How I've Tested
Common for all tests was the following configuration:
- 64 MB RAM
- Matrox Millennium 4 MB, BIOS 2.5
- Quantum Fireball 1280 for Windows 95 and DOS Tests
- Seagate Cheetah ST34501W at DPT P6x86MX (M2)144UW plus RC4040 w/32 MB EDO RAM for Windows NT Tests
The 6x86MX (M2) ran under the following conditions:
- Windows 95 (Build 950, Matrox Millennium Driver 3.41), Windows NT (Build 1381(Service Pack 2). Matrox Millennium Driver 3.06) and DOS Tests:
The K6 ran on different boards depending on the test configuration:
- Windows 95 (Build 950, Matrox Millennium Driver 3.41) and DOS Tests:
- Windows NT (Build 1381(Service Pack 2)., Matrox Millennium Driver 3.06) Tests:
The Pentium MMX ran under the following conditions:
- Windows 95 (Build 950, Matrox Millennium Driver 3.41), Windows NT (Build 1381(Service Pack 2). Matrox Millennium Driver 3.06) and DOS Tests:
- The Pentium Pro (256 kB L2 cache) ran onWindows 95 (Build 950, Matrox Millennium Driver 3.41), Windows NT (Build 1381(Service Pack 2)., Matrox Millennium Driver 3.06) and DOS Tests:
The Pentium II (512 kB L2 cache) ran on
- Windows 95 (Build 950, Matrox Millennium Driver 3.41), Windows NT (Build 1381(Service Pack 2)., Matrox Millennium Driver 3.06) and DOS Tests:
Winstone, Winbench and IMB ran for each operation system under 1024x768x65536x75. Each Winstone test was ran 3 times to get reliable results.
Summary
Better late then never is the best thing to say about the 6x86MX (M2). We didn't expect it, but the 6x86MX (M2) is a very attractive CPU. I have to say that I'm personally impressed with Cyrix, because after all the trouble they had with the 6x86 I never would have expected a CPU as good as the 6x86MX (M2). My long term visitors certainly remember that I wasn't really a friend of the 6x86 when it came out. The thermal problems, the problems running NT 4.0 and some nasty software incompatibilities as well as the weak FPU couldn't really wake my interest in this CPU. When eventually the 6x86L was released, that was solving almost all of the 6x86 problems, it was just too late to play a role in the high performance market anymore. The Pentium MMX as well as the new K6 were much much faster than a 6x86.
Now the 6x86MX (M2) comes out and it's even faster than the AMD K6 at really pathetically sounding clock speeds. Have you expected that? I certainly haven't and I know that AMD hasn't either. The question will now be if Cyrix and IBM will be able to produce a decent number of these chips and especially when they will be able to release the 200 and 225 MHz version. The 225 MHz 6x86MX (M2) will not only be a tough competitor to AMD's K6, but it will reach dimensions of Intel's Pentium II as well.
As long as you don't run highly FPU depending software like CAD or 3D rendering, the 6x86MX (M2) is an excellent CPU. It has less trouble with motherboards than the AMD K6, because it runs at the Pentium MMX voltage of only 2.8 V. In other respects it shares a lot with the K6, including the inability to run in multi CPU systems.
The Pentium Classic, which is anyway much slower than the Pentium MMX and the Pentium MMX itself are CPUs that nobody should waste his money on anymore. It makes me smile thinking of the soon released Pentium MMX 233. It will already be slower than the 6x86MX (M2) at 166/66 MHz. It's also funny to realize another thing. Just imagine an 6x86MX (M2) at 300 MHz. It is obvious that it would be much faster than the Pentium II at the same clock speed. Isn't it surprising that little Cyrix can develop a CPU for a pathetic Socket 7 board, running with the old fashioned and slow external level 2 cache at only 66 MHz, that is faster than giant Intel's high sophisticated 'SEC' CPU Pentium II, with it's complicated and of course expensive cartridged level 2 cache running at half the CPU speed??? Doesn't it look as if Intel needs this complicated and expensive level 2 cache technology to cover their inability of designing a decent CPU core? How fast would a 6x86MX (M2) be with a level 2 cache like the Pentium II ? Should we really be stupid enough to even pay more money for Intel's lack in CPU design? I really wonder ....
For Intel the times are getting tougher and tougher. This year is far from a good year for the world's largest semiconductor manufacturer. Trouble with this guy of that hardware website, trouble with the whole press, AMD's K6 launch, Pentium II bug, lawsuit from DEC and now the Cyrix/IBM 6x86MX (M2). Well, Andy, I guess things can only get better really.
The Jedi are back and the reinforcements from Cyrix and IBM will hopefully jeopardize the plans to the new Death Star called 'Slot One'. Even the little Ewoks form IDT will come and help with their C6 against the bad Emperor. The Force seems to be with them.
