“Latency is like 20 to 30 milliseconds for first read on hard disk… on SSD it gets down to closer to SSD latencies… 30 microseconds, 50 microseconds.”

Peter Mattis · The Pragmatic Engineer · Distributed databases with Peter Mattis

Play from 23:58
23:58

Audio streams from the publisher’s own feed. Verbatim does not host or copy it.

What was described

Measured first-read latency of Colossus-based storage on HDD vs SSD

20-30 ms on HDD, 30-50 µs on SSD (NVMe)

Said nearby on the same subject

“You can send 20 gigabyte file to each other. Private network, point-to-point, no latency.”
BL
Barrett Lyon
The a16z Show ·
Hard number 20 GB file transferred point-to-point with no latency Share this moment
“With Reed Solomon, you can get that down by a bit lower… I think it was like essentially 2x… versus three copies storing three times as much storage.”
PM
Peter Mattis
The Pragmatic Engineer ·
Why it worked 2x storage vs 3x replication Share this moment
“Gives us five-fold reduction in latency time, 80% reduction in cost, and all of these benefits.”
♫
Said on The Neuron: AI Explained
Why it worked 5-fold reduction in latency; 80% reduction in cost Share this moment
“It's easier to deal with an issue in the earliest possible time than it is to wait. Point easy vs point hard.”
♫
Said on The Good Builder Podcast
How they work Share this moment

Moments like this, every Monday

The strongest new moments on your topics, in the speakers' own words. Free for three topics.

No card. No password. One-click unsubscribe.

Track a competitor?

Pro watches up to 25 names or phrases and emails you the morning after they are said.

See Pro, $29 a month

Verbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.