Yes. M means 10^6 in computer applications when referring to communications speeds such as Ethernet at 10 or 100Mbps. The only area where k, M, G take on their binary meanings of 2^10, 2^20, and 2^30 are in areas that are naturally binary, such as memory addressing. Since disk sectors are read and written to/from memory buffers, they also follow binary size conventions such as 128, 256, 512, 1024, etc. Similiarly, record sizes of many serial protocols such as x/y/zmodem, bisync, hdlc, etc. tend to be powers of 2 as well, although this isn't always true. For example, Ethernet with a maximum frame size of 1518 or a datagram size of 1500 certainly isn't a power of two. Generally, whether binary or decimal base is used is clear from the context.