Market Research on market size and growth

Assuming a useful life of cheaper drives at 3 years as they have 3 year warranty that means a 20TB drive would have to cost less than $864 / 5 USD. That is $172.80 USD each. I know in Australia we could not get 3yr warranty 20TB drives for $172.80 USD each, no way.

But is that 100TiB stored or available.

But if you have 100TiB available then you are likely to also have it in a RAID meaning overheads play a factor. Overheads would include at least a RAID 5 system (one extra drive) and either installed in the PC or as a NAS. Either means an additional overhead.

Keeping to a minimum then additional hardware at approx $300USD and 6 drives.

Now NAS grade drives need to be less than (864 - 300)/6 which is $144 USD each drive.

You think correctly

And all that before factoring in electricity cost and maybe extra cooling in summer months

4 Likes

I am at the stage I ask why? Not technically, I get it and know tis is random stuff, but basically it’s sad that many networks are just uploading stuff, good data, crap data and random data and think, that’s it, we have won.

Recently I am much more keen on knowledge, not data. Personal knowledge, which may be just crap to other folk, but knowledge to the user or group using it privately. Public data too, knowledge, so good quality knowledge that happens to be stored as data.

I think much of the community are missing a trick here. It’s a bit like AI training, it’s not quantity, but quality that matters. I think decentralised data networks MUST become decentralised knowledge networks.

5 Likes

In Chia AFAIK it’s just data that is only useful for mining. So all hash type stuff, but not usable data.

3 Likes

At least with Safe it’ll cost the uploader so (hopefully) that will be the rarity rather than often.

1 Like

How can Safe be that?

That’s the big question. Given where LLM’s are going, they will eat everything. But I think this is all still 10-20 years away at least.

So there is time for SAFE to evolve.

DeStor has 2.1 EiB stored for 1826 clients using filecoin
https://destor.com/en-us/filecoin-network-client-explorer

Some big data sets being stored, the biggest client storing nearly 25 petabytes. This is a pretty nice list:
https://datacapstats.io/clients?sort=[[%22initialAllowance%22%2C0]]

10 Likes

Some stats about permanent data storage network Arweave

https://viewblock.io/arweave

Storing 192 TiB of data

Supplying 372.52 PiB of storage

Which is 0.05% of available space being used for storage.

Store cost is 0.749 AR/GiB which at a rate of $24.56/AR is about $18/GiB

One of the things about their block explorer that impressed me is you can see data that’s stored. eg block 149147 has transaction aT3hok2… which shows the image being stored by that transaction.

They’ve done a total of 8,329,765,620 transactions and started 2020-05-31 which is 1545 days of activity, translating into an average rate of 62 tx/s for the entire life of the network. The daily stats show 346 tx/s.

Pretty impressive network. But still a drop in the ocean compared to the 20 EiB chia farming and 22 EiB filecoin mining.

12 Likes

Also seems very expensive.

3 Likes

It does, but like Autonomi it is pay-once-store-forever, so it doesn’t seem as bad… will be interesting to see how Autonomi compares after launch in terms of cost.

8 Likes

Chia seems to be having a bit of trouble

Farming capacity is down to 17.75 EiB after a local peak of 35 EiB 10 months ago.
https://xchscan.com/charts/netspace

Looks like the reason for this is lack of farmer profitability due to the premine being continuously sold, some conversations at:
https://old.reddit.com/r/chia/
https://chiaforum.com/latest

Not exactly sure what lesson to take from this or how it could have been different, but maybe a) smaller premine and b) putting more of those premine funds into ecosystem development? Any thoughts?

I guess another useful point here is the rate of exodus - 50% capacity lost in 10 months. I wonder how this might affect the design for sustainability of autonomi network.

7 Likes

I don’t think there’s anything to be done with the current version because of the impossibility of predicting what people will do - it’s possible to make a full Bitcoin (I mean the crazy price increase) and it doesn’t matter how much the nodes earn - everyone will want a piece of the pie.

The other extreme is that it turns out that eternal data storage is not sustainable and copies of the technology find more sustainable economic models to human psychology.

Everything in between is also possible.

For example, if the price is volatile, which makes predicting the yield impossible and the network loses, say, up to 50% of the data per year due to periods of farmer outflow, but at the same time this allows you to upload data cheaply, it may turn out to be possible to upload the same data x100-x1000 times (with the appropriate modifications to the data to make this technically possible), which will protect them from the loss…


Check out the Dev Forum

1 Like

We luckily also don’t got farmer profitability

But we do need a utility token

Copying bitcoin dust is…