VMware and virtualization kn/s slowdown

[Kaj Soderberg]

Hi virtual specialists,

I wonder if it is known, at least in general terms, what a normal engine slowdown would be using VMware as Windows virtualization on a Mac.

Thanks for any information.
Just want to know before spending money, what i might do anyway.

PS I know about the bootcamp possibility, but i have situations where switching on the fly between OSX end Windows is handy.

Best regards,
Kaj

[bob]

Hi virtual specialists,

I wonder if it is known, at least in general terms, what a normal engine slowdown would be using VMware as Windows virtualization on a Mac.

Thanks for any information.
Just want to know before spending money, what i might do anyway.

PS I know about the bootcamp possibility, but i have situations where switching on the fly between OSX end Windows is handy.

Best regards,
Kaj

Should be none for the engine itself. It executes natively. System stuff will be slower however. But the NPS ought to be real close to normal.

[royb]

Hi virtual specialists,

I wonder if it is known, at least in general terms, what a normal engine slowdown would be using VMware as Windows virtualization on a Mac.

Thanks for any information.
Just want to know before spending money, what i might do anyway.

PS I know about the bootcamp possibility, but i have situations where switching on the fly between OSX end Windows is handy.

Best regards,
Kaj

Should be none for the engine itself. It executes natively. System stuff will be slower however. But the NPS ought to be real close to normal.

I concur with Bob here. I'm not a Mac user but I've done lots of virtualization with Windows hosts using LInux guests and Linux hosts using Linux guests and Windows guests: the CPU-intensive functions (which obviously chess engines qualify as) run at full speed or just a fraction of a percentage point slower. Disk I/O is a whole different ball game though -- that is MUCH slower in my experience, thankfully engines don't often do lots of I/O (TB accesses may be one exception but I'm not a huge user of them so I cannot comment on that in this case).

[Kaj Soderberg]

OK, thanks Bob and Roy.
I'm very happy with what i read.

Best regards,
Kaj

[krazyken]

For just running windows based chess engines on Mac, I'd recommend using wine instead. performance is as good or better than when run under windows. The biggest difference you will have between running under wine or running under VMware Windows, is that Windows eats a huge chunk of RAM that might be better used for hash tables. Using wine also makes it easier to use Windows engines under Mac OS GUIs.

[PerroMaster]

AMD 6000+ ATHLON 64 X2 DUAL, ASUS, 4000 (2.87GB), (00:00:21.0) [4385595 pps.] PowX= 13.663

AMD 6000+ATHLON EN UN SOLO PROCESADOR,(sobre WINE/LINUX) , 2.87Gb, (00:00:25.5) [3611667 pps.] PowX= 11.252

The program “Axon-Benchmark-4” shows that under Linux/wine there is a lost of about 20%.

Linux/wine:13.663
Windows:11.252

[krazyken]

AMD 6000+ ATHLON 64 X2 DUAL, ASUS, 4000 (2.87GB), (00:00:21.0) [4385595 pps.] PowX= 13.663

AMD 6000+ATHLON EN UN SOLO PROCESADOR,(sobre WINE/LINUX) , 2.87Gb, (00:00:25.5) [3611667 pps.] PowX= 11.252

The program “Axon-Benchmark-4” shows that under Linux/wine there is a lost of about 20%.

Linux/wine:13.663
Windows:11.252

where does one get this "Axon-Benchmark-4"?

[bob]

AMD 6000+ ATHLON 64 X2 DUAL, ASUS, 4000 (2.87GB), (00:00:21.0) [4385595 pps.] PowX= 13.663

AMD 6000+ATHLON EN UN SOLO PROCESADOR,(sobre WINE/LINUX) , 2.87Gb, (00:00:25.5) [3611667 pps.] PowX= 11.252

The program “Axon-Benchmark-4” shows that under Linux/wine there is a lost of about 20%.

Linux/wine:13.663
Windows:11.252

where does one get this "Axon-Benchmark-4"?

Also it is obviously not a compute-bound benchmark or that would never happen...

[PerroMaster]

http://playwitharena.com/download/aac-bench_v4b.zip

[royb]

AMD 6000+ ATHLON 64 X2 DUAL, ASUS, 4000 (2.87GB), (00:00:21.0) [4385595 pps.] PowX= 13.663

AMD 6000+ATHLON EN UN SOLO PROCESADOR,(sobre WINE/LINUX) , 2.87Gb, (00:00:25.5) [3611667 pps.] PowX= 11.252

The program “Axon-Benchmark-4” shows that under Linux/wine there is a lost of about 20%.

Linux/wine:13.663
Windows:11.252

As someone who runs Linux and Windows chess engines under Wine I can say that this has not been my experience at all. In fact, Linux running Rybka3 and many other windows engines under wine shows that the engine runs a little bit FASTER than the same engine running under Windows.

I'm not sure what the Axon-Benchmark-4 is or what it is designed to test, but for chess engines my comments have held true for the last 2 years as Linux gets faster and faster with each new kernel put out.