• Forum has been upgraded, all links, images, etc are as they were. Please see Official Announcements for more information

Dash versus Bitcoin versus Darkcoin wallet update benchmark times

Sub-Ether

Sub-Ether

Well-known member
This is a work in process so will be updated as more data comes in from the latest wallet versions..

For the purposes of clarity and to help measurements when measuring speeds of different coin's blockchains when updating, allow me to introduce a new term which is a dimensionless time based ratio called Tao, this is defined as the wallet update time ratio,
where Tao = (seconds in real time spent updating) / (hours of blockchain data behind in wallet)

Local internet connection speed 100 megabits/second.
Machine 1 : Windows 7 64 bit, SSD (500 megabytes/s up+down, SATA 3) 4.4 GHz i7 4820k
Machine 2 : Windows 7 64 bit, HDD (3 gigabits, SATA 2), 2.4 GHz Intel Duo E4600

7th March 2015

Machine 1,SSD: Darkcoin (0.11.1.25) = Tao ratio = 1.346 seconds / hour of update + ( 9.85 seconds indexing startup time )
Machine 1,SSD: Bitcoin (0.10.0) = Tao ratio = 3.304 seconds / hour of update + ( 34.4 seconds indexing startup time )
Darkcoin is 2.4 times quicker than bitcoin

Machine 2,HDD : Darkcoin (0.11.1.25) = Tao ratio = 3.817 seconds per hour of update + ( 19.07 seconds indexing startup time )
Machine 2,HDD : Bitcoin (0.10.0) = Tao ratio = 26.14 seconds per hour of update + ( 838 seconds indexing startup time )
Darkcoin is 6.9 times quicker than bitcoin on traditional 3 gig sata hard drives of average cpu spec.

27th March 2015

Machine 1,SSD: Dash (0.11.2.10) = Tao ratio = 0.5434 seconds / hour + (15.63 seconds indexing startup time)
Machine 1,SSD: Bitcoin (0.10.0) = Tao ratio = 4.663 seconds / hour + (62.32 seconds indexing startup time)
Dash is 8.581 times quicker than bitcoin on fast solid state memory machines

Machine 2,HDD : Dash (0.11.2.10) = Tao ratio = 1.587 seconds / hour + (32.76 seconds indexing startup time)
Machine 2,HDD : Bitcoin (0.10.0) = Tao ratio = 9.645 seconds / hour + (623.4 seconds indexing time)
Dash is 6.078 times quicker than bitcoin on traditional 3 gig sata hard drives of average cpu spec.

Observations:

Dash has smashed the Tao update speed barrier of unity 1, currently Tao sets a new record = Tao ratio = 0.5434 seconds / hr !!
The cpu of the HDD machine maxs out at 100% with bitcoin, indexing goes very slow especially on the first run from cold start and gives spurious update times (Darkcoin is totally stable at 40% CPU)
Darkcoin update times were statistically closer together and more stable especially on the HDD.
Finding nodes to connect to with bitcoin takes longer than with Dash.
On the HDD dual core machine, the entire Dash blockchain of 427 days downloaded and updated quicker than only 19 days of bitcoin (32.6 times quicker!)

Conclusions:

a) For i7 SSD, Darkcoin (version 0.11.1.25) was 2.4 times quicker than improved bitcoin, now Dash (unreleased version 0.11.2.10) is 8.581 times quicker for updating than bitcoin and also indexing and startup is 4 times quicker.
b) Fast modern machines and solid state drives work well with the bitcoin faster hop times but older machines do not like it, and use too many resources bottle necking the bus bandwidth.
c) Bitcoin blockchain length is too large for older,slower machines with less bus and memory bandwidth, especially on a first time windows start of the day.
d) Bitcoin is slowing down and Dash is getting quicker, best Dash Tao ratio = 0.5434 seconds/hour, best bitcoin Tao ratio = 4.663 seconds/hour.
e) for the entire blockchain update, Dash is only 14% slower on old HDD dual core machine compared to new fast SSD setup,
so it is strongly resistant to high latency time traditional hard drive powered setups with limited internet, bus and memory bandwidth.
f) bitcoin indexing and startup on magnetic hard drives from cold start is very slow, Dash starts up 19 times quicker!

Speculations:

The 60 times speedups to bitcoin's hop times have been filled to the brim in the memory pool/buffers and now the bitcoin transactions limit of 7 per second has already maxed out.
It is invitable given the approach that this will only get worse over time.

Sundry

Total time for entire blockchain of Dash (v0.11.2.10) to download on SSD i7 = 42 minutes, complete Tao ratio = 0.2476 seconds / hour
Total time for entire blockchain of Dash to download on dual core HDD machine = 50.5 minutes, complete Tao ratio = 0.2956 seconds / hour
 
Last edited by a moderator:
Tao Sync ratio and update times for Dash version 0.11.2.23

10th May 2015

Time behind = 15 weeks 5 days 9 hrs = 2649 hours
Time to update and fully sync = 10 minutes and 32.74 seconds = 632.74 seconds (additional 16.12 seconds startup time)
Machine 1,SSD: Dash (0.11.2.23) = Tao sync ratio = 0.2389 seconds / hour + (16.12 seconds indexing startup time)
Machine 1,SSD: Bitcoin (0.10.0) = Tao sync ratio = 4.663 seconds / hour + (62.32 seconds indexing startup time)

The Dash sync ratio is 19.52 times quicker than Bitcoin , this is an increase of 2.28 times more than Dash version 0.11.2.10

The Dash sync ratio is 59.51 times quicker than Monero, this is an increase of 2.28 times more than Dash version 0.11.2.10

Monero Wallet

Time behind = 51.383 hours
Time to update and sync = 12 minutes and 11 seconds
Tao sync ratio = 14.22 seconds / hour

Sundry

First full year of the blockchain to download and update = 26 minutes 2 seconds
Machine 1,SSD: Dash (0.11.2.23) = Tao sync ratio for first year = 0.1788 seconds / hour
 
Last edited by a moderator:
With version 0.11.0 of bitcoin coming out and our version 0.12.0 of dash with all the new features, I figured it was update time:-

Tao Sync ratio and update times for bitcoin (v.0.11.0) and Dash (v.0.12.0.49)

28th August 2015

Machine 1,SSD: Bitcoin (v0.11.0) = Tao ratio = 4.144 seconds / hour + (9.3 seconds indexing startup time, needed 3 runs for best time)

Machine 1,SSD: Dash (v0.12.0.49) = Tao ratio (last 12 weeks) = 0.1345 sec / hour (and 14.5 seconds startup)
Machine 1,SSD: Dash (v0.12.0.49) = Tao ratio (last 10 days) = 0.1688 sec / hour (and 13.9 seconds startup)
Average Tao sync ratio from 2 samples = 0.1517 sec / hour


Comparisions

Average Tao bitcoin to Tao dash ratio = (4.144 / 0.1517) = 27.32 times faster
(best tao / tao ratio = 4.144 / 0.1345 = 30.8 times faster)
Average Tao Monero to Tao Dash ratio = (14.22 / 0.1517) = 93.7 times Faster
(best tao / tao ratio = 14.22 / 0.1345 = 105.7 times faster)


Total time for entire Dash (0.12.0.49) blockchain = 35 minutes and 2 seconds
Machine 1,SSD: Dash (0.12.0.49) = Tao sync ratio for complete wallet = 0.1507 seconds / hour
% Tao sync improvement from earlier Tao ratio (v0.11.2.23) = (0.2389 - 0.1517 ) /0.2389 = + 36.5%
Download time improvement from Dash (v0.11.2.10) in March 27th to Dash (v0.12.0.49) 28th August= +16.6 %

Observations:-

Bitcoin locks onto outdated wallet peers such as 0.8.2.2 and 0.9.2.1 probably resulting in a slower update times, where are the fast seed nodes ??


upload_2015-8-3_14-30-7.png
 
Last edited by a moderator:
Sub-Ether said:
With version 0.11.0 of bitcoin coming out and our version 0.12.0 of dash with all the new features, I figured it was update time:-

Tao Sync ratio and update times for bitcoin (v.0.11.0) and Dash (v.0.12.0.49)

28th August 2015

Machine 1,SSD: Bitcoin (v0.11.0) = Tao ratio = 4.144 seconds / hour + (9.3 seconds indexing startup time, needed 3 runs for best time)

Machine 1,SSD: Dash (v0.12.0.49) = Tao ratio (last 12 weeks) = 0.1345 sec / hour (and 14.5 seconds startup)
Machine 1,SSD: Dash (v0.12.0.49) = Tao ratio (last 10 days) = 0.1688 sec / hour (and 13.9 seconds startup)
Average Tao sync ratio from 2 samples = 0.1517 sec / hour


Comparisions

Average Tao bitcoin to Tao dash ratio = (4.144 / 0.1517) = 27.32 times faster
(best tao / tao ratio = 4.144 / 0.1345 = 30.8 times faster)
Average Tao Monero to Tao Dash ratio = (14.22 / 0.1517) = 93.7 times Faster
(best tao / tao ratio = 14.22 / 0.1345 = 105.7 times faster)


Total time for entire Dash (0.12.0.49) blockchain = 35 minutes and 2 seconds
Machine 1,SSD: Dash (0.12.0.49) = Tao sync ratio for complete wallet = 0.1507 seconds / hour
% Tao sync improvement from earlier Tao ratio (v0.11.2.23) = (0.2389 - 0.1517 ) /0.2389 = + 36.5%
Download time improvement from Dash (v0.11.2.10) in March 27th to Dash (v0.12.0.49) 28th August= +16.6 %

Observations:-

Bitcoin locks onto outdated wallet peers such as 0.8.2.2 and 0.9.2.1 probably resulting in a slower update times, where are the fast seed nodes ??


View attachment 1681
Click to expand...
Tweeted.

https://twitter.com/taoofsatoshi/status/637576753465176064
 
I appreciate the effort, but this Tao unit doesn't really tell you much about how efficient different clients sync up, because you're not taking into account the amount of transactions it has to verify. It's no suprise that Tao_Dash < Tao_Bitcoin because the average Tx/Hour on the Bitcoin blockchain is much higher.
 
wsmithston said:
I appreciate the effort, but this Tao unit doesn't really tell you much about how efficient different clients sync up, because you're not taking into account the amount of transactions it has to verify. It's no suprise that Tao_Dash < Tao_Bitcoin because the average Tx/Hour on theI do agree with you that until Dash is under equal load, the test could be said to be not on a level playing field. Bitcoin blockchain is much higher.
Click to expand...
Perhaps I wasn't clear what Tao was aimed as representing, it is not meant to be efficiency as such although there are similarities, as you infer, strictly speaking efficiency could be described something like :-

efficiency = (current_tx) / (Max_tx) * 100
if each tx >= 2kbytes then Max_tx = 1 meg / 2k = 500 per block = 100% efficiency

But the way I see it, Tao does take into account the number of transactions per hour, because if the network is under heavy load (e.g 0.9 megabyte blocks for bitcoin), the value will increase, but the result is 'weighted' and none linear as opposed to directly derived from tx/hour like with running at maximum with every block full.

Tao was meant to show the 'quality' factor of the sync speeds. Compare in the amplifiers, how well they react to sharp sounds with slew effect etc,basically how responsive the 'circuit' reacts to a given signal level over a timeline.
A high 'Q' of say 200 in electronics is very good and considered highly tuned where as a 'Q' = 4 is poor.
Using this analogy Tao = 1 / Q but I wanted the value to lesson as the network became more tuned to show increasing performance, then the value would tend towards zero.
Do you see what I mean?

http://www.radio-electronics.com/info/formulae/q-quality-factor/basics-tutorial.php


 
You must log in or register to reply here.
Back
Top