ARMini Benchmarks

Using version 1.1 of Richard Spencer’s benchmarking programme I have added some further benchmarks to the original table.

In the table below, the benchmarks are expressed in percentages, where 100% is equivalent to a Strong Arm 202MHz Risc PC running RISC OS 4.02

Two of the benchmarks proved sensitive to screen resolution and so I repeated these at various resolutions:

The ARMini offers two major improvements over the basic Beagleboard XM: file system read/write access is very much improved and screen resolution is now as good as the Iyonix.

How old is the ROM image that you’re using with the BB-xM? Recent ROM images should be identical to the images used in the ARMini (unless Andrew is doing something that I don’t know about!). I.e. the Portable module should be switching the CPU speed between 300MHz & 800MHz depending on load, 32K modes should work (although there’ll only be 4 bits of blue, same as the ARMini), and I raised the video pixel rate limit from 86.5MHz to 100MHz (although since that’s overclocking it there’s no guarantee that any machine will be stable at that speed).

How old is the ROM image that you’re using with the BB-xM?

About September 2010. I haven’t retested it.

Edit: sorry about January 2011 version 5.17 CRC 35F3ED86

Can I ask why FS read/writes are so much faster on the ARMini? Is it because it has a USB attached HDD inside instead of flash memory?

As far as I know it’s all down to the ARMini using an SSD instead of HDD. Of course there might be something that Andrew’s managed to tweak that I’m not aware of, and things still stand to get much better generally once SCSIFS is updated to support background transfers.