Performance Mandel FPU | x.264 transcoding
Mandel FPU test
The Mandel FPU benchmark measures double precision (also known as 64-bit) floating-point performance through the computation of several frames of the popular 'Mandelbrot' fractal. The code behind this benchmark method is written in Assembly, and it is extremely optimized for every popular AMD and Intel processor core variant by utilizing the appropriate x86 or SSE2 instruction set extension.
Now if you come from the Commodore 64 / Amiga era like me (Peek & Poke Commands FTW!), you can probably remember rendering Mandelbrot graphics, a mathematical formula that much like a paradox, never ends and thus is repetitive. Back in the 1990s it took me a full day to complete one Mandelbrot image. Amazing where we are right now as the same set of calculations can be done in a split second & even real-time.
The FPU Mandel test again is Hyper Threaded, multi-processor (SMP) and multi-core aware. Again 100% baseline performance, good performance.
Multi-threaded Video Transcoding H.264 (DD5.1) to x.264 AC3 5.1
Movie encoding x.264 - x.264 is a free library for encoding H.264/MPEG-4 AVC video streams. Encoding/transcoding to that format is one of the most intensive tasks a processor can perform. As such this probably is the best test of the entire review. We encode a h.264 Dolby Digital 1080P trailer of 150 MB to Matroska x.264 with 5.1 channels AC3. It's compressed in such a way you can play it back with Haali media splitter and/or FFDSHOW codecs. We use handbrake software which is multi-core aware... the more processor cores it sees, the faster it can and will transcode. This software also is a perfect benchmark for CPU and memory testing.
The displayed number is the number of frames rendered per second averaged out over the encoding process. The higher the number, the faster performance is.
Core i7 processors dominate here thanks to Hyper Threading, put it in perspective and you'll notice that the 1000 USD Core i7 975 is twice as fast as today's tested Athlon II X4 620, yet 10x more expensive.
