FYI, Postgres fires a shot at MDB claiming faster performance/lower latency. —Kyoami…


Interesting. I’ll add a data point from my days at Oracle on the SPARC processor core. Our processor had 8 hardware threads per core. (AMD and Intel, by comparison, have a paltry 2 hardware threads per core. We looked down on them.) And when it came to multi-threaded work loads we would handily beat the x86 world. You would fully expect this given how many threads we supported simultaneously in hardware. But we were on-par in a single-threaded world. Why did this turn out to be a big deal? Because we discovered that many of our heavy hitters were using SPARC based machines in a single-threaded manner. So our advantage just didn’t matter much. Likely one reason that Oracle dumped everything related to SPARC and Solaris just over a year ago.

So that brings up this interesting point… does it really matter that Postgres is significantly faster the more concurrent threads are used? I have no idea, honestly. I don’t know if the typical use model with those data bases is indeed multi-threaded, or if it is single-threaded (as Oracle found out). If anybody knows I’d be curious. TIA.


First, Postgres is a relational database. Second, it’s comparing ACID (Atomicity Consistency Isolation Durability) transactions which MongoDB added the feature last year. My understanding is that ACID is not the reason people choose Nosql database. It’s just a nice to have feature for nosql.


Giant meh, I’m just shocked, shocked I tell you that a postgres outfit did a benchmark that showed postgres was faster!!!

here is a different perspective by someone who is a little more neutral.…

Don’t get me wrong, Postgres is great and probably a better choice than mongodb in some cases. One database will never rule the entire market.