oh for sure there’ll be reason for it but to opt for that beyond 64GB pushing the limit of what is normally available and then expecting many miners, seems rather odd as in risky.
Reminds me of a niche project “emonie” which I watched just for the pure comedy of it; really odd for having a single rich dev and an unwieldy project and seemed to be centered on the idea of each node hosting the entire network state in memory!
tldr; every limitation, reduces the potential engagement… many users with hardware they already own, is a winner.
I would not say they are that different in all regards. Filecoin is basically the same as private data. In that case the cost should be the same weather you charge it to store it or fetch it. Of course in reality they won’t be because that assumes resources cost the same on both networks. But in theory if they are both competing for the same job the competition should make the cost exactly equal in a frictionless market. So maybe different animals, but like a bat is to a bird, they occupy alot of the same niche.
But it most likely just highlights greed and ridiculous overhead.
The average yearly cost to operate a large data center ranges from $10 million to $25 million. A little less than half is spent on hardware, software, disaster recovery, continuous power supplies and networking. Another large portion goes toward ongoing maintenance of applications and infrastructure. The rest is spent on heating, air conditioning, property and sales tax, and labor costs. ref
Do you have an estimate for what we might expect for the relative cost of storage on Safe Network vs S3 (even for a really wide range / confidence interval)? Just curious, no pressure at all to commit to anything Anyone else with a guess feel free to chip in!
Given that the initial price will be set by MaidSafe and then will self-regulate according to the supply of resources, I dare say that if we want we can set a price 1000 times lower than S3 and I am almost sure that this price will remain so low due to low network usage, because the people who are in the crypto world now are here as speculators, not as users, so I speculate that the price will remain so low for years…
When you say cost, do you mean the amount necessary for farmers/the Network to breakeven (i.e. no profit) or do you mean the price (i.e. the amount necessary for farmers/the Network to cover their costs with margin)?
I think it’s would be great to have a breakdown of the relative advantages/disadvantages as well as utility (i.e. applicable use cases) of SAFE vs Filecoin vs AWS vs Storj vs Sia. I think that will be very informative in terms of understanding hypothetical market size and growth as well as target market/user.
I agree that:
For enterprise applications, there is no room for error. Downtime costs money, and so I don’t suspect Safe will see enterprise applications until reliability has been proven for quite some time. Startups, however, will provide a great avenue to test out servicing business applications as well as proving out reliability over time.
I wouldn’t try to put numbers on it or confidence levels, but my thinking is this:
cloud storage is far from a monopoly and my impression is that there’s a competitive market there with at least Google, Microsoft and Amazon plus many smaller players. Just an impression though, so worth examining.
If so the question would be how much “excess profit” could Amazon be making, because Filecoin providers will be facing much the same or higher costs than Amazon.
For Filecoin providers to be breaking even, there costs would be no more than ~5% of Amazon’s, meaning Amazon are able to charge enough to make >95% net profits on sales. (That’s Filecoin costs and Amazon costs being even, and Filecoin charging 5% of AWS, leaves Amazon with a profit of 95% on sales).
So how much profit do we think Amazon might be making? Well 50% sounds pretty damn good for this kind of unsupported service, but let’s say they have some big cost of scale and technology advantages (in which case another monopoly is growing), and could be somewhere between 50% and 90% of sales.
That’s looks tough for Filecoin providers, and looks like over supply pushing the price will below cost. Not good for them or their network, but maybe this will be temporary. It does make for an attractive or storage option, but I’m guessing that’s not the only reason big cloud providers are popular.
My guess is the situation is worse from their point of view and that even 50% profit for Amazon on this is way too much. Should be possible to find out from public accounts or surveys. The lower Amazon’s profit on sales, the bigger the losses being made at those price levels by Filecoin providers, and they’ve invested a lot of capital that is an additional cost (probably including debt with accumulating cost of servicing) if their hardware is lying idle.
Fascinating to watch, so thanks for keeping an eye on it and updating us @mav. I’ve no idea if my logic or figures stack up so please feel free to pull them apart.
This sounds true if the filecoin network was not farming a coin to subsidize the difference. In reality, we plan to do the same in the Safe Network. People will pay a symbolic price out of their own pockets, and the rest of the pay will come from inflation. There are currently 30 million filecoins printed, with a maximum of 2,000 million.
It would be interesting if someone could calculate the difference that farmers get from inflation…
And considering all of the other benefits that Safe N would offer over Arweave (e.g., convenience, communications, speed, etc.) I would guess that storage on Safe would end up costing somewhere in that range. Let’s say that there’s no additional increase in WTP from those additional benefits and go with the low end of $5/GB.
There’s argument to support the pricing being less than $5/GB since the convenient farming would allow a much larger number of farmers and idle resources being put to use.
But there’s also argument to support that the pricing would be greater than $5/GB as bandwidth becomes the limiting resource. Note that after their initial entry into the market and using VC money to subsidize the cost of their products, AirBnB and Uber, despite freeing up idle resources, end up costing about the same as hotels and cabs respectively. So Safe N storage might be cheap initially but end up costing more as subsidies run out and its popularity rises.
Another way I like to think about the eventual price of storage on Safe N is 10 to 30 times the cost of S3 per month, i.e., $3 to $8 per GB, which also around the $5/GB thesis above.
This is news to me. What price are they going to set it to?
Do you think we will have a lower upload rate than Filecoin?
The price remaining ‘so low’ would suggest lots of people would want to upload data, right?
Either. Both.
Asking ‘will safe node operators run a loss / profit / breakeven’ is a different question than ‘what will be the cost ratio with S3’.
I wonder if when Safe will become another multicloud option? That could be an easy way to get people to get started with very low risk because they have redundancy from the other clouds.
I guess the other way for filecoin providers to break even is if the filecoin price has been speculated very high, so it’s not related to their vs amazon costs, it’s related to their vs amazon income. I feel like eventually filecoin costs:income will converge towards amazon levels, but for now I’d guess viability is probably not a costs thing, probably more of an income thing. @Dimitar seems to agree: “the rest of the pay will come from inflation”.
Yeah, this would probably indicate filecoin storage costs are currently low because of too much income rather than very low operating costs. Oversupply would only happen if suppliers were being paid too much income.
Certainly initially. Filecoin had a lot of publicity, while very few people know about Safe. But if there are enough brave people in the community to place ads on social networks and if we organize free distribution of storage, we can change that.
Cheap price for storing means that the token will be expensive. People keep expensive money (gold, bitcoin) and spend cheap ones (fiat). I continue to speculate that in the first stage of life span the Safe Network will be used mainly as money because:
people have free storage at the moment (unfortunately it is free because people are the product that is sold to advertisers);
there is not enough good money in the world (Fiat does not work to storing value, gold is better than Fiat, but it is only for the rich, bitcoin is better than gold, but again it is only for the rich - safecoin will be the first currency to be for the poor)
I agree that filecoin has a lot of publicity but would be interested in how to quantify the difference. None of the numbers I have found seem to show very few people knowing about Safe relative to filecoin. But I agree with you intuitively, so I’m not sure where the revealing data would be.
