Ask Blinders
Yesterday
1985
How big are the balls of Google to lay off thousands and then do 70 BILLION in stock buyback?
Tech Industry
Yesterday
879
I watched two Google Engineers refuse to give up a seat for a lady holding a newborn baby at the airport.
Tech Industry
Yesterday
844
Is meta really worth the wait?
India
Yesterday
846
Ideal indian parents
Tech Industry
Yesterday
3273
Pray for folks at Tesla
I’ve been seeing more and more cases of smaller companies announcing that they are moving back to having their services run on-prem via blog posts or just hearing through general word of mouth. I assume this is stimulated by cost-cutting measures due to the recession. I mean it makes sense if you have a low traffic site or can somehow manage the risk of availability (blackouts, users in only one region, getting DDOSed) but is this a feasible future even for smaller start ups? Does the future of cloud companies look grim, or do the benefits with reliability and tooling outweigh the costs for most cases? TC 205K
Nope, it could be cyclical process with periodical pull backs as software is maturing as well as some hardware could have better reliability over the time, so it would be easier to build and operate hybrid cloud than ever before.
this isn’t an effective long term solution. what needs to happen is more competition in the cloud space. right now margins for cloud are insanely high
Hmm there seems to be lots of competition, but dominated by AWS. Theres azure, GCP, Baba, IBM, OCI, Heroku (although this ones kinda dead…) are the margins not any better for the alternatives?
Not necessarily on-prem, but owned hardware in a (or multiple) professional collocation datacenters has its place. $0.08/gb for egress can kill a company with bandwidth fees.
SAP sees that all the time. Cloud is not just a net positive for everyone. Never was never will. Basecamp’s arguments are valid
I've worked in a small deep learning startup 5 years ago that ran everything on prem. People used to look at me with the weirdest face when I told them, but it was seriously way cheaper. We had something like ~200 GPUs training 24/7, but no real live traffic, so it didn't matter if one of the machines went offline for a bit or whatever. Under those circumstances, the price of hardware + someone to keep it rolling is much lower than what our AWS bill would have been. It makes sense, in some cases, to have things on prem.
Can you provide some links of the blog posts? This is the first time I’ve ever heard of going back to on prem (other than defense, govt)
https://world.hey.com/dhh/why-we-re-leaving-the-cloud-654b47e0
Thanks!