Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
For normal multi-threaded apps, obviously more cores is better.
Even though in our case we have hundreds of connections per second, the length of those connections is extremely short and the primary component for us is the single-threaded PHP app that delivers the content. Yes, it pulls from a database (another core via mariadb) and serves it (another core via nginx), but those times are almost zero compared to the time spent in PHP. Each page load for a thread is something like 80% PHP and 20% MySQL.
So throwing a lot of cores at it doesn't improve speed for us, as we are bound to a single threaded app. Our server load is already low most of the time, so we don't need more cores to serve more connections per second. I focus on response time, time to first byte, time it takes for maintenance to complete and various scripts...
During our busy times of day, we see thousands of requests per second, but our response time and TTFB remain largely the same. So while we are CPU-bound, it's a direct result of a single request -> PHP response time, nothing is being "queued up" waiting to execute per se.
You should take a look at AMD Epyc CPU's to replace your Xeon's if you are utilizing multiple cores efficiently. I have been a fan off and on of AMD for a long, long time -- the very first build that I built myself from scratch was an AMD 386/40, and since then they've beat Intel every now and then. This is one of those times, they are handily whooping Intel's butt.
It's funny that AMD was first to release PCIe 4.0 boards this year as well, before Intel...
I'm pretty sure loading times for threads are faster by several multiples. Could be confirmation bias after reading here, but I doubt it. Plus you will probably have a measure of that to prove it anyways.