| ||
0x12, 0x56, 0x89, 0x23, 0xEF, 0x89, 0x28, 0xC3 0xE2, 0xFF, 0x04, 0x23, 0x49, 0x41, 0x74, 0x3F 0x56, 0x89, 0xAA, 0xB2, 0xC7, 0x38, 0x28, 0x2A 0x28, 0x28, 0x42, 0x73, 0x82, 0xDE, 0xF3, 0x28
Show 128-bit data for SSE in proper endian order, for 8-bit, 16-bit, 32-bit, and 64-bit block sizes.
128 bits = four single-precision floats. How wide would the next generation processor have to be for quad vector double-precision? Write a memory handler for this new hypothetical type processor.
Future super matrix processors will be able to number-crunch entire matrices at a time. How wide would the data path be? Write an alignment macro for that width. Hint: There are two primary solutions!
Earlier in this chapter the function HexDmp32() demonstrated a 32-bit (8 ASCII byte) hex dump. Write a slightly different function that prints a string of 8-bit bytes in hex of a specified count.
void HexDmp(byte *pData, uint nCount);