Mini-ITX MiniMig 2.0 and the Yaqube ARM board...

[wolfchild]

motorollin wrote:
Out of interest, why does it require a hardware controller? Why can't the hardfile access be done in software? How does the controller interface with the rest of the board and make the hardfile support available?

--
moto

The original PIC is a slow processor (it gives less than 5 MIPS at 20MHz) and has very little RAM.  So little that it cannot even cope with folders on the SD card.

On the other hand, the ARM processor is much faster (I reckon 18MIPS+) and has enough RAM to allow buffering of the read/write streams, and possibly support for folders on the SD card as well.

In Yaqube's video you can see around 400k transfer rate in SysInfo using the ARM processor.  If using the PIC this would most likely have been 40k or less!

[wolfchild]

Belial6 wrote:
Partly, it appears, but there is also a USB connector on the board, so I assume that that requires at least some circuitry.  It is also possible that the ARM does not have everything in chip that the PIC does, so that could add circuitry.  Someone would with more experience with the two would have to tall us that though.

Well, the ARM has more peripherals than the PIC, though for this application, it's likely that use was made of onboard SPI and USB device.  (Not USB host!)  Obviously Yaqube could clarify on this.  I'm merely speculating.

If you'd like to have a look, datasheets are here:
PIC18F252
AT91SAM7S

Mind you that Yaqube has mentioned previously that he plans to change ARM processor from an Atmel to a Philips LPC2388.  The Philips part looks like a very nice chip.  It has Ethernet, USB Device, USB Host, SD card interface, battery backed clock and lots of extra goodies!

Regards,
Edwin