For forcing 'little endian' upon us, coupled with windos.

Ermm what exactly is the Register involvement with this case?

MMMmmm cases...

It would get *really interesting* if Intel was responsible for the problems AMD had with the Opteron's late last year, and proof of it was found and usable in court. ;->

Little endian is harmless. Imagine you have to add two very large numbers on a bigendian CPU. You start by adding the least significant words together, and adding the carry bit into the addition of the next most significant words. You can have your pointers point passed the end of bigendian numbers (error prone). You can add the size of the numbers to the pointers (waste of time). You can use mixed endian (bigendian bits, little endian words) (Extra hassle to find the Nth bit). Little endian is marginally simpler for arithmetic.

The as for Intel and windows, the damage was already done by the time IBM chose DOS.

Intel lost my vote with the 487 (A complete 486 which only works when connected to a 486 that has had its internal math coprocessor disabled, and requires a motherboard with a coprocessor socket when half the advantage of the 486 was that it did not need an external coprocessor). All because people would not pay £50 for a CPU. Now we have twits willing to pay £500 for a CPU, and Intel laughing all the way to the bank.

Actually, motorola architecture in 'big endian' proved to be much easier for me than the intel little endian ever did. Looking in memory, the bytes are in the order I would expect them to be instead of being swapped around. It's a matter of preference, but mine's clearly BIG.

You're right about DOS, but for me it's winDOS that really crapped IT.

Hi, I have heard second hand about the deals made with distributors not to sell AMD for better prices, with OEM not to build with AMD, etc. To this day Super Micro makes AMD OEM server boards but does not advertise them on their main website website. The lawsuits has got Intel covering thier butts and these antitrust abuses are stopping but not gone. And for the Opteron/Phenom cache errata, Intel has the same issue in another cache which has never been fixed.