4 ms·

I understand it is faster on RAM, but I am asking why IOW what makes it faster to access sequentially on SSD

• gniv 3y ago
  It is faster in RAM, on SSD, on HDD, for the same reason: the reads are done in blocks. Even if you want a single byte, the system (the controller) will read an entire block, typically on the order of 100k bytes. So if your B-tree node is all stored in one block, all reads from that node will be fast.