Irontiga
|
|
September 21, 2014, 07:43:03 AM |
|
Thanks. Any body got something else they think that burst needs? (Not things we need Uray for, but rather easy things that I can do)
A chart showing the distribution of recent block winners ( like this) would be useful for verifying the democratizing effect of PoC. Either you must leave that to Uray....or....hmmm...maybe i can do...I'll try...but I don't think I'm good enough at coding yet....
|
|
|
|
burstcoin (OP)
|
|
September 21, 2014, 07:48:36 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped. Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news.
|
BURST-QHCJ-9HB5-PTGC-5Q8J9
|
|
|
yellowduck2
|
|
September 21, 2014, 07:51:45 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped.  Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst   If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news. There is a reason why most people are using small stagger. Because the tool available now take too much time and for some people the merge broke their plot. I am using 2k stagger because I can't go any bigger for gpu plot and the merge program is taking ridiculous amount of time compare to gpu plotting. I wouldn't want to make bigger stagger with existing tool simply because it take way too much time for "optimize" many many TB. For small miner , they can play with it. But bigger miner not possible to use.
|
|
|
|
Chode
Newbie
Offline
Activity: 42
Merit: 0
|
|
September 21, 2014, 07:55:56 AM |
|
@Irontiga - where did you heard about WD being bought by Seagate or you're just joking? The only thing I could find was a post from April 1st, which makes it invalid news...
|
|
|
|
burstcoin (OP)
|
|
September 21, 2014, 07:56:31 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped. Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news. There is a reason why most people are using small stagger. Because the tool available now take too much time and for some people the merge broke their plot. No, the reason most people are using small stagger is because they couldn't be bothered to use 64-bit java. The information to use 64-bit and why was stated clearly many times before the client was even released, and most people ignored it. This capped them to ~500 instead of the 8191 the java plotter allowed which although still not great, is much better.
|
BURST-QHCJ-9HB5-PTGC-5Q8J9
|
|
|
yellowduck2
|
|
September 21, 2014, 07:59:31 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped.  Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst   If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news. There is a reason why most people are using small stagger. Because the tool available now take too much time and for some people the merge broke their plot. No, the reason most people are using small stagger is because they couldn't be bothered to use 64-bit java. The information to use 64-bit and why was stated clearly many times before the client was even released, and most people ignored it. This capped them to ~500 instead of the 8191 the java plotter allowed which although still not great, is much better. I am using 64 bit Java but anything above 2k stagger will crash my gpu driver. I am using NVIDIA 750 ti What we need is a fast merge tool which can merge 5-10 Tb in less than a day. Maybe a gpu merge tool A fast merge tool is important for miner because it help increase hardware life span in long term. The current merge tool is like the CPU plotter. Too slow for big TB. We need an upgrade
|
|
|
|
unsoindovo
Legendary
Offline
Activity: 1932
Merit: 1042
https://locktrip.com/?refId=40964
|
|
September 21, 2014, 08:01:00 AM |
|
Pretty sure that the Win plot optimiser has fubar'd my plots again. Recommend not using.
Anyone who has used the latest cpu plotter can inform of max stagger.
Thanks
yeah the win plot is fucked up, create big stagger also, uselesses Amph Are you saying plotting with high stagger it is usefull? Why you think so? I'm replotting with 50000 stagger size with last cpu plotter... but if you explain me... I stop it. Thanks
|
|
|
|
burstcoin (OP)
|
|
September 21, 2014, 08:07:10 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped.  Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst   If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news. There is a reason why most people are using small stagger. Because the tool available now take too much time and for some people the merge broke their plot. No, the reason most people are using small stagger is because they couldn't be bothered to use 64-bit java. The information to use 64-bit and why was stated clearly many times before the client was even released, and most people ignored it. This capped them to ~500 instead of the 8191 the java plotter allowed which although still not great, is much better. I am using 64 bit Java but anything above 2k stagger will crash my gpu driver. I am using NVIDIA 750 ti What we need is a fast merge tool which can merge 5-10 Tb in less than a day. Maybe a gpu merge tool Well that's a little different situation. That didn't exist until much later on, after many people filled drives with 500 stagger plots. GPU won't help staggering. Staggering is bound by disk seeking. Ideal situation would be plotting to a large ssd, then staggering reading from the ssd and writing to a hdd.
|
BURST-QHCJ-9HB5-PTGC-5Q8J9
|
|
|
Irontiga
|
|
September 21, 2014, 08:21:03 AM |
|
@Irontiga - where did you heard about WD being bought by Seagate or you're just joking? The only thing I could find was a post from April 1st, which makes it invalid news...
Oh wow....I'm an idiot, so sorry! I didn't notice the april fools bit...
|
|
|
|
Amph
Legendary
Offline
Activity: 3248
Merit: 1070
|
|
September 21, 2014, 08:37:38 AM |
|
Pretty sure that the Win plot optimiser has fubar'd my plots again. Recommend not using.
Anyone who has used the latest cpu plotter can inform of max stagger.
Thanks
yeah the win plot is fucked up, create big stagger also, uselesses Amph Are you saying plotting with high stagger it is usefull? Why you think so? I'm replotting with 50000 stagger size with last cpu plotter... but if you explain me... I stop it. Thanks 50k is fine, but 1M is too high, with uray pool
|
|
|
|
Amph
Legendary
Offline
Activity: 3248
Merit: 1070
|
|
September 21, 2014, 08:38:16 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped.  Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst   If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news. There is a reason why most people are using small stagger. Because the tool available now take too much time and for some people the merge broke their plot. No, the reason most people are using small stagger is because they couldn't be bothered to use 64-bit java. The information to use 64-bit and why was stated clearly many times before the client was even released, and most people ignored it. This capped them to ~500 instead of the 8191 the java plotter allowed which although still not great, is much better. I am using 64 bit Java but anything above 2k stagger will crash my gpu driver. I am using NVIDIA 750 ti What we need is a fast merge tool which can merge 5-10 Tb in less than a day. Maybe a gpu merge tool A fast merge tool is important for miner because it help increase hardware life span in long term. The current merge tool is like the CPU plotter. Too slow for big TB. We need an upgrade you can replot faster than gpu plotter, with the last cpu plotter(13k nonce/minutes on my i5 haswell)
|
|
|
|
fanepatent
|
|
September 21, 2014, 08:42:04 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me
|
BURST - BURST-58XP-63WY-XSVQ-ASG9A
|
|
|
vaxman
Member
Offline
Activity: 99
Merit: 10
|
|
September 21, 2014, 08:57:49 AM |
|
I am using 64 bit Java but anything above 2k stagger will crash my gpu driver. I am using NVIDIA 750 ti What we need is a fast merge tool which can merge 5-10 Tb in less than a day. Maybe a gpu merge tool A fast merge tool is important for miner because it help increase hardware life span in long term. The current merge tool is like the CPU plotter. Too slow for big TB. We need an upgrade
Well, to merge some 10 TB/day the data rate would have to be ~127 MB/s reading AND writing concurrently. As the process is all about de-interleaving data, the reads naturally are scattered all about the disk - the access times and data rate depend on the original stagger size, your disks RPM and possibly filesystem fragmentation. Then you need enough RAM to hold a decent chunk to be written sequentially. I'd say that for stagger 2048 the data rate can't be higher than 20 MB/s on a 7200 rpm drive, so there is the limiting factor, it is not the software.
|
|
|
|
MsCollec
Legendary
Offline
Activity: 1400
Merit: 1000
|
|
September 21, 2014, 09:13:02 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me Funny, Someone already sold it to me Don't follow the crowd, be a smart trader.
|
|
|
|
Amph
Legendary
Offline
Activity: 3248
Merit: 1070
|
|
September 21, 2014, 09:15:11 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me Funny, Someone already sold it to me Don't follow the crowd, be a smart trader. at 5k? i doubt you bought at that price
|
|
|
|
MsCollec
Legendary
Offline
Activity: 1400
Merit: 1000
|
|
September 21, 2014, 09:18:17 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me Funny, Someone already sold it to me Don't follow the crowd, be a smart trader. at 5k? i doubt you bought at that price No. Just paid little over the market price.
|
|
|
|
fanepatent
|
|
September 21, 2014, 09:22:55 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me Funny, Someone already sold it to me Don't follow the crowd, be a smart trader. at 5k? i doubt you bought at that price No. Just paid little over the market price. Well, if you bought over the market price why didn'y you buy on the exchange? It would have helped the price a bit
|
BURST - BURST-58XP-63WY-XSVQ-ASG9A
|
|
|
familiaverde
|
|
September 21, 2014, 09:24:50 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me Funny, Someone already sold it to me Don't follow the crowd, be a smart trader. at 5k? i doubt you bought at that price No. Just paid little over the market price. Well, if you bought over the market price why didn'y you buy on the exchange? It would have helped the price a bit Dont worry ... im buying a lot at 130-170 satoshi
|
|
|
|
yellowduck2
|
|
September 21, 2014, 09:36:45 AM |
|
For the most part small miners are of the game, this coin didn't go the way Id hoped.  Servers and data centers took over.
What's wrong with getting your own server? I'm pretty sure that there was a lot of GPU buying going on when all the other coins came along. But not only GPU's, but also efficient CPU's and mobo's to support lots of GPU's. I hope you're not scared of buying a server.. As far as data centers go, I doubt it. They usually don't stock up large on HDD's, trust me, I've looked at renting high storage servers, they're all either unavailable, or waayy too expensive. Also, running huge data center for this is a very big risk because of hard disk failure and reducing life span mining this coin 24/7. 6 months down the road. We will see how many people post their HDD die off here. We will know which brand is the best and worst   If they can't break even before hard disk eventual dead , they loss huge amount of money i think HDD mining downside compare to GPU is the hardware turn over rate. HDD turn over rate should be 2 - 3 times higher than GPU. In longer term, HDD mining could be more expensive than GPU mining because of hard ware failure I put my bet on seagate being able to break even before it die seagate is the worst i ever bought. had to rma 12 of 32 hdds during the last 2 years. wd only 2 of 60. Toshiba is the worst. Bought 6 of them once, 50% failed in under a week. I think mining relating failed is going to be directly related to people's stagger. If plots are optimized, I doubt a single sequential read of 1/4096th of the capacity every ~4 minutes is really going to do much, but the lower stagger sizes many people are using is bad news. There is a reason why most people are using small stagger. Because the tool available now take too much time and for some people the merge broke their plot. No, the reason most people are using small stagger is because they couldn't be bothered to use 64-bit java. The information to use 64-bit and why was stated clearly many times before the client was even released, and most people ignored it. This capped them to ~500 instead of the 8191 the java plotter allowed which although still not great, is much better. I am using 64 bit Java but anything above 2k stagger will crash my gpu driver. I am using NVIDIA 750 ti What we need is a fast merge tool which can merge 5-10 Tb in less than a day. Maybe a gpu merge tool A fast merge tool is important for miner because it help increase hardware life span in long term. The current merge tool is like the CPU plotter. Too slow for big TB. We need an upgrade you can replot faster than gpu plotter, with the last cpu plotter(13k nonce/minutes on my i5 haswell) That is pretty good. What is the exact parameter u run on the bat ?
|
|
|
|
AnonymousEconomist
Full Member
Offline
Activity: 154
Merit: 100
Add me on Twitter! @AnonOnAMoose
|
|
September 21, 2014, 09:42:45 AM |
|
WTB burst. Large number 250k and above
What price? PM me your offers You can't sell large number on the exchange without driving down the price Looks like I'm the only one willing to sell to you. I hope 5k sat is a price you're willing to work with. 4900 sats last price from me Funny, Someone already sold it to me Don't follow the crowd, be a smart trader. So you are insulting whoever sold you the BURST? Only one person wins in a trade. If you are calling yourself the smart trader, you are implying he/she is stupid.
|
Add me on Twitter! @AnonOnAMoose
|
|
|
|