Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Is the RAM shortage killing small VPS hosts?

fourplex.net

Post reply on HN
Print view View original

201–210 of 260 posts

Re: Is the RAM shortage killing small VPS hosts?

#201

Post by graemep »

direwolf20 wrote: post #113
graemep wrote: post #33

Earlier quoted context omitted.

Low cost, simplicity and customer service. AWS does offer Lightsail which is similar pricing.

The terms of use of Lightsail say you can't use it in a way that's intended to reduce your costs over EC2. So they have more free bandwidth, but you're not allowed to intentionally send your traffic through Lightsail to avoid the extortion of EC2 bandwidth pricing. They have cheaper CPU, but you're not allowed to use 100% of your CPU.

Very much like a similarly priced VPS though. It lets me persuade people to use a VPS because it is from a familiar big brand. I have come across a lot of people who use EC2 (usually just a single instance) when a VPS would work fine for them.

You cannot use Lightsail to work around EC2 costs, but you can use Lightsail instead of using EC2.

Re: Is the RAM shortage killing small VPS hosts?

#202

Post by miyuru »

aitchnyu wrote: post #192
Gigachad wrote: post #160

Earlier quoted context omitted.

Vultr has one that's $2.5/month v6 only. Probably good if you just need something tiny to run some automation.

What are the issues faced by v6-only hosts and are there countries where it is a non-issue?

GitHub is the main problem currently. Some software like composer does not work due to this.

https://github.com/orgs/community/discussions/10539

For countries, if you meaning connecting to VPS, lot of countries have good IPv6 connectivity now. For me both ISPs I use have native v6. This will differ from person to person.

https://www.aelius.com/njh/google-ipv6/

Re: Is the RAM shortage killing small VPS hosts?

#203

Post by graemep »

bombcar wrote: post #32
saidinesh5 wrote: post #11

Earlier quoted context omitted.

Interesting. I'd have thought these giants would have better pricing because of the scale...

They're selling all their capabilities; using them as a VPS is like using a battleship to cut cheese. But if all you really do with cloud stuff is "ssh into a server I have" (which covers a ton!) then you'll find much cheaper/more performant elsewhere.

> They're selling all their capabilities; using them as a VPS is like using a battleship to cut cheese.

A lot of people do it.

People feel the battleship is safe and familiar. For most businesses the extra cost is not even noticed. Even a small business spending $500/month on hosting instead of $50 is not going to notice.

Also, if something goes wrong (e.g. your AWS region goes down) its far easier to explain to a manager or client that "its Amazon's fault and lots of stuff is down", rather than "its Digital Ocean's fault".

Re: Is the RAM shortage killing small VPS hosts?

#204

Post by direwolf20 »

rjzzleep wrote: post #175
llmthrow0827 wrote: post #174

Earlier quoted context omitted.

This option is available to any sovereign country.

In fact, it is what most European countries used to do for their strategic industries.

America does it too for its strategic financial business, bailing out the banks after 2008.

Re: Is the RAM shortage killing small VPS hosts?

#205

Post by matkoniecz »

filloooo wrote: post #74
ZenoArrow wrote: post #4

I'd see Chinese RAM manufacturers like CXMT filling the void left in the market for consumer-grade RAM modules, I appreciate they face challenges (like lack of access to cutting edge EUV machines), but the RAM just needs to be fast enough and affordable enough for the average user for these companies to make significant inroads into the market that Micron, Samsung and SK Hynix are abandoning to chase the AI server ma…

Their scale is simply too small to affect the market outside China, majority of their chips will be eaten up by HBM3 production with yet unknown yield rate. They are forbidden to buy foreign equipment beyond their current process node, which is already obsolete, die size is 40% bigger than Samsung, not to mention lithography, the big 3 are using EUV while they are stuck with lobotomized DUV. They can start making som…

> die size is 40% bigger than Samsung

likely naive question: why it is a problem? I would be fine if RAM in my PC has 10 times larger physical size if it is overall cheaper.

I guess that larger may have more power draw, but given costs of RAM and electricity and power draw of RAM it sounds unlikely to be a problem.

At a high end it would run into real-estate prices - at some point using half of room for computer stops making economical sense, given costs of rent or buying flat space. But just doubling size of PC does not sound like a bad tradeoff if it would be say 20% cheaper. Or 50% cheaper.

Is it about not fitting existing motherboards?

Is there reason why they cannot just make memories physically larger? It is "only" 40%, not 40000%

Re: Is the RAM shortage killing small VPS hosts?

#206

Post by lachiflippi »

ianseyler wrote: post #61

Can we get smaller VM hosts? I’ve seen some minimums at 512MB for a host. I need 8MB at most sometimes. Update: Fourplex (this host) uses a 1GB minimum.

Running modern full-fat Linux on anything sub-512MB isn't a great experience unless you're willing to do a lot of tweaking, or running specialized distros like alpine. If dropping the whole "vm" thing is an option, you can go much lower -- I've been running perfectly usable alpine lxc system containers on as low as 32MB -- though container-based vps kind of fell out of favor in the last couple years, probably due to the issues that come with not having your "own" kernel in "your" vps. Virtuozzo/openvz was everywhere back then, now it's pretty much all kvm/vmware/hyperv.

Re: Is the RAM shortage killing small VPS hosts?

#207

Post by ZenoArrow »

filloooo wrote: post #74
ZenoArrow wrote: post #4

I'd see Chinese RAM manufacturers like CXMT filling the void left in the market for consumer-grade RAM modules, I appreciate they face challenges (like lack of access to cutting edge EUV machines), but the RAM just needs to be fast enough and affordable enough for the average user for these companies to make significant inroads into the market that Micron, Samsung and SK Hynix are abandoning to chase the AI server ma…

Their scale is simply too small to affect the market outside China, majority of their chips will be eaten up by HBM3 production with yet unknown yield rate. They are forbidden to buy foreign equipment beyond their current process node, which is already obsolete, die size is 40% bigger than Samsung, not to mention lithography, the big 3 are using EUV while they are stuck with lobotomized DUV. They can start making som…

> Their scale is simply too small to affect the market outside China, majority of their chips will be eaten up by HBM3 production with yet unknown yield rate.

They don't currently have the tech to compete on HBM3 production, but they can produce DDR5 memory, and they will undoubtedly be scaling up production on this.

Re: Is the RAM shortage killing small VPS hosts?

#208

Post by kunley »

dpedu wrote: post #62

While not a small host, I thought I would mention what I observed with OVH's VPS offering. I was considering their line of VPSes recently because of how generous the cores/ram quantities were given the price. For example, the smallest offering is 4 cores / 8GB at just over $4 a month. What I found is that it is cheap because the cores, and presumably ram, is old. Like, 2013 era Xeon E3-1275 v3 old. But that's fine! O…

Btw, Hetzner seems to have slightly cheaper offer for such small VPSs, and their disk i/o is much better than at OVH. (My own tests; I always compare the storage speed)

Re: Is the RAM shortage killing small VPS hosts?

#210

Post by pjc50 »

bee_rider wrote: post #59
actionfromafar wrote: post #37

Earlier quoted context omitted.

Turns out their wind "opercapacity" maybe isn't. Maybe they are trading chip efficiency for raw power.

Something I’ve been sort of wondering about—LLM training seems like it ought to be the most dispatchable possible workload (easy to pause the thing when you don’t have enough wind power, say). But, when I’ve brought this up before people have pointed out that, basically, top-tier GPU time is just so valuable that they always want to be training full speed ahead. But, hypothetically if they had a ton of previous gen G…

This is a problem with basically all "spare power" schemes: paying for the grid hookup and land on which you situate your thing isn't free, as well as the interest rate cost of capital; so the lower the duty cycle the less economic it is.
Post reply on HN