchedMD CEO Danny Auble described the acquisition as a way to “enhance the development of Slurm,” but emphasized it would remain open source. NVIDIA Blog
Nvidia’s position in the AI spending surge only raises the stakes. For the January quarter, sales soared 94% to $68.13 billion. The chipmaker is now projecting current-quarter revenue to hit $78 billion. Two customers—unnamed in the disclosure—made up 36% of fiscal 2026 sales. CEO Jensen Huang put it bluntly: “This new way of doing computing is not going to go back.” Reuters
With Nvidia’s dominance in AI hardware under scrutiny, customers and competitors are actively searching for other options. Broadcom on Monday announced a long-term deal with Google to build custom AI chips that runs through 2031. Reuters reports demand for Google’s tensor processing units, or TPUs, is picking up as companies look for ways around Nvidia’s more expensive graphics processors. Reuters
Back in December, Nvidia framed its SchedMD acquisition as a piece of its wider open-source move, with rivals circling. But as reported by Reuters, CUDA—Nvidia’s own software platform—still stands out as a key chip draw. The bottom line: software control can weigh about as heavily as the hardware. Reuters
That’s part of the reason a modest software acquisition has drawn attention well beyond the supercomputing crowd. Late last month, Reuters pointed out that Nvidia’s forward price-to-earnings ratio dropped to its lowest point since early 2019. Investors, the analysis noted, are uncertain whether the heavy AI infrastructure bets from Microsoft, Alphabet, and Amazon will actually translate into revenue and profit at the pace many had been counting on. Reuters
The deal isn’t one-dimensional. Nick Patience, who leads AI platforms at Futurum, noted that pulling SchedMD closer could help Nvidia streamline its tech stack. Still, there’s a risk: any step away from Slurm’s vendor-neutral roots might turn into an “optimized barrier” for rivals’ hardware, and that could spark pushback from open-source advocates. Futurum
At the start of trading Tuesday, Nvidia was holding steady near $177.64 a share, putting its market cap around $4.53 trillion. Investors are still letting the company chart its course, though questions hang over whether AI’s next phase depends more on chip speed, software dominance, or a mix.
