Starting the round of Liquorix 3.2 kernel testing were several disk benchmarks. With the Threaded I/O Tester running eight threads of 64MB random writes, the performance across the two Intel notebooks were quite close between the two kernels. However, for the AMD FX-8150 Bulldozer setup with a Serial ATA 3.0 solid-state drive, the stock Linux 3.2 kernel from Ubuntu Precise was many times faster while the Liquorix kernel was a great deal of time. It is a loss for the Liquorix kernel, but this kernel is what is said to be designed for "desktop, multimedia, and gaming workloads" -- not multi-threaded server/workstation disk tests.
Interestingly though, with the PostMark disk benchmark, the Liquorix kernel performed favorably. On the Core 2 Duo T9300 notebook, there was a measurable gain with the Liquorix kernel while from the Sandy Bridge notebook the speed-up was more than doubled. The SATA 3.0 SSD on the FX-8150 system did not successfully complete PostMark with either kernel.
Moving over to gaming -- one area where the Liquorix kernel is focused upon optimal performance -- the results were surprising. For the Intel Sandy Bridge system with its in-kernel DRM graphics driver, there was a measurable loss in performance when running Nexuiz. When running the AMD FX-8150 Bulldozer system with AMD Radeon HD 6570 graphics on the open-source driver, there was no measurable change in performance. With Ubuntu 12.04 there remains problems using Nexuiz in conjunction with the Nouveau graphics driver, so no results are available from the NVIDIA-powered Lenovo ThinkPad notebook.
When running OpenArena, the Intel Sandy Bridge graphics continued to be slower than the stock Ubuntu kernel while the Radeon graphics did better on the Liquorix kernel.