10EiB now (chia)
I am not sure, but is not only way to contribute tpurchase HoloPort ? The thing is, they are not available, while it seems expensive for what is inside.
I read that you can install it in your own system
This website provides official public statistics from Storj DCS satellites.
The statistics are provided in JSON format:
Privacy. Security. Freedom
1 Like
Might be too early to call it but maybe filecoin has peaked. Filecoin was on the cusp of 6 EiB (peaked at 5.955 EiB) total storage but in the last 6 hours has gone significantly backwards to 5.8 EiB. Chia sure is capturing a lot of storage space.
5 Likes
To leave Filcoin is not that easy like Storj, when your FIL are locked to several months to years. But good thing is that to leave Chia is very cheap and we will attract many people which can farm here and there…
3 Likes
This is built into Node age: rewards are paid only on section split in proportion to Node age which seems like a good way of doing this without straying into a kind of SNT staking which favours those with deep pockets and is less democratising.
8 Likes
Filecoin at 6 EiB today
0-1 EiB after 87 days
1-2 EiB after 58 days
2-3 EiB after 45 days
3-4 EiB after 35 days
4-5 EiB after 26 days
5-6 EiB after 29 days
2 Likes
Why are days decreasing?
1 Like
If the farmer is in profit and have still resources to invest, it is much less risky to increase own farm.
1 Like
I’d assume it’s days since last EiB and not total days.
3 Likes
Chia has seen some pretty impressive improvements in plotting speed lately with MadMax software. It may lead to plotting-on-demand (ie proof of space becomes proof of work).
This conversation is fascinating:
https://old.reddit.com/r/chia/comments/nzxatl/the_madmax_plotter_didnt_break_chia_but_its/
[with assumptions…] one GPU is capable of spoofing the equivalent of 150 TB of space
I was impressed with how much effort chia put into the VDF before release (competitions), but still we’re seeing huge improvements after the network goes live. Goes to show that a live network is the only thing that really matters for proving robustness. I wonder what we can learn from this?
There’s been a response to this topic by chia in the blog plotting speed and security.
9 Likes
I’m not fully understanding this but one reply from a chia employee later in the thread says
[–]TheRealHaitchchia employee 44 points 1 day ago
Your math is flawed as you’ve forgotten an important point:
A 10 second plotter (well < 9.5 second GPU plotter) would be equivalent to 50TB, not 150TB. There is a new signage point every ~10 seconds, if you’re taking ~30 seconds to generate one plot, you’re missing out on two-thirds of the signage points, slashing your effective capacity.
You need to be able to create and submit a plot in less time than there is between signage points to emulate a 512 Plot farmer.
So it’s not an effective 5PB farmer for $60K - it’s a 1.66PB farmer, and 1.66PB of real disk space is cheaper to purchase than that, and a whole lot cheaper to run.
Any chance you can give a TLDR ELI5 on this @mav?
2 Likes
I can’t comment on the particular maths by TheRealHaitchchia regarding 150 TB vs 50 TB, but one thing that’s really clearly missing from the analysis is the operational costs of gpu vs hdd (the video link from the blog post at 23m16 covers this really well). Yes you can simulate a lot of space using just one gpu, but it takes a lot of energy inputs compared to just buying and using a hdd, so in the end the economics and incentives of the gpu probably will never work out.
The big news to me was the magnitude of improvement; the plotting was so much more improved and in such a short time after release. If we anticipate for safe network a certain amount of improvement in, say, bandwidth… maybe we should expect more than that once a live network comes online. Always seems to be some new trick that isn’t covered by testnets or optimization-oriented competitions.
It also says a lot about the cracks and the edges - if there’s a way to game it then it will be discovered. Simple will always beat complex in that respect, we should aim for simple mechanisms wherever we can. Although simple is not always simple (eg emergent behavior)!
The growth and tooling and exploratory work happening in chia is so fascinating.
15 Likes
100% and this is great.
Music, pure music to the ear.
I was asked a while back to look at joining the Santa fe institute for complexity. I had not realised what an honour that would be and turned down the suggestion. I would love it and could wallow in that environment, like living a dream. However, Safe had started (why I was over in CA) so had no time in any case. But a dream place for sure, heaven if there ever is one.
Lately, though I have thought of the opposite, set up an Institute For Simplicity to answer all the same problems, but focus not on the journey to solutions which IMO go from confusion->complexity->more confusion->loop->Simplify (re-evaluate the “problem”) → simplify → simplify → solution (or nature does not understand or need calculus 
)
Where many times the solution uses the problem to create better outcomes. i.e. not all problems are problems if analysed properly. Also nature and in particular complex emergent behaviours are apparently very simple algorithms multiplied in a colony/society/body to form a complex machine. Kinda like cells->molecules->tissue->organ->organism where cell is simple as can be and organism may be as complex as a human, all from simple cells and connectionty of them (even neuron cells and connectomes making up cortical columns in the brain).
One day, one day I will create the institute for simplicity to actually understand, not complexity, but the evolution of intelligence and the harmony and beauty of not only cell–>orgnaism, but of the basic cell of a rock/plant/water/bee etc. and how it relates to and is required to allow the cell->human to happen and survive.
Economics though, economics, I think may be the most stupid thing we have ever done. Fast start to evolve intelligence and grow as a species, but with a hard stop!. I am not sure, but that’s for later.
12 Likes
Ever hear of the Forth (spelling?) Language back in the 70’s (a threaded language) where there was only ever a very small set of primitives and one built up more and more complex instructions. Your description reminded me of it. Quite easy to build a forth machine and so many possibilities. Guess it was never efficient enough nor really had any applications it would solve.
4 Likes
I have an old friend back in the States who’s an eccentric (maybe genius level) Elec. Eng. - he loves Forth and I agree it is awesome how compact it and simple it is.
The problem IMO with languages is a human one. Forth is a step above assembly (as I understand it - and maybe I’m wrong here, just my backseat take on it), and many of us dumber humans (the vast majority) maybe have a difficult time with, not the language itself, but cooperative coding in more complex projects.
So working with teams of people is hard - keeping track of code and nomenclature is fine if the project is very rudimentary, but more complexity and more people of varying ability seems to require a more abstracted language. E.g. Rust really seems to be geared for group efforts.
1 Like
Economics is the study of resource allocation. There are many camps of course. The real problem IMO is that some humans are the equivalent of AI “paperclip maximizers” in terms of accumulating wealth/power and these people favor particular schools of economics that allow for that to happen. The ideas of these particular schools are then implemented at the State level with guns used to enforce the economic policies.
None of this means that the study of resource allocation - “economics” - is bad or wrong … or ‘badong’ (for you “Kung Pow” fans out there). I’d say the opposite, as if we didn’t analyze resource allocation, we wouldn’t have tools to be able to challenge the accumulator mindset.
1 Like