Does anyone feel like this is sort of an apples to oranges table? It compares reading one item from L1 cache to reading 1M byte from memory, without adjusting

Does anyone feel like this is sort of an apples to oranges table? It compares reading one item from L1 cache to reading 1M byte from memory, without adjusting for the amount of data being read (10^6 more). It looks like the data was chosen to minimize the number of digits in the right column.