> I'm not too raid savvy (that should be obvious ;); but, what precisely > does that chip on the motherboard painted 370 on it do, then? It is an ide controler + a bios that changes some "int"s at boot time. When you are coding in asm you comunicate with hardware calling an interruption. if the bios changes the hard disk access "int" you will see the raid as a normal hd. If you boot in DOS and run some partition program you can see your raid partitions there. > What would > be the difference between a HPT370-enabled motherboard and another board > with simply an additional IDE controller on it? The raid is unimplemented (in the bios software). > There is surely some > hardware component present, as when the system boots up you can hit ctrl-S > I think it is to build your raid arrays and examine the disks. It simply writes some kind of info. somewhere in the hds to allow know in the next boot the raid config. You can use your raid HDs in an other hpt37x raid and boot without enter the bios. (Probably dangerous if the bios versions are diferent :P) > Perhaps I'm oversimplifying, but I insist on drawing a distinction between > a HPT370 controller and simply adding another standard IDE controller. Really they are diferent,... but HPT370 / HPT372 are software raid controlers. > Maybe I'm missing your point, but while the HPT370 can indeed just work in > "normal" additional IDE controller mode, it presents to the system in raid > configuraton a device that apparently can't be seen without special > drivers. This seems to put it squarely in a class of hardware-based > implementation to me. In DOS / Windows 9x you can work without this drivers. Linux (at last the gpl driver) dont use the HPT37x bios software, ... it access the drives directly,... the driver do all the job. > I've heard the same criticism of the Promise raid solutions- are they too > manufactured by HPT? "cristicism"? what's that? I think that Highpoint and Promise are different companies.
