12:19
Iridescence
I updated as well
12:21
Paul N. | BAND (🅑) - I'll never DM you
Sounds like it's mining pool's fault. This should be disclosed if the payout would be incorrect. Withholding this information makes a HUGE impact on the reputation of the pool.
12:23
Zack
I think that Mandel shouldn't bother paying people back for that time period.
12:24
It was my fault for pushing the code that made the fork happen. If our users should be upset with someone, it is me.
Jon Snow joined group by link from Group
12:43
Jon Snow
What’s the latest guys?
12:46
Iridescence
Kassel still getting all the blocks
12:48
Mr Flintstone
can we make a market for if he abandons the chain at 6k
12:49
Iridescence
maybe he'll just throw more hashpower at it
12:50
can't wait to see what the new difficulty will be
13:06
Brian Brian
What’s this critical hard fork for?
13:38
Deleted Account
is the amoveo restart?
13:52
Roger That
Yeah I think I will have to stop mining too. This doesn’t make sense anymore.
13:55
Deleted Account
hardfork cannot prevent batman or kassel
13:56
Jon Snow
Agreed. I will stop my mining as well.
13:58
Roger That
It can stop an asic if you change algo.
13:59
Deleted Account
how could you know that they used fpga or asic?
14:00
but a algo that only suitable for GPU is great
14:02
or use algo only for CPU
14:04
kdmlll
Would it not most likely be a large number of GPUs? The best outcome would be to lower the barriers to entry so that there is more competition. This would be achieved once everyone has easy access to efficient GPU mining software. This will come with time.
14:10
Iridescence
they are using the equivalent of roughly 700 top-end GPUs
14:25
Roger That
Yeah, I think we have discussed these a few times already with very little incremental to add anymore. Will stick around for a week or 2 to see where this is going. Otherwise 2 veos it is for me.
Tarrence van As joined group by link from Group
14:33
Deleted Account
If we start over I would second the idea of changing algorithms while amoveo is evolving, the thing is for it to succeed it's must achieve widespread adoption that it's not currently getting thanks to Kassel
14:34
Iridescence
Well to be honest it's not achieving widespread adoption because Amoveo hasn't done much marketing
14:35
We're just sad at Kassel because we're not getting any VEO
14:38
Deleted Account
Are vertcoin or groetslcoin less secure for them being asiscs resistant?
14:55
Greg
Yes you should mine about 1 VEO in 24h or less if they run at 240MH each. Give it a try
15:01
Roger That
That’s right and that’s how most folks who won’t have access to 700 gpus or an asic will feel like. Which is the majority of folks in crypto or otherwise.
15:02
Jun Dam | 100x Trade | bitcashBank
Hey guys it would be good to have a more even distribution early on to generate more interest so a memory hard algo like equihash or cuckoo cycle would be ideal. I was excited early on with Aeternity partly because the plan was to use Tromp's cuckoo cycle
15:03
Roger That
Mainly if a govt wants to attack and kill them with supercomputers. Which is an extreme case.
15:23
Iridescence
With dominating hashpower even if you find a block, you probably will not be able to get that block included in the chain since the larger hasher will be able to build a longer chain in a shorter amount of time
15:25
If you find one block in 24 hours, the large hasher will just find a block at the same height a few minutes after you, and then build more blocks after that
15:26
Thus the block you found will be irrelevant
15:26
Roger That
I am sure the pool I am mining at is affected by that big time.
15:27
Pools need to have roughly similar hashpower to find blocks together
15:36
Roger That
yeah so mined all day for 0.35 veo
15:36
over and out captain!
15:40
This block is uncharacteristically long
15:41
Nvm, just a fluke lol
15:54
Greg
yes it could but we still mine some blocks on Mandel's pool
16:02
TownAlwaysGet
Cpu miner still working right?
16:04
Greg
the code is working but you will not mine much! there is highly optimized GPU farms running out there, somenone mines > 90% of the blocks
16:05
TownAlwaysGet
It does? On the post-fork code?
16:05
Without any update?
16:06
Somebody is throwing the equivalent of about 700 GPUs of hashpower at Amoveo
16:06
TownAlwaysGet
Okay... It still hard though.
16:07
Greg
Mandel's pool just mined blocks 4956 and 4961!
16:08
Greg
I'm sure the rest of the miners have more than 700 GPUs combined
16:08
Iridescence
My optimized 1080 gets 1 GH/s
16:09
I haven't seen anybody report higher GPU numbers
16:09
Mandel's pool reports about 100-200 GH/s
16:09
The rest of miners are at most another 300 GH/s
16:09
Greg
you need to compile it for Windows! :) Mandel's miner do 500-550 MH/s on a 1080.
16:10
Iridescence
Yeah....
16:10
Windows is such a pain
16:10
Greg
yes but what is the hashrate of Zach's pool without Kassel?
16:10
Iridescence
Can't tell
16:11
Zach's pool doesn't report hashrate of any sort
16:11
Deleted Account
Well if you do maybe we can even the field
16:11
Iridescence
He estimated that network hashrate is at 1 TH/s
16:11
I only have a few hours a day to work on this during weekdays
16:12
Deleted Account
Feel your pain 🤨
16:13
Iridescence
Optimizing CUDA code for Amoveo has actually been quite fun
16:13
I got to understand the hashing algorithms better
16:13
Also getting 1 GH/s was pretty cool
16:14
Though somebody else with a 1080 was getting lower performance
16:14
I think there are a few knobs to turn for Nvidia
16:24
Greg
yep, great job!
16:48
Gonzalo
updated and OK
17:15
CeloStarter Bandits
Zack Wait😀 you can join.. Just didnt want you in the Trader Swamp....
17:16
We got 2 Trades so far.
17:16
last trade 68 Dollar 12 VEOS
17:19
To all: Trader Discord is free to speak.. Any opinion... of Amoveo.. There is also an official one.. You should join too
17:48
CeloStarter Bandits
17:48
thats the official one
18:07
Zack
Thanks for calculating it.
18:10
It looks like Kassel is pushing hard again
Deleted joined group by link from Group
18:28
Deleted Account
Hey everyone
18:28
Zack
This Discord link will not expire. This is good evidence that Can Sen is running a fair market. He isn't trying to manipulate membership in order to trick us into trading at bad prices.
18:28
Deleted Account
Pretty stoked to have found this project
18:29
Is there docs anywhere for mining? Or do I need to scroll up? :)
18:31
Paul N. | BAND (🅑) - I'll never DM you
Are we really moving forward with 95% hash power and mining rewards going to 1 guy? I mean ... imagining 90% of veo gets locked up by 1 player, then he can freely manipulate the price isn't it?
18:31
Zack
see the pinned link.
18:31
Deleted Account
My bad thanks. Whitepaper is solid. Nicely done man.
18:47
Deleted Account
Hey guys, im a bit of a n00b, how does the wallet work?
18:52
The network has about 890 gh/s
18:52
Greg
Zack is there a way to get the distribution of payouts on your mining pool? it could be interesting to be able to track percentage-wise how much VEO each address is mining.
18:53
CeloStarter Bandits
Thank you.. But if you Trade be carefull.. People may exploit your good will..
18:54
This is crytpo.. So it is likely people offer you Veo they dont have..If somebody with rock solid reputation want to make escrow.. Please..
18:55
Zack
We could write script to look up every block, and find all the transactions that my mining pool paid out, and then sum up how much went to each address.
19:01
Greg
okay, thanks will release it if I code something
19:01
Zack
im already making it
19:04
Greg
so I will just use yours, thks
19:05
it seems to have decreased a bit
19:08
Deleted Account
Zack do you calcualte timestamp in the SHA256 hash?
19:09
Zack
the timestamp is in the header.
mining is done on the hash of the header.
19:09
Deleted Account
hmm are we solo mining on this pool?
19:10
so a miner must win a block to get the reward? not pooled / shared rewards?
19:14
Deleted Account
are u sure? how do we see our distribution?
19:14
Deleted Account
so i was looking at the c version of your pow
you add data + diff + nonce
so data here is the blockhash right?. is this calcualted by the full node? and what is stored in it?
19:14
Deleted Account
is the hard fork happened?
19:17
The blockchain had 1 hard fork so far.
19:23
Deleted Account
thx
19:29
Deleted Account
i dont see any unconfirmed balance even though the pool has won blocks while ive been mining..
19:30
Paul N. | BAND (🅑) - I'll never DM you
I can confirm it's a bit delayed.
19:30
Deleted Account
ahh ok, thanks
19:31
Deleted Account
how long have u been mining?
19:31
Deleted Account
lol
19:31
not super productive 🙂
19:31
whats the total supply for Amoveo?
19:32
Zack
a little under 6000 Veo
19:32
Deleted Account
oh that explains it..
19:32
Gonzalo
I mined more than 70% with a cheap VPS 4 days ago, for the other part i'm with 4 gtx1080
19:32
Deleted Account
Zack whats the reasoning for that? any specific reason for 6000 only?
19:33
Paul N. | BAND (🅑) - I'll never DM you
This is day 4 of Mainnet
19:33
Zack
only 4 days of mining so far. so there aren't many tokens yet.
19:33
Gonzalo
why not? 1 VEO has 8 decimals
19:34
Deleted Account
sorry i meant whats the max supply, not circulating
19:35
Zack
the block reward is set by a governance mechanism. the community decides the block reward.
19:36
Deleted Account
right. fascinating. and how does one see what the community has decided on?
19:40
Zack
trees:dict_tree_get(governance, block_reward).
157546925
19:48
Deleted Account
sorry for the n00b questions, but whats the max supply of veo?
19:49
Zack
your guess is as good as mine. Who knows what the community will decide to do.
19:49
Deleted Account
oh right
19:54
Mr Fox
who is the member of the community? how to vote for the reward?
19:57
Zack
Amoveo has no voting.
19:57
Deleted Account
How doe s the community decide?
19:59
Deleted Account
Read it. Can't say i understood it 🙂
20:02
Paul N. | BAND (🅑) - I'll never DM you
Zack How can I bet that the block rewards will go down?
20:02
like, what are the steps to achive that
20:04
The block reward is already in the process of being decreased.
20:04
Paul N. | BAND (🅑) - I'll never DM you
Like api:new_governance_oracle(block_reward, 0.00001).
this?
20:05
Zack
the minimal cost to start an oracle is around 0.33 Veo.
20:05
trees:dict_tree_get(governance, oracle_initial_liquidity).
33826071
20:05
Paul N. | BAND (🅑) - I'll never DM you
and what happen if the 95% guy keeps betting against the governance oracle to prevent the decrease in mining rewrd?
20:06
Thanks for the info :)
20:06
Zack
If the community is able to rally together and disagree with the 95% guy, then we can do a hard fork so he loses the bet.
That means the 95% guy is now only a 90% guy, and we all doubled our money.
20:06
Deleted Account
whats this 95% guy? is someone mining 95% of the hash?
20:07
Someone using an IP address out of Kassel Germany is finding most of the blocks
20:07
CeloStarter Bandits
Kassel ... Germany... Then he is in this Channel...
20:07
Deleted Account
whres the network stats?
20:07
CeloStarter Bandits
You can write directly to him
20:08
Zack
he has not identified himself to me.
20:08
Paul N. | BAND (🅑) - I'll never DM you
or any of us for the matter
20:08
Zack
I don't know what you mean.
20:09
Deleted Account
i can see the pool hash rate - around 52 gh, but whats the total network hash?
20:10
Zack
block:hashrate_estimate().
20:10
Deleted Account
windows user here
20:10
Zack
my node is syncing. give me 4 minutes.
20:11
do u know his wallet address?
20:11
Zack
he uses like 10 different addresses.
20:11
Deleted Account
rigght. whats his hashrate?
20:11
is Amoveo nicehash resistant?
20:12
Zack
over 90% of the network
20:12
Deleted Account
could be nicehashing..
20:13
Deleted Account
impressive
20:13
so he's doing 1.1 th/s?
20:13
Zack
seems like it.
20:13
Deleted Account
thats an impressive number of GPU's or nicehash power
20:14
Zack
I think if it was nicehash, more people would be doing the same?
20:15
Deleted Account
maybe, hard to say... this project is still pretty under the radar
20:15
Zack
the block reward will decrease in the next 20 minutes or so.
20:15
Deleted Account
no ann on btctalk or even an official website
Aizen Sou joined group by link from Group
20:16
Deleted Account
this is like the reverse of every new coin ive seen lately. thats why i like it 🙂
20:17
would be nice to win a block or two 🙂
20:19
my maths says he has about 60 rigs
20:19
each with 7 or 8 x 1070s
20:21
Aizen Sou
Hi Zack I'm curious how did u calculate that?
20:22
Zack
network hash rate - Mandel's pool hash rate - Iridescence pool hash rate.
20:23
CeloStarter Bandits
i have one lausy Veo
20:25
Zack
I decreased the reward in my pool to 0.8 Veo, because the block reward is lower now.
20:26
Deleted Account
Any idea why its taking so long to show pooled rewards?
20:27
Zack
My pool pays out rewards instantly. I never hold anyone else's money.
20:28
Deleted Account
im so confused. which one is your pool?
Markus joined group by link from Group
20:29
Deleted Account
thats a link to the explorer?
20:32
Zack
looks like someone else did a bet in the oracle.
20:34
Aizen Sou
The stats are wrong according to my observation ( I mined Amoveo with my GPUs as well )
20:35
CeloStarter Bandits
@Mandelhoff can you lower the mininumpay out or somehow sent the rest to me if possible . 0.30?
20:36
Aizen Sou
Did u check his balances, Zack ?
20:47
Zack
it looks like no one messed with the oracle. I think the problem is that I failed to make the necessary bet in the oracle.
20:48
no, the oracle is of type 2.
20:49
when I look up the order book, it seems empty. but the oracle says that it has bets of type 2 in it.
20:54
What is especially weird, is that I made the txs a few minutes ago, and they seemed valid.
I guess the order book got messed up when these txs were put into the tx pool and then dropped.
For now I think I can fix this by resyncing from genesis, and then inserting the oracle txs again.
20:56
$33 per veo. 6000 Veo.
our market cap is still around $200k
Deleted joined group by link from Group
21:05
Zack
reinstalling the blocks didn't seem to help
21:14
my guess is that someone bet against me in the oracle to stop it from closing.
I made a new bet just now, so we should be able to close the oracle and decrease the block reward in 1000 blocks.
21:19
I will do some transaction analysis and identify whether or not someone messed with the oracle.
21:27
Greg
isn't that a normal thing that people bet against you?
21:36
Zack
everyone is allowed to bet whichever way they want.
21:39
Mr Flintstone
what if someone else bets against you again?
21:40
Zack
It is also possible that we discovered a bug in the oracle. I will do a little more research to find out what is going on.
21:42
the oracle started at 4134.
I made a bet of 0.4 Veo at 4137.
21:54
Deleted Account
Is miner
1.0.0.2 still functional? Haven't updated it yet.. this evening
21:57
Zack
There is a bug in oracle_unmatched. this txs shouldn't be callable until the oracle is closed.
I will do a hard fork to fix this.
21:59
I am thinking we should undo some blocks to fix this.
So we can undo the mistake that made the market not closed.
22:01
that way we can lower the block reward immediately.
22:03
Mr Flintstone
how many blocks
22:03
Mandel Hoff
I'm not completely against a block rollback, but I am not at all prepared to rollback my pool payouts to handle that sort of event.
22:03
Mandel Hoff
Yes, the
1.0.0.2 miner still functions fine.
22:03
Zack
about 80 blocks
22:03
Mr Flintstone
approx what time was block 5133 mined?
22:04
Mandel Hoff
I just confirmed 5191, but not sure how many else in that window.
22:04
Zack
2 hours, 28 minute
22:05
How many Veo would this cost you Mandel?
22:05
did you find a block in the period in question?
22:06
Mandel Hoff
1 for s ure, I'lll have to check on more - one second
22:08
Just 1 block confirmed after 5133, and I had block reward set at 1.0. Pool cost is just the 1 veo.
22:08
Zack
So, what if you had to choose between these options:
1) we undo 80 blocks
2) we can't lower the block reward for 3 or 4 days
22:08
I can pay you 1 veo to make up the difference.
22:09
Mandel Hoff
I'm fine with the undo option.
22:09
im almost done testing the changes
22:13
Mr Flintstone
were there any other big transfers in that 80 block period? I just hope someone didn’t do a big otc deal and is about to get screwed
22:14
Zack
good point. lets check.
22:14
O K
Asking the tough questions
22:14
Zack
I sent one to myself. of 30.
22:14
Gonzalo
I see 30 VEO on block 5185
22:15
but I guess this was between Zacks accounts
22:15
pool wallet to own wallet
22:15
so no problem I suppose
22:17
Mandel Hoff
5139 has a big one too, but that looks like Zack's as well.
22:22
Zack
ok, I did the restart. we are back at 5138
22:22
light nodes wont work until 5220 or so.
22:22
Gonzalo
ok, do we need to
make prod-clean
22:23
Zack
correct. this time we do need to do that.
22:23
Mandel Hoff
ok, my pool is off for upgrade
22:34
Zack
success:
trees:dict_tree_get(governance, block_reward).
100227592
22:35
Zack
This event is evidence that Amoveo will be able to recover from flaws discovered in the protocol going forward.
22:37
Mandel Hoff
Whoever has this address: BHurv8cnvGRKFt0... you had a 1 veo payment tx that got rolled back. I confirmed your wallet went back to "new", so I moved your unpaid balance back to 1.X and set your total paid to 0. No one else had rolled back txs from me.
22:38
Blocks are flying, but I'm a couple blocks back from Zacks' node on sync. Still watching it
22:40
Mandel Hoff
Looks like I'm staying in sync - just normal prop delay. Just looks worse when Kessel finds fast blocks.
22:45
Gonzalo
My node stoped at block 5141
22:46
Maybe I synced to the wrong node
22:46
O K
got stuck at 5153
22:46
Zack
5153 is the correct high.
22:46
O K
Your node shows 5216 deoesn't it
22:47
Gonzalo
but headers are at 5216
22:52
O K
what does block absorber died!
mean again?
22:53
Zack
that is happening a lot now because we are recovering from a big fork. it is fine.
22:53
O K
ok sync_mode:normal().
now
22:59
Gonzalo
Zack, api:height=5216 // block:height=5163? is ok?
23:04
Zack
Kassel contacted me.
23:04
O K
You have my full attention
23:04
Zack
It sounds like good new
23:05
Gonzalo
this is so interesting
23:06
Zack
He says he wrote a miner for GPU that is 2.4x faster.
He sent me some tokens from <<"BLvrQn16BFfhZo7q2ju165mRd62nXu9UA6E9EB7h7DgDw/LiWmPbkpyG4lXG5YSkH04f3gt9ddXx2DhgC/OuZAE=">>
23:06
He seems supportive of the project
23:06
Mr Flintstone
our boy
23:07
O K
Not coins, code
23:07
Zack
he says he will make the gpu code open source in a couple weeks
23:07
Mr Flintstone
I had a feeling it was greed...
23:08
O K
I don't see a problem with it, but at some point it will benefit him to strengthen the community and network
23:08
And that point is right now!
23:10
Zack
That is hard to know for sure.
23:10
O K
I was being facetious
23:10
Mr Flintstone
No problem with him keeping the gpu code to himself for a while..
23:10
O K
It's easy for me to say as a minnow
23:10
Mr Flintstone
/ her
23:10
Gonzalo
he/she sees his VEOs could worth nothing, thats why he's trying now to be friendly
23:11
Zack
Maybe we can reach some compromise.
Like, we agree to lower the block reward as long as he agrees to open source the miner right now?
23:11
O K
I like the idea of negotiating
23:11
Mr Flintstone
@zack it’s not like he is the sole guardian of knowledge
23:12
O K
Well, I mean we can improve the code we have
23:12
Mr Flintstone
It is not unconceivable that someone finds out how to 2.4x the current code
23:12
O K
I've been reading the code, but I'd be lying if I said it wasn't above my comprehension at this time
23:12
Mr Flintstone
yeah I’m not well versed either
23:12
O K
No not at all, I mean GPU improvements have been way more than "2.4x" the last few days
23:13
Mandel Hoff
2.4x still means that person has ~35x the hardware vs the rest of the network.
23:13
Is make a pool, and have his code default to his pool, no?
23:14
Does he have AMD working code?
23:14
The community has no AMD working code
23:14
Zack
yes, I think is is not cuda code.
Maybe we should get my opencl miner working
23:15
O K
Half my hardware is offline and I'm a minnow. AMD code doubles my hashrate. I can't be the only one. Surely this would help distribution
23:19
Zack
he says his hardware is 30x V100 currently
23:20
Jim
Hm, that explains some things. That’s not cheap.
23:23
Zack
so that is like $250,000 of mining hardware
23:25
so the V100 card is something you can get over AWS? is that right?
23:26
Gonzalo
makes sense 54ghs for kassel?
23:26
Jim
A V100 is approx $1/hr at AWS at cheapest
23:27
Gonzalo
gtx1080ti (3584 CUDA cores) on Mandels GPU miner = 530mhs
23:27
Zack
it is like $0.07 an hour at paperspace
23:27
Jim
The price dropped from $5 or so
23:28
Mr Flintstone
hash rate is about to skyrocket
23:28
Deleted Account
that would be good. But this guy doing nothing wrong, but mining too much blocks
23:28
Gonzalo
V100 (5120 CUDA core) on Mandels GPU miner = 757mhs x 2.4 = 1817mhs x 30 = 54.5ghs
23:29
Zack
oh, $2.30 an hour. my mistake.
23:29
Mr Flintstone
ok lol if it was 7 cents we would see 10 gh today
23:29
O K
So this guy has a major opportunity cost
23:29
Zack
if he has 30, that is like $69 an hour
23:29
Mr Flintstone
that’s around 25k
23:30
Zack
or like Jim says, if amazon is $1 an hour that is a better deal than paperspace
23:30
so only $30 an hour
23:30
Mr Flintstone
this seems like a cheap price to pay for a huge percentage of float
23:30
it would cost us 2.4x more, because our software is worse
23:30
assuming he is honest
23:31
O K
I think that's very likely that he's made 2.4x optimizations
23:31
scott_l | DeFi Pulse
or just restart the chain with an algo you can buy on nicehash
23:32
and let everyone bid for what they think veo is worth..
23:32
Deleted Account
i prefer mining locally
23:32
scott_l | DeFi Pulse
you still could
23:35
Zack
Kassel seems to have invested a lot of time and energy into his success. He has valuable skills to be able to do so well.
I think that things are going well.
23:35
Mandel Hoff
Good math. Sounds like our code base isn't too far off.
23:35
scott_l | DeFi Pulse
all he has done is destroy the early community
23:35
Mr Flintstone
I think this is too rash
23:36
O K
I don't think anything has been destroyed
23:36
Mr Flintstone
yeah, I think this is just an example of someone capitalizing on information asymmetry for their own gain
23:36
scott_l | DeFi Pulse
yea, but you've just been arguing and telling people they are stupid when they said they didnt like the direction things were going
23:37
I don't think I've ever insulted anyone personally
23:37
scott_l | DeFi Pulse
maybe you calling people stupid is a bit of hyperbole
23:37
O K
If I'm not valued in this space, I would be happy to leave
23:38
scott_l | DeFi Pulse
i think that is exactly how most people that have left have felt.
23:38
O K
I think many people thought they were going to get rich quick, and were upset when they weren't in the top 10 off their CPU
23:39
This project has merit, and they thought it would remain a secret, and when it became widely known they would do 1000x
23:39
But it has merit, and so it did not remain a secret
23:39
I don't see that as a reason to start over, that is all
23:39
scott_l | DeFi Pulse
im not just randomly guessing about people's motives
23:40
im talking about actual individuals that are no longer interested in building on this project
23:41
CeloStarter Bandits
Kassel.. kannst mich mal kontaktieren oder im Discord PM danje
23:42
CS https://discord.gg/yTzGDWX
23:42
Zack
I already wrote enough layer-2 tools to carry us through the first 6 months. the javascript light node can use the market smart contract with single-price batching.
At this point it is not important if people want to build on top.
23:42
Mandel Hoff
I suppose releasing the code is one way he can stop without the network freezing. Diff increases going to hit soon.
23:44
Mr Flintstone
@Mandelhoff I don’t think he would stop mining though
23:44
Zack
If you mine with v100s on Amazon and sell the tokens, do you make enough profit to keep mining?
23:44
Mr Flintstone
only if the price goes up a lot each day
23:45
Mandel Hoff
Hard to say - just pointing out it's not necessarily altruistic if he does share code.
23:45
Zack
Kassel says it costs like $1 per veo right now
23:45
Mr Flintstone
wow that is super cheap
23:45
Zack
It seems like they can be sold for $20+ fairly easily
23:45
Even if our mining code is 2.4x less efficient
23:46
Mr Flintstone
30 dollars per hour and he is finding a block every few minutes, that sounds about in the right ballpark @zack
23:46
Zack
it is still almost 10x profit if you push money through the cycle. cash -> rent gpu -> amoveo tokens -> cash
23:47
Mandel Hoff
Good to have some real math to encourage continued gpu mining on the community side.
23:48
Gonzalo
+1 He's protecting his bag full of veos.
23:48
Iridescence
Fuck that explains it
23:48
I was thinking of using the cloud V100s too
23:48
Zack
We should test his claims to see if they are accurate
23:48
Iridescence
But they require a time delay to check why you are wanting to use it
23:49
Nice to be able to spend $1000 a day on mining
23:50
Jim
Or bored deep learning researcher
23:50
Zack
I thougth it was $1 per hour for the gpu?
23:51
Mr Flintstone
If you tell them you want to use gpu compute for mining cryptos will they reject you ?
23:51
O K
I was also wondering this, what kind of customer are they trying to avoid
@Iridescence
23:52
Mandel Hoff
Here's Azure's prices. They are not available to my subscription though tot test.
23:52
Deleted Account
v100 is like 3$ per hour from ama
23:53
Iridescence
$1 per hour is if you commit to 1 year
23:53
I'm getting a Node is not running error
23:55
Iridescence
Even from a fresh clone
23:58
Zack
$3 per hour, times 24 hours per day, times 30 gpu, times 30 days, $64k per month.
23:58
Deleted Account
i got about 500mhs with v100 on ur not upgraded code but didnt seem worth it so stopped
23:59
Iridescence
With the upgraded code it's probably a few GH/s
23:59
Deleted Account
nice
23:59
Iridescence
I don't know, just guessing
23:59
I can try to test if I get approved
00:01
Iridescence
@zack make prod-attach fails with a "Node is not running!" Error when I do a fresh clone
00:02
Gonzalo
did make prod-go before??
00:02
Iridescence
But fuck I can't afford 30 V100s on Amoveo
00:04
Deleted Account
plz gimme upgraded code i try ❤
00:04
Iridescence
Haha sorry, testing period for upgrade is over
00:05
Maybe Kassel's code is better than mine
00:06
O K
What is the advantage to keeping code private?
00:06
Deleted Account
why trust kassel
00:06
"yeah im burning tons of money here"
00:06
Iridescence
I have no incentive yet to make the code public
00:06
O K
More hashrate for your pool, more cut for you
00:06
Deleted Account
right
00:07
i wanna mine iri pool
00:07
Iridescence
Mandel has a better pool than me, he can just copy my code
00:07
Gonzalo
you can lower the fees
00:08
Iridescence
My fees are lower, but mining is not shared
00:08
So it's less appealing
00:08
I'm just using Zack's mining pool code
00:08
Gonzalo
ok, well Im going to switch off my full node, I've no incentives
00:09
O K
Maybe you could make a deal with Mandel to get a % cut to release your code to him
00:09
And spare yourself the labor of running an unshared pool
00:09
Well nothing is going to get done if I can't start my node
00:10
O K
Maybe we can help, what problems are you having
00:10
Deleted Account
what does enter a seed string mean? in mandels miner
00:10
Iridescence
I get a "Node is not running!" error after make prod-attach
00:10
O K
It's to develop randomness so you get fewer nonce collisions
00:11
Iridescence
I'll restart my computer and see if it works
00:11
Deleted Account
so i need to input random characters and press enter?
00:12
It says mash keyboard right
00:12
Deleted Account
yes.
00:13
Iridescence
Same thing happens
00:14
O K
Did you make prod-clean since the hard fork earlier
00:14
Iridescence
Welp, still get the same error after restarting
00:14
I'm cloning a fresh repo straight from github
00:17
Deleted Account
I can support a cloud mining op if someone has the know how to set up an efficient one. Pm me
00:17
Zack
I do make prod-restart
00:17
Deleted Account
My experience has been that aws and the like will ban such operations so I don’t see how the big miner is using cloud
00:18
Zack
It's true. Even if he wasn't burning money, it would be in his incentive to say this.
This project is very high risk. Don't invest too much.
00:18
Maybe we should do a insured crowd funding smart contract to pay you to release it
00:19
Iridescence
Maybe we should =)
00:19
O K
Who all is using the secret Iri code?
00:19
Are the numbers truly that much better from Mandel
1.0.0.3
00:19
Iridescence
I have provided the upgraded code to about 4 testers so far
00:20
The numbers are on my github
00:20
I get 1 GH/s on my 1080
00:20
Greg
Zack, I'm very sorry to propose this idea but I work with crypto everyday and I was wondering, what if Kassel is you. He mines on your pool, have coding advantage (2.4x algo), was here since the start, only you were able to communicate with him. Please prove me the opposite. ❤️
00:22
Zack
It is impossible for me to prove that I don't own crypto.
You are right to be suspicious.
I am not kassel.
00:22
Mike
Zach gets rewarded regardless of who finds the blocks so I don’t see it as advantageous to corner the market
00:22
Greg
lol Zach's is first defensor of decentralized governance so I don't think he will be shocked by my idea lol
00:22
O K
Well Amoveo has to survive a year so he can get the big dev payout
00:22
Zack
Follow the instructions for turning on the node on the README. Then do the instructions for syncing.
Sometimes I forget a step too.
00:23
Iridescence
I figured out the problem
00:23
It's because I upgraded / changed the Erlang install for some other project earlier on
00:24
Zack
Great. I'm glad you figured it out. Can you fix it?
00:24
Mr Flintstone
lol I love this
00:25
Greg
this is funny. I love your project and all the work, will try to better understand the code.
00:25
Denis
You can hire AWS p3 instances at spot price at ~1$
/hour. They uses tesla v100 GPU. Amazon doesn't prohibits mining on their servers
00:26
Zack
Awesome. Thanks for sharing
00:26
Iridescence
It is fixed
00:30
Deleted Account
i actually tried to get p3 awhile back, was rejected
00:31
i guess i can use spot pricing though
00:31
ill have to learn how to do that
00:32
Mike
If anyone can get the cloud mining working and make a guide I’d definitely pay for it. I’ll figure it out this weekend otherwise
00:32
O K
How much are you offering?
00:34
Iridescence
Omg the Amoveo node is running on my vps
00:35
Iridescence
Hmm stuck at block 5210
00:35
Mandel's pool shows the same
00:36
Mike
I’d do at least $100 if it’s simple, I’m sure others are interested in it
00:37
O K
I'm at 5210 as well
00:37
Iridescence
So is Zack's node
00:37
Greg
Kassel stopped mining, new drama
00:37
O K
it's from the fork
00:37
Tarrence van As
Im at 5216
00:38
Iridescence
Oh so 5216 is you
00:38
O K
5216 is headers
00:38
Zack
We undid 80 blocks because I left a big flaw in one of the Oracle transaction types. It is all fixed now.
00:38
Iridescence
But we're stuck at 5210 now?
00:39
O K
the correct blockheight is 5210, the apiheight is 5216, these are the expected numbers right now
00:39
Zack
I got 5211. We aren't stuck
00:39
Just slow in catching back up
00:39
Zack
once we get to 5217, then the fork will be healed and the light node will work normally again.
00:40
Tarrence van As
How do I confirm I'm on the right chain?
00:41
Zack
if you are on the wrong chain, you are stuck at 5216, and the height wont increase
00:41
Tarrence van As
Ok I think I'm on the wrong chain then
00:41
How do I remediate?
00:41
Zack
I wrote a document about updating
00:42
Tarrence van As
Where can I read it
00:42
Tarrence van As
Ok so if I am using the Amoveo mining pool I should be on the right chain?
00:43
Zack
Are you saying that you are running a copy of the open source mining pool software? or do you mean that you are connected to my mining pool?
00:44
Tarrence van As
I am connected to yours
00:44
Using the decryptoed cuda miner
00:44
Zack
since you aren't running a copy of the Amoveo full node, you don't have to do anything when the Amoveo full node updates.
00:45
Denis
AWS has restrictions on how much spot instances you can launch. New accounts has very low limits. Have test launched 3 spot p3.2xlarge in Oregon region, it worked out. But my request for p3.16xlarge with 8 teslas was restricted, guess some developer who often use AWS can easily get bunch of this instances.
00:46
Zack
good to know.
How fast is the hashrate?
00:52
Tarrence van As
@Iridescence How do I upgrade the 1080? I'm getting 150MH/s, seems you are getting more?
00:52
Denis
Haven't tested miner yet, just tested will amazon let me hire some instances. Will give it a try once get back home. Hope
@Iridescence will share his optimised miner with community, or give it for test.
01:03
Gonzalo
What about people mining during this 80 new blocks, are they going to get a block reward?
01:03
Zack
79 blocks were Kassel, one was Mandel's pool.
01:04
im sorry the product isn't perfect. I will try to do better going forward.
01:04
O K
You're doing great
01:06
Greg
You don’t owe us anything
01:07
Mr Flintstone
It doesn’t need to be perfect
01:14
Mike
Down for me too.
01:15
Mandel Hoff
I verified my pool only had 1 payout tx in this time peroid. I credit that person the veo that was rolledback. My pool found 1 block in that region, and Zack said he'd send the pool 1 veo to compensate for that missing credit.
01:15
Zack
yes. thank you for noticing.
01:16
yeah, give me an address
01:18
Mandel Hoff
Please send the block reward to my pool adress. Miners were credit so it'll just payout normal from the pool.BKyLqCvF5vt7KBiPLId7xlW64Bb/uV+icPnvn5mgEKncqlCKiaXUTkRcZv7iYK3GKxSspNh0leIhc4YA+DJAY/s=
01:19
Mike
My payment from 5139 came through it wasn’t showing for a bit
01:21
Mandel Hoff
👍 Glad that worked. Pretty sure that was an easy fix with an obvious payout amount for a single address on my end.
01:22
Greg
So much devotion 👍🏻
01:24
Gonzalo
I love conspiracy theories, better if they cant ever be proved... more fun! I've either reasons to believe that Zack is Batman&Kassel and that he is not. Anyway, for me this is not important at all. Only the smooth running of the network and adoption is important.
01:32
Anyway, If I were Kassel I would balance my miners to work both on Mandels pool and on Zacks, just to not be so noisy and also to allow people mining on Mandels to see sometimes a few satoshis on their unpaid balance. My 2cts
01:32
You hear me, Kassel??? 😂
01:32
Iridescence
Kassel can't hear you over all his GPUs
01:33
Greg
You killed me 🤣
01:33
Iridescence
He's too busy swimming in his VEOs anyway
01:38
Mandel Hoff
Superior capital has this sort of extreme advantages all the time. Hard to know if all of this is true, but does seem the most probable.
01:39
Mr Flintstone
@zack block explorer acting strangely
01:39
Mandel Hoff
Zack, I tried your hash-score function from opencl and it was oddly slower "out of the box". I just put it inline in my kernel. Worked great but went from 110mhs to 73mhs.
01:39
Mr Flintstone
showing block height in 1300/
01:39
Zack
yes. I am having serious issues keeping the node alive.
01:40
Mr Flintstone
what is the issue?
01:40
Zack
oops. I guess my optimization made it worse
01:40
it crashes when I try to send 1 veo to Mandel
01:41
Mandel Hoff
Did I paste in the wrong address? I'll check
01:42
Zack
the wrong address shouldn't make the entire node crash
01:42
Mandel Hoff
BKyLqCvF5vt7KBiPLId7xlW64Bb/uV+icPnvn5mgEKncqlCKiaXUTkRcZv7iYK3GKxSspNh0leIhc4YA+DJAY/s=
01:42
Aizen Sou
Mandel ur pool is down ? A lot of error messages
01:42
Zack
BKyLqCvF5vt7KBiPLId7xlW64Bb/uV+icPnvn5mgEKncqlCKiaXUTkRcZv7iYK3GKxSspNh0leIhc4YA+DJAY/s=
01:42
Mandel Hoff
I think my pool is alive. My miners are doing work.
01:43
Zack
I think it has something to do with how much mining power is pointed at my node.
01:43
Mr Flintstone
It’s an I/o issue?
01:44
don’t miners only send a valid preimage?
01:46
Zack
this time I am turing the mining pool off first, then syncing.
01:48
O K
We should bet on it
01:49
I P
guys sorry to just break in like that, i'm new to this. wanted to mine with gpus but could not find link to the optimized cuda miner
01:49
O K
What OS are you running IP
01:49
Aizen Sou
Lol if Zack were Kassel why did he say about it ? He could keep silent and fool like you will never have a clue what happened behind the scenes.
01:50
Deleted Account
Zack,I would love if you could expand a bit more on this....
01:50
I P
i have both windows and linux machines
01:51
Zack
It is a pretty serious problem if true.
01:51
Zack
I sent the Veo
01:52
on reddit i saw " An upgrade available to provide the most optimized CUDA code for Amoveo GPU mining, and typically gives a 3-5x performance improvement, depending on your GPU"
01:53
O K
It's been applied IP
01:53
I P
ah ok thank again
01:53
Mandel Hoff
Thank you. I see it on block 5221.
01:55
Gonzalo
oh, and why not? At the end this is a big chess play we are playing, so all is possible.
01:55
Anyway I dont believe that Zack is Kassel
01:56
Deleted Account
So you suspect Kassel is doing that... Anyway I said it before if we do restart we should have a good discussion about the mining algo
01:56
Mr Flintstone
I don’t think this is what he is saying
01:56
Deleted Account
But that it can happen
01:56
Mr Flintstone
it’s just that u can apparently mine a memory hard and non memory hard PoW algorithm at the same time
01:57
Greg
I don’t want to play this game but the fact that Zach is taking care of his community doesn’t refute 100% this idea. But like
Gonzalo, it doesn’t matters to me.
01:58
Deleted Account
Nah Zack is not Kassel he has the dev reward no need
02:00
Zack
yes, that about sums it up.
It looks like the same economics as merge mining with namecoin for example.
02:20
Gonzalo
Any update on current pools hashrate?
02:21
Jim
The problem seems to be not that they can be dual mined, but that only one ETH pool is doing it - further centralized and provides unfair advantage
Dan Boabes [RO] joined group by link from Group
02:28
O K
0.835 blocks per minute since launch
02:37
Zack while the PoW is close to sha256, it is not exactly sha256 right?
02:37
Deleted Account
it's called reverse psychology if you don't know the strategy. Zack is pro capitalist. May the strongest wins. 😉
02:38
Iridescence
The pow is sha256
02:39
Mr Flintstone
So sha256(x) == amoveoPoW(x)
02:39
Iridescence
Sha256 is an operation
02:40
What it operates on is different for Amoveo vs Bitcoin
02:40
Mr Flintstone
so why can’t you use bitcoin ASIC?
02:40
O K
Could you expound on this? I have been wondering as well
02:40
Iridescence
Hang on
02:42
Amoveo's header is structured differently
02:42
to mine a block, you have to find a nonce such that the SHA256 of the header has a certain number of zeros
02:42
Mr Flintstone
but, for the same input, the hash amoveo uses is equivalent to sha256?
02:42
Mandel Hoff
sha256 is like an oven. The X that's put in the oven is different for Amoveo than from btc. If X is bread and cheese, you get cheese bread. If X is flour and eggs, you get a cake.
02:42
Iridescence
so if there is an ASIC that is specifically tailored for Bitcoin's header structure you can't use it for Amoveo
02:42
Mandel Hoff
Asics are like ovens that can only take a specific set of ingredients as inputs.
02:42
Mr Flintstone
got it
02:43
I always thought the ASIC could take an arbitrary input
02:43
Iridescence
I think there are general ASICS that can do SHA256 as well
02:43
Mandel Hoff
That's closer to fpga.
02:43
Iridescence
but I think they are not as fast as the bitcoin-specific ones
02:44
Mr Flintstone
makes sense
02:44
Mr Flintstone
so if you were skilled enough, you could take a bitcoin asic, change some of the hardware, and it can mine amoveo?
02:45
Iridescence
it is hard to change an ASIC
02:45
Gonzalo
not if its harcoded, but i dont know for sure
02:45
Iridescence
You need to design the ASIC and then send it to be manufactured
02:46
once the design is set, if you want to change the design, you have to get it re-manufactured
02:46
which can be costly
02:46
TBH I haven't manufactured an ASIC before so I don't know much about it
02:50
Zack
It is single sha 256 with a header shaped different from the bitcoin header
02:51
Bitcoin is double-sha256. And the header is shapped differently.
03:06
Deleted Account
I agree
03:16
Deleted Account
Just the processor with the single sha 256 algo
03:16
Deleted Account
I guess Zack sees it as a possible scenario for Amoveo in the future....
03:18
Iridescence
Zack has explicitly said that he wants ASICs to mine for Amoveo
03:18
Deleted Account
Even though what it's happening now it's very close as Kassel is racking almost all supply
03:18
Iridescence
Personally I disagree that ASICs are a good method of securing POW for a cryptocurrency, but it is his project and he decides as he chooses
03:19
Mike
I think it’s good but I’ve researched the process to making an ASIC a year or two ago and it’s a bitch
03:20
Only someone with manufacturing connections would be able to do it imo
03:42
Greg
Mandel's pool didn't find any block since 2 or 3 hours, we need optimized miners and more hashpower 😒
03:47
Mike
Stop whining. Go read CUDA docs like some of us are doing RN ;)
03:48
Monk
Zack, did your IP chance? I can't find neither the pool, nor the light wallet under your IP any more.
03:48
doesn't work for me
03:58
If you need a node to mine off of temporarily you can use mine
159.65.173.9 but you should switch back to zack's when it's up
04:13
Mandel Hoff
Yes, I think so.
04:29
O K
My node is at 5259
04:30
Greg
the block height seems correct on 'pool stats', don't know if we find blocks but my unpaid balance didn't move since 2 hours. never happened before
04:30
O K
mmm, no that happens to me sometimes
04:31
Was kassel still mining through zacks pool?
04:31
zack's pool being down should give mandel's pool an advantage right?
04:32
Mr Flintstone
Unless Kassel solo mines
04:53
Mandel Hoff
Anyone else running a node? If yes, what's your IP?
04:53
Matthew Slipper
yes
04:53
Zack
sorry I let it die for a while guys. I rebooted just now.
04:53
Matthew Slipper
it's noy syncing right now, though.
04:54
Mandel Hoff
My pool looks right. Everything is "on" and normal sync. No errors anywhere - no blocks for 50 minutes.
04:54
Aizen Sou
@Mandelhoff what is the IP of ur node? Thanks
04:56
So weird that we mined 50 minutes and no blocks.
04:56
Zack
The only error messages I see when it dies are those that appear if someone is trying to contact me too rapidly
04:59
Mandel Hoff
There 5260 just as Zack came back online. Other nodes should be able to mine if your node is down, right?
05:00
That was kind of freaky that the pool got 5260 just as your node came up after 50 min no luck.
05:00
Zack
yes, my node was probably making blocks so quick that others were being blocked
05:00
That is the thing about randomnes
05:00
Greg
we just got 2 blocks! should be variance
05:00
Zack
and humans are made for seeing patterns, so randomness can get us.
05:01
Mandel Hoff
Right... well my node is a good state then. My peers all has many people, so I do see lots of others online.
05:01
Zack
I am rebooted, but I think no miner is connected to me.
05:01
Mandel Hoff
Sure - could be random luck timing.
05:01
Zack
I guess Kassel's miners went down after enough time
05:02
Zack
nevermind, I got a block
05:02
Greg
wrote some nodejs code to extract block information from 8080 API. array parsing was pain in the ass when you are used to json lol
05:03
Mandel Hoff
Oh, it was a 97 minute gap, not 50 minutes.
05:04
Zack
arrays are JSON too. you are just used to json dictionaries
05:04
yeah, it was a great nap too
05:05
Greg
yeah this is was I meant
05:06
I'm using Batman node for this, will not make too much requests, I promise
05:07
Greg
thanks, this will help me display more info
05:15
Tarrence van As
@Zack have you considered using protobufs for data interchange? Would make it very easy to build libraries around the protocol
05:16
Mr Flintstone
are we still at 5261?
05:16
Tarrence van As
yes
05:16
Mr Flintstone
Kassel not mining?
05:23
Zack
you think that protobufs is easier than JSON?
05:23
protobufs is more compressed than JSON.
05:23
Tarrence van As
yes and versioning is easy
05:24
JSON is hard to parse in a lot of langauges
05:25
Zack
what language do you want to do Amoveo stuff in that has a tough time with JSON?
05:26
Mike
Perhaps protobuf is an overkill for simple data exchange. Consider MessagePack:
https://msgpack.org/
05:26
Tarrence van As
I'm just saying generally specing, protobuf makes data much more portable
05:27
Zack
I am a very lazy programmer. I am not going to do an upgrade unless someone has a need for it right now.
05:27
Tarrence van As
Haha fair enough
05:28
Zack
the mining pool is a program that runs on the same computer as Amoveo. it contacts amoveo through http, and it has an http server.
We could make a similar program that just translates between protobuf to JSON, forwarding messages to the Amoveo server, and returning them to you.
05:37
Mandel's pool finally got some blocks. faster than 10 minutes per block too.
05:37
Mr Flintstone
maybe Kassel heard our prayers and started mining on Mandels pool
05:37
Zack
Kassel's miners went down when my node crashed for 100 minutes straight
05:38
I mentioned to him, he should be restarting them shortly
05:43
Mr Flintstone
I hope he mines the right chain cuz I just moved a decent amount lmao
05:43
Zack
You are worried that my mining pool is on the wrong chain? I assure you, it is not.
05:48
Mandel Hoff
Wow, you did just move some veo. Nice to see some action happening.
05:49
Zack
that is 100 veo mr flinstone. good job aquiring so many
05:52
Mr Flintstone
thank you for the assurance
05:55
Mike
@Mandelhoff I have a duplicate “recent received transaction”, differing block #’s
05:55
Iridescence
has Kassel stopped mining?
05:56
Zack
he was off for a while, but I think he is back now.
05:56
Iridescence
my main worry is that Kassel is mining unsustainably
05:57
I don't know how sustainable 30 V100s are XD
05:59
Zack
around $21k a month we estimated earlier.
I think Amoveo can handle a 10x reduction in mining power, it would take about 2.5 months for the block time to go back to normal.
I believe that I can afford the $5.2k of mining costs to do this by selling Veo I have accumulated.
05:59
scott_l | DeFi Pulse
30, try 900
05:59
Zack
he claimed that he is using 30
06:00
Mandel Hoff
Are you the person I manually set payment for on the block rollback about 9 hours ago? When I confirmed the payment, I saw that. If that's you, that's fine - the dupe is in my database, not the block chain.
06:02
scott_l | DeFi Pulse
yea. im aware of the claim.
06:03
Iridescence
Hopefuly I will be able to get a V100 from amazon to test on
06:07
Tarrence van As
What is the supply of VEO right now?
Rachid Ajaja | Will never PM you first joined group by link from Group
06:20
Aizen Sou
Zack ur node is down again
06:20
Zack
thank you for telling me.
06:21
eheap_alloc: Cannot allocate 85562808 bytes of memory (of type "heap").
06:22
it looks like my ram suddenly spiked from 60% to 100% and made me crash.
06:22
Aizen Sou
U need better hardware 😅
06:23
Zack
I think it is a software problem.
06:24
there is no reason that the memory requirement would jump that much in such a short time.
06:24
it happened in less than a minute.
06:24
normally the memory requirement sits around 60% all day
06:24
the only errors I see in the log are that someone was sending requests too frequently to the server.
06:25
Iridescence
I reported the same error to you yesterday
06:26
Is that an internal erlang error, or do you explicitly allocate memory somewhere in the node?
06:30
Zack
I think all 200 page of erlang in Amoveo use at least some RAM.
06:31
I will find out which part is using too much.
06:47
Zack
I found a problem with the request_frequency thing.
If too many people were talking at me too quickly, I think it might suddenly overflow my ram like we saw.
So I fixed that.
This also heavily reduces the number of error messages in the log, so if the bug is still in the code, hopefully next time it happens the log will show me what is wrong.
06:58
O K
First successful update without starting from 0, thanks for the new instructions
06:58
Zack
if you updated just now, you might have broken code. I am doing some experiment.
06:59
well let me know if I can help or when I should update
06:59
Is
@vocarity speaking into this channel?
06:59
Zack
ok, ill let you know.
07:00
Mr Flintstone
looks like we’ve been producing blocks at a reasonable rate
07:03
I didn't even know this was a thing
07:04
I see the feature now
07:08
Zack
looks like the experiment was a success. so your code is fine. we will all be using that code soon.
07:58
Monk
I fired up 4 V100s on Amazon. They reach about 3.3GH with the optimized miner from
@Iridescence, 430MH unoptimized. At 12.20 per hour for four cards, that makes them a bit more expensive than the P100s on Google, which can reach 1.37GH per card, at a price of a bit more than 4$ per hour for four cards.
07:58
still hard to believe that Kassel should get 26GH per card
07:59
O K
Do you get to choose what OS to run?
07:59
Monk
I used Ubuntu, there's a special AMI that comes with Cuda installed.
08:00
There are Windows AMIs for those machines too, but I don't know whether the miner supports Windows yet.
08:00
Zack
that is a lot bigger than the 2.4x improvement that Kassel claimed to have
08:00
O K
Mandel's miner works well on Windows, and Iri is hoarding the "optimized" miner
08:03
Monk
To reach Kassel, I'd have to run 30 p3.16xlarge instances with 8 cards each, cost: $730 per hour. I'm pretty sure mommy will not like Kassel's CC bill at the end of March.
08:04
Cuda 9.1. The "Deep Learning AMI (Ubuntu) Version 5.0" AMI.
08:04
Gonzalo
this is why I feel that $30 per VEO is too cheap.
08:05
a least if you do cloud mining...
08:05
Monk
I was joking. But I think I saw Zack mention that he's using EC2 somewhere.
08:09
EC2 has a relatively bad price/performance ratio, because the V100 has a relatively bad price/performance ratio and because the whole infrastructure around it is overdimensioned for mining. The V100 performs roughly like 3 1080s, but it costs roughly 16 times as much (at least where I live).
08:10
so cost should be divided by at least 5 to get "optimum" mining costs on GPUs
08:10
a dedicated miner that doesn't fully calculate deprecation and his time will get a fair bit lower
08:14
Iridescence
I speculate that Kassel meant that he is running 30 p3.16x large instances (each instance having 8 V100s), and not 30 V100s...
08:16
Zack
sorry, I don't know much miner jargon.
Kassel said this: "Yes they are all from ec2, P3-instances"
08:16
Monk
If you take "optimum" costs (full deprecation and work time, but a real rig in a cheap power environment), it costs about 150 bucks per hour to hash the entire network right now, as a very rough guess. This isn't so irrational, so we'll likely be in for even harder mining conditions.
08:16
Zack
so maybe Iridescence is right?
08:17
Zack
so does that mean he has 240 V100s running?
08:17
Iridescence
that would make more sense
08:17
I believe my optimized miner is close to globally optimal for Amoveo minining
08:18
Iridescence
and we have measured 3.3 GH/s on one V100
08:19
Zack
that would make about 800 gh/s which is about what he has.
240 * 3.3
08:19
So we don't need his code if we can get Iri's code.
08:20
even with my code you still need the hardware
08:20
O K
Some of us do have hardware
08:20
Iridescence
with my unoptimized code V100 gets around 430 MH/s
08:20
MarcoMrm ⛓️🕸️
Kassel has to get a very special discount at Amazon to throw that much power on a experimental project...
08:22
Zack
$730 per hour * 24 hours * 30 days
1/2 million per month.
That is pretty expensive.
And he would have already spent over $30k
08:22
this is exactly what I was warning people to not do.
08:23
Iridescence
Maybe he is a mega whale who doesn't really care
08:23
MarcoMrm ⛓️🕸️
Something is off I would expect that numbers once the project stabilizes
08:23
Zack
but maybe he can afford it
08:23
Iridescence
$100k is small change for some people in this world
08:23
O K
I'm assuming this has paid off for him in the past
08:23
This isn't their first rodeo
08:23
Deleted Account
what % is kassel these days
08:24
Mr Flintstone
permissionless chains will let people do this :)
08:25
Zack
looks like the whole network is 370 gh/s right now.
But my node crashed like 5 times in the last few hours.
08:26
MarcoMrm ⛓️🕸️
strange for a whale to work on a cuda mining algo...
08:26
Iridescence
if you think about, people who would know how to optimize mining algos for CUDA would be the sort of people who are crypto whales
08:26
MarcoMrm ⛓️🕸️
yeah but why bother if you are already rich?
08:27
Mandel Hoff
That's some epic capital. Love the number crunching.
08:27
Mr Flintstone
rich is not an absolute word
08:27
they may feel not rich enough
08:27
Deleted Account
thing is
08:27
Zack
like a drug. the more a person has, the more they tend to want.
08:27
Deleted Account
is you think this mining is bad, you increse the reward
08:28
if the reward changes to 10 per block, this past week has been a burn of money for all miners
08:28
O K
Rich people get rich by having a certain kind of attitude
08:28
Zack
If we destroyed all of Kassel's work and money, then what incentive does anyone have to invest in Amoveo in the future?
They will think we will destroy their work and money again.
08:29
Mr Flintstone
agreed
08:29
Deleted Account
like i said, if you think its bad that he has 90%+ or whatever
08:29
technically, anything over 51% should be very bad
08:29
Iridescence
I think that we want to incentivize sustainable mining
08:29
Mr Flintstone
that wouldn’t change the distribution though Ryan
08:29
He can still keep mining
08:29
Deleted Account
yeah
08:29
Iridescence
if Kassel can sustain this hashpower throughout a whole year, then good for him
08:29
Zack
51% hashpower is bad. 51% coins doesn't matter.
08:29
Iridescence
but if he abruptly stops, then we are screwed
08:30
Deleted Account
and until hes lower than 51% hash power we just keep increasing the rewards
08:30
Iridescence
think about how long it will take for the difficulty to readjust
08:30
Deleted Account
if the hash power distribution is good, you decrease rewards to zero to improve coin scarcity
08:30
Mr Flintstone
his hash rate share is going to go down soon
08:30
Zack
we can hardfork in that situation. It wouldn't be a problem.
08:30
Mr Flintstone
unless he wants to spend even more money
08:30
Deleted Account
well presumably he wants to sell some coins eventually
08:31
but it seems better for us as a community to help eachother get set up with competitive mining contributions
08:31
Mr Flintstone
maybe we could make a market for when his hash rate drops below 50%?
08:31
Deleted Account
i think righ tnow most of us are just terrible miners and either not participating or cant afford to
08:31
the EC2 set up sounds interesting, i want to learn more about it
08:32
if thats really what this other guy is using then any of us can be competitive
08:32
Mr Flintstone
I was cpu mining on ec2 but all of the gpus are permissioned...
08:32
Iridescence
you have half a million a month to spend on cloud GPU mining?
08:33
cant load wallet or mining stats on it
08:33
O K
website works for me, my miners are mining
08:34
Deleted Account
hmm it loads but doesnt show me any stats
08:34
also, the wallet seems down
08:34
Deleted Account
well maybe the miner is just trying to spend a certain amount quickly to acquire a certain % stake
08:35
Deleted Account
this could mean he's spending $50K within the first few days, to acquire his portfolio desired stake, and then stop or reduce mining
08:36
O K
Have we asked him what his intentions are in the future
Zack
08:36
It might be best to just assume he's being honest at this time and ask what his intentions are
08:36
MarcoMrm ⛓️🕸️
doubt it I believe he is not paying that much but no way to know
08:38
Mandel Hoff
Yes, the Azure GPUs are not available to all subscribers too. I can use the K80s on one level, and none on a second level, and none of my Azure accounts allow the highest tier like V100
08:39
Iridescence
yeah, V100 isn't easy to get
08:40
let alone 30 instances of the largest V100 instances
08:43
Deleted Account
interesting
08:46
Gonzalo
the easyest asnwer is the more probable as well, so I believe he is just greedy, and this is fine
08:46
Iridescence
his intentions are probably the same as ours
08:46
he just has more cash to play with
08:47
Zack
yes, thanks for telling me.
08:49
the mining algo does not use much GPU memory
08:52
but I think the matter at hand is not "dual mining", but merely the existence of an actor that has much more capital than others
08:52
after all, spending half a million a month to mine Ethereum doesn't make sense either
d.d joined group by link from Group
08:54
Deleted Account
He might be mining monero or another top coin not necessarily ethereum
08:54
Gonzalo
its all about expectations... if you believe you could make money this way, why not? So I conclude he is very confident with Amoveo
08:54
Zack
people are already spending more than that mining eth.
ether is $800, blocktime is 15 seconds,
So the daily rewards are like $4.6 million.
08:55
if any eth miner can simultaneously mine Veo at the same time for no extra cost, that is a big problem.
08:55
O K
It didn't hurt Sia or Decred like Pascal
08:56
I know, I dual-mined these things with ETH
08:56
Zack
what was the difference between Sia and Decred vs Pascal?
08:56
O K
And it's exaggerating to say it doesn't impact ETH hashrate at all
08:56
It did, it was just worth it, and used more energy and made more heat and made cards less stable
08:57
I don't know, I hadn't heard of pascal closely before that article
08:57
More than one pool?
08:57
The claymore miner was easy to configure with any pool
08:57
Deleted Account
It seems like economics of mining is not an easy task:
= double sha256 is ASIC not resistant
= CPU only algo are incentive for botnets
= GPU algos may be victims of high concentration of merged mining miners...
08:57
Zack
maybe this guy is just trying to get everyone to use his hashing algorithm, and the problem isn't actually that bad.
08:59
I should say, I've never specifically heard of these problems with Sia or Decred, not that it didn't happen for sure
09:00
Deleted Account
What about increasing the reward per block?
09:01
It can probably discourage "rational miners" if they will mine significantly more per kWh in future than now
09:01
Still it doesn't help with centralization ;<
09:04
Mr Flintstone
centralization of coins is not really a problem
09:04
what is a problem is centralization of hash rate
09:04
which will go away over time
09:05
as long as this miner is not malicious I’m not sure there is a problem
09:05
if this miner is malicious then we hard fork
09:06
Gonzalo
Im not 100% sure but I think that I agree...
09:06
Iridescence
I was thinking a bit, and it's not too surprising that somebody has spent this much none in mininf
09:06
Think of Amoveo first 10k blocks as the "ICO"
09:06
Roger That
@Zack - i promised myself i wont butt in here with my opinions anymore but I cant help seeing this dicussion and probably will exit after this to stop myself. you are experiencing an economic attack (like in says in the pascal link and i said yesterday). and an economic attack isn't always meant as an attack but can turn out to be one in certain circumstances like this. i think some of you (and i dont mean any disrespect by saying this) are trying to come up with forced reasons to see this in a different light. already it seems the cloud mining story he gave doesnt add up and you guys have to plug in the holes to make it stand in a wobbly way. if you still want to trust him 100% it's your call but being a skeptic is a wise decision - always. smaller coins wont survive this kind of attack while larger coins will. if you could dual mine monero with an eth rig, won't matter because still you will be probably less than 5% of total hash. so probably thats why sia survived while pascal had issues. what you are facing the exact same issues mentioned in the pascalcoin post and will face the same - community not expanding, listing issues etc if this continues. this is close to a premined coin even if kassel goes away after a while. also if he can spend to dedicate/ dual mine with a rig that costs upwards to 200K / 500K imagine how many accounts he can create here to sway decisions. everyone is driven by greed but that doesn't mean greed is always good. a famous quote exists that says" capitalism is not safe in the hands of capitalists". be a skeptic - always. and ciao and good luck
09:07
Iridescence
There are many people whales that are willing to put 10k, 100k or 1 million into a good project
09:07
And we are all here because we think Amoveo is a good project
09:07
And so large whales are willing to invest that much into mining power to get VEO
09:08
O K
@RogerFun I don't see you as butting in
09:08
Mr Flintstone
your reasoning is part of the reason why I’m still buying
@Iridescence
09:08
Yeah, any opinion is welcome
09:08
so long as you’re not spamming lol
09:09
Zack
I think Emacs is what was crashing my server, not Amoveo.
I guess I should switch to VI.
09:09
Iridescence
I disagree that it is a premine. The large hasher has had to commit significant resources to get VEO
09:10
Mr Flintstone
significant resources in excess of the normal per unit cost to mine as well I believe
09:10
Iridescence
Zack had been a very good dev for the project so far
09:10
Mr Flintstone
Maybe I’m wrong tho
09:10
O K
I agree that a wider distribution would be a positive thing, I think the issue can be addressed by changing the algorithm or with changes going forward, without starting over.
09:11
Mr Flintstone
Possibly part of the solution is to be patient for a few weeks
09:11
this is still less than a week old
09:11
O K
I thought reducing the block reward was an interesting idea
09:11
Zack
Thanks for sharing. Yes, I am also concerned that this attack could be happening.
09:11
O K
until we can have a more level playing field with our optimizations
09:15
Zack
nevermind, it wasn't emacs.
09:15
Roger That
quick addendum : someone mentioned yesterday that folks are complaining because they couldnt make it to the rich list by mining with their CPUs. That's not true. I mined with my CPU and my average GPU without any aspiration to make it to the rich list in the remotest possible sense. but in 3rd day I feel getting pushed out by plain economics then its a different story. Crypto based on my limited understanding so far stands at the intersection of multiple fields of study, economics and psychology are 2 of them. You can only ignore them for so long at your own peril. Tagging my opinions as sour grapes from not making it to the rich list is a bad read of my/ small miner psychology. I was actually starting to lean erlang (and hopefully still will) due to amoveo. so if nothing else, I thank you for that.
09:16
O K
Roger, that was me that made that statement and it was not about you or anyone specifically
09:16
It was a response to someone who said that all I was doing was arguing and "calling people stupid"
09:16
Iridescence
I have learnt more CUDA and a lot about SHA256 because of Amoveo, so thank you too!
09:17
O K
And
@RogerFun I have some experience with GPU mining on a relatively small scale, I encourage you to keep mining
09:17
Zack
I learned a lot about writing secure scalable software, Thank you everyone for creating this opportunity.
09:19
O K
What I was really trying to do
@RogerFun was offer some perspective on supplies, that this is not ripple, and 1 VEO in terms of supply would be about the same as 4886494 ripple
09:19
There isn't much veo
09:19
I'm mining with small amount of hardware getting about a veo a day
09:20
Deleted Account
@the pool has any fee?please
09:21
Zack
I think Mandel's pool is charging 2.9%
you are using windows, right?
Iridescence pool is charging 1%
09:22
Deleted Account
tks if i run a pool,how can i set the fee?
09:29
Deleted Account
thanks
Deleted joined group by link from Group
09:32
what is 80000000 mean?
09:33
Zack
right now it is paying 0.8 Veo
09:33
Deleted Account
that is mean the fee is 20%?
09:34
Zack
yes, the fee is 20% be default.
09:34
Deleted Account
thanks,i got it
09:36
i want build a pool and a group, popularize the veo in china
09:37
Deleted Account
Does the NVIDIA GPU Miner work with a single GPU or multiple? Also can you recommend OC settings in afterburner?
09:40
Gonzalo
Since competence is a good thing, I'm thinking about opening another pool. The cheapest one for sure 😁
09:50
Iridescence
0% fee?
09:50
Negative fees for promotion?
10:01
Gonzalo
no, no, but will charge only 0,69% fee, with a free welcome drink and full of hot CATwoman pool dancers around. And waiters will wear a BATman dress as well.
10:14
block height is 0,it is right?
10:17
Deleted Account
i run the api:height then i wait?
10:18
api:height().
is not even on the page I linked you to.
10:20
Zack
oh, it is on the page
10:34
I just did an upgrade that should make block propagation about 10x quicker.
If you are running a mining pool, this is an update you will want.
10:35
Hopefully this will let multiple pools coexist, even when the blocktime around 15 seconds.
10:36
Kenny
hey zack, is it required for me to have port 8080 open if i want to stay in sync? or is 8080 only needed if i want to serve light clients
10:36
Gonzalo
updated & synced 👍
10:37
Zack
8080 is required to stay in sync.
you can manually return to sync without it.
10:37
but every time you want to sync, you will have to type commands
10:38
Zack
great. thank you Gonzalo
10:39
ok, I will make it more clear. Thanks for the advice