New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
In addition to premium/newer hardware, somehow I always see better results with VDS than with shared cores, even though my average usage is always in the 10-20% range. Never figured that one out either. Both use KVM, right? Or is there any radical difference in the technology used to deploy VDS by various players, and the regular KVM VPS?
From my experience of using KVM and VMware vps, KVM has more overhead, and if it's not that, then providers are skimming a bit from each deployment, allowing them to cram more customers on the node rather than to have a little extra capacity spare. That's both vds or VPS.
That's not surprising. I am not against sharing unused resources either. But for some weird reason, it does seem to have some effect in my case. Or perhaps it's just the better/more updated network hardware used by the 2-3 VDS providers that I've used, and nothing to do with the fact that it's a VDS deployment as different from a regular KVM with shared core.
On the topic of KVM, I doubt if anyone provides anything else these days.
However, I remember using SoftSyshosting in Singapore at one point, and remember having excellent results at that point. I think they used to use Leaseweb SG.
Despite being on very mass-market hardware -- I think 2650 or something similar -- the VPS gave very good results. I checked the virtualization technology, and turned out, it was Hyper V.
Similarly, I believe Linode used to use Xen in the old days (5-10 years back), and I was a Linode customer for 2-3 years in a row. Used no one else in those days.
I really think it's down to the provider, they can choose how much you get for your money, I have tried dedicated core vps with 4 different providers with similar ryzen so far, all similar hardware, so theoretically you can expect the same from each one, I will allow 10% difference as wiggle room, but I have seen as much as 30% difference in single core performance, to me, that is a lot, is a measurable and defineable amount of time running my own workload, like 2 jobs for every 1 on the slower vps.
I have found a provider that does provide value for money for my particular workloads, data bandwidth is not that important, that baseline CPU performance is very important though.
So yeah, I think it's down to how much they give you and how much is skimmed so they can deploy 1 more vps out of the node.
I stand corrected at any time, what do I know, I just use very CPU heavy tasks
@WebHorizon have coffeelake vps, if you need dedicated core, buy 3 core vps in SG, you could probably use 1 core and not bother anybody, just a suggestion, it's a very powerful CPU.