Blender 4.1 Benchmarks Confirm Even Faster CPU Render Times Under Linux

Blender has long enjoyed faster CPU rendering under Linux compared to using Microsoft Windows. Across many different processors over the years consistently we see faster Linux CPU render performance than under Windows, though that's typically the case for most renderers. With yesterday's release of Blender 4.1, there is even faster Linux CPU render speeds. Here are some initial Blender 4.0 vs. 4.1 benchmarks.

With Blender 4.1 on Linux there is the ability to have even faster performance than prior releases thanks to now using huge pages in jemalloc for faster allocations. This requests the Linux kernel to use 2MB huge pages for large allocations rather than 4KB pages. Developers found this could speed-up the CPU rendering times for Blender 3D on Linux by about 5%. The change doesn't impact Blender on Windows.

Curious about the new version, I ran some Blender 4.0 vs. 4.1 benchmarks on a few Ubuntu systems for reference. My initial tests were done on a System76 Thelio Major Workstation with an AMD Ryzen Threadripper 7980X processor and then DIY desktops around the AMD Ryzen 9 7950X and Intel Core i9 14900K processors. Here's a look at how those three different systems perform with Blender 4.1 compared to the prior Blender 4.0 release:


For the simpler scenes with Blender were slight reductions in render times...



While there are more meaningful gains when rendering the larger benchmark scenes with Blender 4.1.

Blender 4.0 and prior was already consistently faster on Linux than Windows while now with Blender 4.1 that delta is set to be even larger... Should be fun for the next round of Windows vs. Linux benchmarking.

Those interested can find more of my Blender 4.1 benchmarks data building up. Fresh Blender Linux GPU rendering benchmarks soon.