14 DAY TRIAL // AMD is just about ready to launch its 8-core FX-Series central processing units and, in the meantime, the lucky few who got it early have supplied the web with benchmarks.
A certain user from Overclockers.ua, going by the handle of blood_angel, put the soon-to-be launched AMD FX-8120 CPU through some benchmarks.
For those that don't know, this item is an 8-core beast with a total of 16 MB of cache memory (8 MB L2 and 8 MB L3).
During the tests, it had a clock speed of 4 GHz (4,003 MHz) and a multiplier of 20x.
The results were fairly impressive in AIDA64 (see the high-resolution screenshot) and, in CineBench R11.5, the score was of 6.55.
One might look oddly at the 6.55 number (OpenGL score was 46.82), but they should remember that, unlike in 3DMark programs, where scores reach thousands of points, Cinebench returns much lower renders.
Indeed, the highest achieved, according to this list, is 27.20, and that belongs to a system with 48 cores (Opteron 6174) and just as many threads. Meanwhile, top OpenGL render is of 56.04 (Xeon X5680 12-core system).
Go here to check out the discussion that these latest shots spawned (it is many forum pages long already).
For those who want to know what the chip's base settings are, it is an 8-core model with a base clock of 3.1 GHz and a Turbo Core limit of 4 GHz.
Of course, enthusiasts are free to start meddling in those settings and push it even higher, though it falls to them to make sure meltdown, so to speak, is avoided.
The CPU also boasts a dual-channel DDR3 memory controller that handles frequencies of 1,866 MHz.
Other specifications include a TDP (thermal design power) is 125W, or only 95W (for the energy efficient version).
AMD will have the Bulldozer units out in two days (October 12, 2011, Wednesday). Until then, follow these steps to make sure your system is ready for it.
