Hacker News new | ask | show | jobs
by avinassh 903 days ago
I understand it is faster on RAM, but I am asking why

IOW what makes it faster to access sequentially on SSD

1 comments

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.