Posted YABS from my Vultr Ampere A1 (4 cores, 24GB, free tier) against my paid Virmach-clone x86 KVM at CloudCone ($3.50/mo, 2 cores, 4GB). The ARM box won on every metric except single-thread latency. Disk especially — the NVMe on Vultr's ARM fleet is noticeably faster, presumably because newer hardware.
Posting screenshots in next comment. IMO this says more about host refresh cycles than architecture superiority, but YMMV. Take it with a grain of salt — I may have gotten lucky with neighbor noise on the x86 box.
Wah free tier can win against paid one? This one cannot be right. You already check is not burst CPU or not? Some provider give fast first 1 hour then throttle dong. Can or not you run YABS again after 24 hours?
Post a reply
You need an account to reply.
Log in or
register to join the conversation.