Author

Topic: ASICMINER Speculation Thread - page 260. (Read 808905 times)

sr. member
Activity: 362
Merit: 250
June 18, 2013, 12:31:12 PM
Quote
Current Bitcoin can't be sustainably much over $400.

Justification: the 1MB block size limit and the fact that the price has pretty much grown hand in hand with the number of transactions per block (in the long time trend). We have free transactions now (for old enough coins). Eventually a minimum transaction fee for ANY transaction will be several dollars, when the block space scarcity really starts to constrict usage. If BTC adoption grows at the current rate, in a couple of years this scenario is going to be real. This will mean, by the way, that the block subsidy will be insignificant far sooner than most people think. At 20,000 tx/block and $1 fee per tx, assuming $400/BTC and current reward level, fees will be two thirds of total miner revenue.

This gives absolutely no explanation for the $400 figure.

Within the last year the number of transactions per day went from ~40k to ~60k (+50%), in the same time the usd/btc price went from ~6.5usd to ~110usd (+1692%). Stating these two values have "pretty much grown hand in hand" is a lie of epic proportions.

Wild theorizing on my part, take it with a truck load of salt: TX fees will not raise to several usd/tx even if the block size limit was not lifted. The reason is that a majority of all transactions will be relatively small sums and people will simply refuse to pay such high fees. Eventually average small time users would rather drop bitcoin and the network would be left exclusively to the movement of larger sums, this development would rather decrease the number of transactions than increase it.

Again, where does the 400 come from??

I'm also waiting for someone to connect dots here. How exactly does the "transactions-per-time-limit" affect the USDBTC exchange rate?
full member
Activity: 294
Merit: 100
June 18, 2013, 11:22:04 AM
I meant adjusting the difficulty to allow for 1 block every 5 minutes, as well as cutting the reward in half. The net effect is the same for miners (ASICMINER), but would this allow for greater throughput?

Ian
Like I said earlier, the reward only affects throughput by affecting change in hashrate and influencing transaction selection.
1 block every 5 minutes would double the possible transactions / second, but it would also double the possible blockchain size increase speed.

Okay, I understand. Thank you for the information.

Ian
hero member
Activity: 784
Merit: 1000
0xFB0D8D1534241423
June 18, 2013, 11:13:16 AM
I meant adjusting the difficulty to allow for 1 block every 5 minutes, as well as cutting the reward in half. The net effect is the same for miners (ASICMINER), but would this allow for greater throughput?

Ian
Like I said earlier, the reward only affects throughput by affecting change in hashrate and influencing transaction selection.
1 block every 5 minutes would double the possible transactions / second, but it would also double the possible blockchain size increase speed.
full member
Activity: 294
Merit: 100
June 18, 2013, 11:00:26 AM
So much for the AM speculation thread...

I'm not sure what to speculate about until we figure out why the hashrate is hovering around 10 th/s over the past 24 hours.   Smiley

With that said -- the two hour number is ~35 th/s. However, that is a rather small sample size.
full member
Activity: 294
Merit: 100
June 18, 2013, 10:57:43 AM
What would the net effect be of a 5M max block size
It would allow for, but not necessitate, 5 times the blockchain size rate of increase and 5 times the transaction rate.

reward halving
The reward has no effect on the block size, except for the fact that it increases the importance of fees.

and difficulty retarget of 1 block/5 minutes?
These are conflicting things. The difficulty retarget happens every n blocks. The block rate is a (blocks / minute) number.

Sorry -- I was unclear. I meant adjusting the difficulty to allow for 1 block every 5 minutes, as well as cutting the reward in half. The net effect is the same for miners (ASICMINER), but would this allow for greater throughput?

Ian
member
Activity: 67
Merit: 10
June 18, 2013, 10:49:43 AM
Quote
Current Bitcoin can't be sustainably much over $400.

Justification: the 1MB block size limit and the fact that the price has pretty much grown hand in hand with the number of transactions per block (in the long time trend). We have free transactions now (for old enough coins). Eventually a minimum transaction fee for ANY transaction will be several dollars, when the block space scarcity really starts to constrict usage. If BTC adoption grows at the current rate, in a couple of years this scenario is going to be real. This will mean, by the way, that the block subsidy will be insignificant far sooner than most people think. At 20,000 tx/block and $1 fee per tx, assuming $400/BTC and current reward level, fees will be two thirds of total miner revenue.

This gives absolutely no explanation for the $400 figure.

Within the last year the number of transactions per day went from ~40k to ~60k (+50%), in the same time the usd/btc price went from ~6.5usd to ~110usd (+1692%). Stating these two values have "pretty much grown hand in hand" is a lie of epic proportions.

Wild theorizing on my part, take it with a truck load of salt: TX fees will not raise to several usd/tx even if the block size limit was not lifted. The reason is that a majority of all transactions will be relatively small sums and people will simply refuse to pay such high fees. Eventually average small time users would rather drop bitcoin and the network would be left exclusively to the movement of larger sums, this development would rather decrease the number of transactions than increase it.

Again, where does the 400 come from??
hero member
Activity: 518
Merit: 500
June 18, 2013, 10:49:02 AM
So much for the AM speculation thread...
hero member
Activity: 784
Merit: 1000
0xFB0D8D1534241423
June 18, 2013, 10:48:26 AM
What would the net effect be of a 5M max block size
It would allow for, but not necessitate, 5 times the blockchain size rate of increase and 5 times the transaction rate.

reward halving
The reward has no effect on the block size, except for the fact that it increases the importance of fees.

and difficulty retarget of 1 block/5 minutes?
These are conflicting things. The difficulty retarget happens every n blocks. The block rate is a (blocks / minute) number.
full member
Activity: 294
Merit: 100
June 18, 2013, 10:32:49 AM
Thanks for the follow-up.

Maybe I'm a bit simple minded, but I do not see how this directly correlates to the BTC USD exchange rate.

I do appreciate the info, however.
Visa handles, at peak, ~4000 transactions per second. With Bitcoin, there are multiple solutions, but none of them involve doing nothing and continuing to use the blockchain alone:

- Move most transactions off-chain
- Increase MAX_BLOCK_SIZE
- Increase blocks per day
- Make smaller transactions somehow


What would the net effect be of a 5M max block size, reward halving, and difficulty retarget of 1 block/5 minutes?

Ian
hero member
Activity: 784
Merit: 1000
0xFB0D8D1534241423
June 18, 2013, 10:26:21 AM
Thanks for the follow-up.

Maybe I'm a bit simple minded, but I do not see how this directly correlates to the BTC USD exchange rate.

I do appreciate the info, however.
Visa handles, at peak, ~4000 transactions per second. With Bitcoin, there are multiple solutions, but none of them involve doing nothing and continuing to use the blockchain alone:

- Move most transactions off-chain
- Increase MAX_BLOCK_SIZE
- Increase blocks per day
- Make smaller transactions somehow
full member
Activity: 294
Merit: 100
June 18, 2013, 10:16:36 AM
I've got no horse in this race, but I believe BTC's max is currently 7 T/S

Follow-up question -- why is there a 7 transaction/s limit?
10min blocks* -> 6 blocks / hour
(6 blocks / hour)(1 hour / 3600 seconds) = 0.0016667 blocks / second

Max block size = 1 MB
(0.0016667 blocks / second)(1 MB / block) = 0.0016667 MB/s
(0.0016667 MB / second)(1,000,000 bytes / 1MB) = 1666.7 bytes / second

According to dooglus, there are (in*148 + out*34 + 10 plus or minus 'in') bytes in a transaction. This equals a minimum tx size of 180 bytes, and a 1-input 2-output minimum size of 225 bytes. Example.

(1666.7 bytes / second)(1 tx / 180 bytes) = 9.26 tx/s
(1666.7 bytes / second)(1 tx / 225 bytes) = 7.41 tx/s

*Note that an ever-increasing hashrate (like the current scenario) means blocks are generated faster, on average, than once every 10 minutes, and a decreasing network hashrate would slow block creation.

Thanks for the follow-up.

Maybe I'm a bit simple minded, but I do not see how this directly correlates to the BTC USD exchange rate.

I do appreciate the info, however.
hero member
Activity: 784
Merit: 1000
0xFB0D8D1534241423
June 18, 2013, 10:10:22 AM
I've got no horse in this race, but I believe BTC's max is currently 7 T/S

Follow-up question -- why is there a 7 transaction/s limit?
10min blocks* -> 6 blocks / hour
(6 blocks / hour)(1 hour / 3600 seconds) = 0.0016667 blocks / second

Max block size = 1 MB
(0.0016667 blocks / second)(1 MB / block) = 0.0016667 MB/s
(0.0016667 MB / second)(1,000,000 bytes / 1MB) = 1666.7 bytes / second

According to dooglus, there are (in*148 + out*34 + 10 plus or minus 'in') bytes in a transaction. This equals a minimum tx size of 180 bytes, and a 1-input 2-output minimum size of 225 bytes. Example.

(1666.7 bytes / second)(1 tx / 180 bytes) = 9.26 tx/s
(1666.7 bytes / second)(1 tx / 225 bytes) = 7.41 tx/s
Most transactions are bigger, so these are high estimates.

*Note that an ever-increasing hashrate (like the current scenario) means blocks are generated faster, on average, than once every 10 minutes, and a decreasing network hashrate would slow block creation. This is due to the fact that difficulty is only calculated periodically (every 2016 blocks).
full member
Activity: 121
Merit: 100
June 18, 2013, 10:04:59 AM
Here's my math:

Max block size is 1 meg.
Minimum transaction size is 154 bytes (1 input, 1 output)
Block frequency is ideally 1 every 600 seconds.

1meg / 154 / 600 = about 11 transactions per second.

11 transactions/sec * 86400 sec/day = 935k transactions/day max.

That's the absolute ideal, however, the average transaction size over the last 1000 blocks is 460 bytes (retrieved from here: http://blockexplorer.com/q/avgtxsize)
Using that size we get 4 transactions/sec and 310k transactions/day.



full member
Activity: 294
Merit: 100
June 18, 2013, 09:51:52 AM
I've got no horse in this race, but I believe BTC's max is currently 7 T/S

Follow-up question -- why is there a 7 transaction/s limit?
legendary
Activity: 1386
Merit: 1000
June 18, 2013, 09:51:15 AM
Here is the link: https://bitcointalksearch.org/topic/m.2220582

And there are other more technical threads with the same debate and more math involved but generally speaking if Max Block Size is not increased then transactions per second cannot increase and how are you going to have Bitcoin reach $1000 a Bitcoin and surpass Paypal or Visa / Mastercard when there are these sorts of caps? It's basically the same stupid problem Mt Gox has with it's limit on transactions per second.

Thanks.

I'm wondering if blocks are even getting close to the max block size? A quick search didn't reveal anything even close.

What is the max transactions per second?

I've got no horse in this race, but I believe BTC's max is currently 7 T/S
full member
Activity: 294
Merit: 100
June 18, 2013, 09:47:38 AM
Here is the link: https://bitcointalksearch.org/topic/m.2220582

And there are other more technical threads with the same debate and more math involved but generally speaking if Max Block Size is not increased then transactions per second cannot increase and how are you going to have Bitcoin reach $1000 a Bitcoin and surpass Paypal or Visa / Mastercard when there are these sorts of caps? It's basically the same stupid problem Mt Gox has with it's limit on transactions per second.

Thanks.

I'm wondering if blocks are even getting close to the max block size? A quick search didn't reveal anything even close.

What are the max transactions per second?
hero member
Activity: 714
Merit: 510
June 18, 2013, 09:24:02 AM
Actually I do and there are discussions about this in other threads. Search the forum and see for yourself.

This is the most terrible response that anyone could post.

I hate googling for something and clicking a link to a post that says "OMFG SEARCH NOOB".. hah

Can you provide a link? I do not see a correlation between BTC to USD & Max Block size.

Thanks.
Here is the link: https://bitcointalksearch.org/topic/m.2220582

And there are other more technical threads with the same debate and more math involved but generally speaking if Max Block Size is not increased then transactions per second cannot increase and how are you going to have Bitcoin reach $1000 a Bitcoin and surpass Paypal or Visa / Mastercard when there are these sorts of caps? It's basically the same stupid problem Mt Gox has with it's limit on transactions per second.

full member
Activity: 294
Merit: 100
June 18, 2013, 09:20:30 AM
Actually I do and there are discussions about this in other threads. Search the forum and see for yourself.

This is the most terrible response that anyone could post.

I hate googling for something and clicking a link to a post that says "OMFG SEARCH NOOB".. hah

Can you provide a link? I do not see a correlation between BTC to USD & Max Block size.

Thanks.
hero member
Activity: 714
Merit: 510
June 18, 2013, 09:19:01 AM
Max Block Size and the limit on the amount of transactions per second means it cannot go over $500 unless this is fixed.

You have clearly no clue what you are talking about.

Actually I do and there are discussions about this in other threads. Search the forum and see for yourself.
sr. member
Activity: 476
Merit: 250
June 18, 2013, 09:04:06 AM
While I appreciate the optimism here, I can't help but be suspicious at the current prices when the demand is so thin.

Of course, today is Tuesday, so the price will probably rise a bit, and stay around 2.9.

I am eager to see what it does after the dividend announcement tomorrow.

meanwhile, hashrate is dropping like a stone.  Huh  We haven't been this low since the beginning of June.
Jump to: