Speed of the safe network

That’s assuming the system allows power vault users to contribute as much throughput as they are capable of. I believe there has been talk of limiting vaults’ allotment of data to a certain fraction of the average vault-owner’s device speed/upload speed/storage capacity score in order to discourage monopolization of data centers.


I must have come across wrong. I meant no sense of frustration at all. It really is my opinion that speed comes second. Sorry if it came across in a different manner everyone, but it’s really just an actual opinion of mine, although paralellising as we do should be fast. I just mean the other goals are more important to me at the moment.


As it should be David.

Yes there may be talk of limitation of bandwidth to discourage mega data centre monopolisation but surely that would be a tweak for later once we have something robust up and running.

So don’t expect the SAFE network to automatically soak up all the capacity on your under utilised datacentre for ever. But it might use a good whack of it for a few months in the short-medium term.
But don’t go to the bank asking for $$$$ to put in another few racks of petabyte storage if SAFE is the only major customer in your business plan.

At least thats the way I understand it for now. Tell me if Im wrong. I don’t mind being wrong, I too frequently am but I hate staying wrong.


Of course. But as David has said in the past, we must always cast an eye down the road when designing for today. Or something to that effect… :upside_down_face:


Seems like bandwidth options really have improved last years. For me in Sweden there are now lot’s of alternatives with good upload speed, 100/100, 200/200, 500/500 and 1000/1000. This change with more alternatives have happened fast, not all are cheap, ranging from 35-100$ month, but with future success of MAID/SAFE there should not be a problem to provide 200/200, 500/500 or 1000/1000 (if it will help the network). The lack of public Ipv4 addresses seems like the biggest problem, with earlier tests it did not work connecting through the ISP carrier grade NAT. I had to ask for a public ipv4 address to make connection to the test.


In the meantime in Aussieland…

Sydney city centre (Newtown)


That is brutal, almost unbelievable poor speed for a large city. Are there any good faster options? is it only bad through mobile/wi-fi or through fiber cable also? I have not seen such low speeds for 10-15 years or more, I hope there are better alternatives available to most people around large cities in Aussie.


Any progress with 5G in your cities?


Bummer. Maybe Elon can help later on.

Started recently in selected areas. Big country, sparse population… doesn’t help for infrastructure. But still…


Fiber is almost at the doorstep…just shocking how slowly things can go down here sometimes…

Bloody hell, is that a 5.4% packet loss?


Won’t caching help load balance storage and connections? Then as a result speed everything up?

Too optimistic?

For what it’s worth it also didn’t come across as hostile to me, and I’m not British.

1 Like

I can get 1 gigabit download and upload for $100 more than my current internet connection. Could be fun to have, but at the moment I can’t really justify the extra cost as I don’t really need it. If I could earn more than $100 a month on farming I’d have a reason to upgrade.

1 Like

Likewise or #MeToo

Residential broadband connection, USA via (dslreports.com speedtest).


@Eureka18 A vault at this connection should supply roughly 1.5 chunk per second to the network. A client at this connection should receive about 11 chunks per second. Parallelism within the network should saturate any individual endpoint connection. I suspect this to hold true until the ISP begins to throttle the connection.

[gratuitous emoji filter disengaged]

So true. It’s the SAFE network, not the SPEED network. :cowboy_hat_face:

SAFE : Secure Access For Everyone :sunglasses:
SPEED : Sometimes Pretty Excellent Except Doomsday :exploding_head:

Is there a safe speed? :wink:

1 Like

Though I agree with you, don’t underestimate the stupidity of mankind. I think most people don’t understand how insecure their data is and what impact that might have. Sure, it’s becoming more apparent these days but everybody just clicks accept on every cookie and give away all their data just to be able to post a meme.

if they have to choose between speed and security, I’m quite sure they’ll choose speed. For the network to succeed I think it should be only benefits without drawbacks.


maybe I’m misinterpreting. but say i wanted to upload a picture i just shot to the network. How long would it take to be accepted by the network? don’t think people are willing to wait for their uploads to be accepted (sorry, maybe i’m misinterpreting)

i’ve been thinking even more about this with regard to parralelism. thanks to @VaCrunch his response:

Yes, but isn’t one man’s download another man’s (mens’) upload?

basically the total required upload speed should be equal to the total amount of downloads ( in mbps) divided by amount of providing nodes. So, yes, upload speed can be lower than download speed, but depending on the amount of supplying nodes and congestion (amount of people downloading).

@tobbetj though some countries are getting mirrored down/upload speeds there are a lot of countries where this isn’t the case. I think @stout77 is a good example of this, living in a big city and having crappy internet.