The CPU doesn't run all cores at full speed because of how tasks are managed and limited by hardware design.
The CPU doesn't run all cores at full speed because of how tasks are managed and limited by hardware design.
That's why Cinebench doesn't feel like a "realistic" benchmark—it tackles everything at once, using maximum resources. Handbrake is similar but I don’t use it; it’s slower and more resource-heavy. I prefer Shutter Encoder because it runs efficiently without overloading my CPU, and it offers a smoother interface (usually around 89%).
It's... challenging. Cinebench breaks the image into numerous tiny squares and processes them one by one, while a video encoder struggles to keep up with such parallelism over extended periods. Many tasks become sequential or hard to speed up in a video encoder. Additionally, comparing Cinebench and a video encoder means handling varying data sizes—depending on your content, you might fit all active work into the CPU's L2 and L3 caches (with 36-64 MB), avoiding slower memory access and keeping latency minimal. With a video encoder, you may need to access main memory more often, which can reduce performance.