Vertcoin - 1 | ASIC - 0 | Lyra2RE | Decentralised | GPU Mineable | Open Source

[BBristow79]

What is the power consumption of 280x (or 7970) with Lyra2RE? (w/ overclocked and w/o)

It is higher than X11, Neoscrypt, but less than scrypt.

I thought there were tests done that showed Lyra2 was less than X11? Perhaps it's different by make/model of gpu's?

[1714285250]

1714285250

[1714285250]

1714285250

[1714285250]

1714285250

[MadGhost]

Vertcoin clone or something new.

[bigradeon]

What is the power consumption of 280x (or 7970) with Lyra2RE? (w/ overclocked and w/o)

It is higher than X11, Neoscrypt, but less than scrypt.

no way, at the same clock frequency lyra cooler than X11,13,15,Neo, approximately 10 degree C. When we ran Lyra at undervolt it took less power

[paulo_thork]

I have a problem I can not start the sgminer in my 280x

when I clickhttp://i.gyazo.com/ab3f5e2c3e933fb8570fedd2c62d08b6.png

after 1 minute is so!http://i.gyazo.com/b822d46a1567add135058a0366d7acb9.png

MY CONF...

{
"pools" : [
   {
      "url" : "stratum+tcp://pool.verters.com:4444",
      "user" : "VhbsVm1????CuoqofZbp23qr",
      "pass" : "xxx"
   }
]
,
"intensity" : "18",
"worksize" : "256",
"kernel" : "Lyra2RE",
"algorithm" : "Lyra2RE",
"lookup-gap" : "2",
"thread-concurrency" : "8193",
"shaders" : "2048",
"gpu-threads" : "2",
"gpu-memdiff" : "0",
"gpu-powertune" : "20",
"gpu-vddc" : "0.000",
"api-mcast-port" : "4028",
"api-port" : "4028",
"expiry" : "28",
"failover-switch-delay" : "60",
"gpu-dyninterval" : "7",
"gpu-platform" : "1",
"log" : "20",
"no-pool-disable" : true,
"queue" : "1",
"scan-time" : "7",
"tcp-keepalive" : "30",
"temp-hysteresis" : "3",
"shares" : "0",
"kernel-path" : "/usr/local/bin",
"verbose" : "true"
}

[djnocide]

Vertcoin clone or something new.

Still vertcoin but an algo switch(From Scrypt-N to Lyra2RE) and a new thread because the Vertcoin account on BTCtalk is banned.

[gregbe]

I have a problem I can not start the sgminer in my 280x

when I clickhttp://i.gyazo.com/ab3f5e2c3e933fb8570fedd2c62d08b6.png

after 1 minute is so!http://i.gyazo.com/b822d46a1567add135058a0366d7acb9.png

MY CONF...

{
"pools" : [
   {
      "url" : "stratum+tcp://pool.verters.com:4444",
      "user" : "VhbsVm1????CuoqofZbp23qr",
      "pass" : "xxx"
   }
]
,
"intensity" : "18",
"worksize" : "256",
"kernel" : "Lyra2RE",
"algorithm" : "Lyra2RE",
"lookup-gap" : "2",
"thread-concurrency" : "8193",
"shaders" : "2048",
"gpu-threads" : "2",
"gpu-memdiff" : "0",
"gpu-powertune" : "20",
"gpu-vddc" : "0.000",
"api-mcast-port" : "4028",
"api-port" : "4028",
"expiry" : "28",
"failover-switch-delay" : "60",
"gpu-dyninterval" : "7",
"gpu-platform" : "1",
"log" : "20",
"no-pool-disable" : true,
"queue" : "1",
"scan-time" : "7",
"tcp-keepalive" : "30",
"temp-hysteresis" : "3",
"shares" : "0",
"kernel-path" : "/usr/local/bin",
"verbose" : "true"
}

try: --no-extranonce

[paulo_thork]

--no-extranonce
can explain better and I do not understand a lot of it!

[daboehla]

--no-extranonce
can explain better and I do not understand a lot of it!

He mean, to start sgminer with the argument --no-extranonce
But please don't ask me, what this do ^^

for example my .bat file to start sgminer:

setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
sgminer.exe --auto-fan --gpu-fan 50-90
pause

[Eastwind]

What is the power consumption of 280x (or 7970) with Lyra2RE? (w/ overclocked and w/o)

It is higher than X11, Neoscrypt, but less than scrypt.

no way, at the same clock frequency lyra cooler than X11,13,15,Neo, approximately 10 degree C. When we ran Lyra at undervolt it took less power

I measured the power consumption. With same core, different memory frequency (750/1500MHz (Lyra) vs 750/1000MHz(Neo)), but higher voltage (880mV vs 840mV) Lyra is about 10% more than Neo. 950W compared to 870W with Neo. That is with 4X7990. I cannot run Lyra with the same voltage as it is not stable.

[bigradeon]

I measured the power consumption. With same core, different memory frequency (750/1500MHz (Lyra) vs 750/1000MHz(Neo)), but higher voltage (880mV vs 840mV) Lyra is about 10% more than Neo. 950W compared to 870W with Neo. That is with 4X7990. I cannot run Lyra with the same voltage as it is not stable.

my friend could run 7990 at 870/1500 ~ 1mb/s at 1.00 vcore, temp 47~52
but in Neo he should run at 1040/1500 ~ 318kh/s at 1080 vcore or higher,  temp 57~65 ( at the same ambient temp with lyra)

in this case your 7990 require a lot more vcore to run stable  compared with neoscrypt, but it doesnt mean that lyra eat more power

[BitmoreCoin]

I measured the power consumption. With same core, different memory frequency (750/1500MHz (Lyra) vs 750/1000MHz(Neo)), but higher voltage (880mV vs 840mV) Lyra is about 10% more than Neo. 950W compared to 870W with Neo. That is with 4X7990. I cannot run Lyra with the same voltage as it is not stable.

my friend could run 7990 at 870/1500 ~ 1mb/s at 1.00 vcore, temp 47~52
but in Neo he should run at 1040/1500 ~ 318kh/s at 1080 vcore or higher,  temp 57~65 ( at the same ambient temp with lyra)

in this case your 7990 require a lot more vcore to run stable  compared with neoscrypt, but it doesnt mean that lyra eat more power

That does not prove that Lyra uses less power. Your friend uses lower voltage and frequency in Lyra.

[BitmoreCoin]

What is the power consumption of 280x (or 7970) with Lyra2RE? (w/ overclocked and w/o)

It is higher than X11, Neoscrypt, but less than scrypt.

no way, at the same clock frequency lyra cooler than X11,13,15,Neo, approximately 10 degree C. When we ran Lyra at undervolt it took less power

I measured the power consumption. With same core, different memory frequency (750/1500MHz (Lyra) vs 750/1000MHz(Neo)), but higher voltage (880mV vs 840mV) Lyra is about 10% more than Neo. 950W compared to 870W with Neo. That is with 4X7990. I cannot run Lyra with the same voltage as it is not stable.

>use different voltages for each algo
>compare results and draw a conclusion about power usage

That just means lyra requires more power to run at the same frequency.

[perezoso]

Ditched X11 with my working GPU rig and now have it mining VTC.

I've got a half dozen more 750ti's that have been gathering dust for months... I might rebuild that rig in the coming days, especially if the news is as good as has been reported when I whip out the kill-o-watt meter tomorrow.

Vertcoin always had the best community and ambition.  The last few months have been trying, but with the hard fork to Lyra2RE, we can start getting excited again.

I'll post power numbers here tomorrow.  AM3+ ASRock boards, Sempron procs, one rig with EVGA cards (running), and another with PNY cards.

The EVGA 750ti's are hashing at 720-725 kh/s each, except one that's stuck at 710-715.

EVGA 750Ti's are factory OC'ed, or more? what clocks?
(I have 690kH/s with Palit StormX Dual, "factory OC" only)

EVGA 750Ti' 620 kHs/s with Gigabyte running Linux ... what dd you do above to achieve higher hash rates ? thanks!

I know ccminer has intensity setting "-i" auto , tried it no effect.. also using latest Cuda 6.5 driver

anyone ? Unix / EVGA 750Ti experience ?

thanks much in advance

I used the unofficial binary from a different fork available here:

http://cryptomining-blog.com/4049-updated-windows-binary-of-the-ccminer-1-5-1-git-fork-by-sp-for-maxwell/

Running just fine for me, but now I'm wondering if it's an illusion.  Still getting 720-725 according to ccminer (except the one stubborn card that's lower), but simplemulti reports me at just over 3 m/hash which, divided by five, is more like 620 k/hash.  Hmmm...

These are the EVGA OC SC cards, on default core speed but increased memory.  It's been a while since I tweaked those settings... it's been in zombie mode mining X11 for months, waiting for something better to come along (like the forked Vertcoin).

PS:  Further overclocking the cores to +40 and the memory to +475 jumps them - according to ccminer - into the realm of 725-735.  I haven't pulled out the kill-o-watt, but I highly doubt the power involved in the overclocking is worth the modest additional hashrate.

[bigradeon]

That does not prove that Lyra uses less power. Your friend uses lower voltage and frequency in Lyra.

we used less vcore to get max speed with Lyra, for example 0.95~1 volt. But in order to get max speed in Neo, we should run at least 1060 ~1100 . ( i tesed many different types of vga) - People who could not run Lyra at max speed and less power need to tweak and waste a lot of time to figure out what optimize config to run.

[ocminer]

Can someone with ccminer have a look at Suprnova's Pool and see if the shares are accepted ? I cannot get mine to run.

Thanks

[Eastwind]

we used less vcore to get max speed with Lyra, for example 0.95~1 volt. But in order to get max speed in Neo, we should run at least 1060 ~1100 . ( i tesed many different types of vga) - People who could not run Lyra at max speed and less power need to tweak and waste a lot of time to figure out what optimize config to run.

I use more power in Lyra because I use higher Vcore and higher memory frequency.
You use less power in Lyra because you use lower Vcore and lower core frequency.
So maybe Lyra uses similar power as Neo.

[bigradeon]

I use more power in Lyra because I use higher Vcore and higher memory frequency.
You use less power in Lyra because you use lower Vcore and lower core frequency.
So maybe Lyra uses similar power as Neo.

  so plz try different config to get max speed and less power, may be your problem in catalyst and OS driver version, or mem/core ratio and other setting like intensity, worksize. We tried trix,giga,ref 79xx,asus,pcs,msi and among them run stable at undervolt and high speed
dont worry maybe in the next few days you'll be able to get better results.
sorry about my poor english!

[Eastwind]

I use more power in Lyra because I use higher Vcore and higher memory frequency.
You use less power in Lyra because you use lower Vcore and lower core frequency.
So maybe Lyra uses similar power as Neo.

  so plz try different config to get max speed and less power, may be your problem in catalyst and OS driver version, or mem/core ratio and other setting like intensity, worksize. We tried trix,giga,ref 79xx,asus,pcs,msi and among them run stable at undervolt and high speed
dont worry maybe in the next few days you'll be able to get better results.

My Lyra hash rate is 7900kH/s with 950W power consumption, or 8.3kH/J. My Neoscrypt hash rate is 2000kH/s with 870W power or 2.3 kH/J. Is your efficiency better than that?

I also run XMR cpu mining with that rig, using 3 threads of 3770k. The power consumption includes cpu mining, which is about 30W.

[q327K091]

I use more power in Lyra because I use higher Vcore and higher memory frequency.
You use less power in Lyra because you use lower Vcore and lower core frequency.
So maybe Lyra uses similar power as Neo.

  so plz try different config to get max speed and less power, may be your problem in catalyst and OS driver version, or mem/core ratio and other setting like intensity, worksize. We tried trix,giga,ref 79xx,asus,pcs,msi and among them run stable at undervolt and high speed
dont worry maybe in the next few days you'll be able to get better results.

My Lyra hash rate is 7900kH/s with 950W power consumption, or 8.3kH/J. My Neoscrypt hash rate is 2000kH/s with 870W power or 2.3 kH/J. Is your efficiency better than that?

I also run XMR cpu mining with that rig, using 3 threads of 3770k. The power consumption includes cpu mining, which is about 30W.

it is absolutely great you are using CPU to mine a coin, very old school, I like it, imagine if half the population of the United States did that, right now is just us a small fraction, someone asked what will it take to propel crypto higher for me is nodes many nodes scattered around and mining  

I am exhausted and tired of short term trades and thinking on exchanges the various pumps and dumps for few hundred dollars profit , few hundred dollars profit although it is money and respected (as behind money FIAT there is work, real human sweat and struggle) is not much in the grand scheme of things

[q327K091]

Ditched X11 with my working GPU rig and now have it mining VTC.

I've got a half dozen more 750ti's that have been gathering dust for months... I might rebuild that rig in the coming days, especially if the news is as good as has been reported when I whip out the kill-o-watt meter tomorrow.

Vertcoin always had the best community and ambition.  The last few months have been trying, but with the hard fork to Lyra2RE, we can start getting excited again.

I'll post power numbers here tomorrow.  AM3+ ASRock boards, Sempron procs, one rig with EVGA cards (running), and another with PNY cards.

The EVGA 750ti's are hashing at 720-725 kh/s each, except one that's stuck at 710-715.

EVGA 750Ti's are factory OC'ed, or more? what clocks?
(I have 690kH/s with Palit StormX Dual, "factory OC" only)

EVGA 750Ti' 620 kHs/s with Gigabyte running Linux ... what dd you do above to achieve higher hash rates ? thanks!

I know ccminer has intensity setting "-i" auto , tried it no effect.. also using latest Cuda 6.5 driver

anyone ? Unix / EVGA 750Ti experience ?

thanks much in advance

I used the unofficial binary from a different fork available here:

http://cryptomining-blog.com/4049-updated-windows-binary-of-the-ccminer-1-5-1-git-fork-by-sp-for-maxwell/

Running just fine for me, but now I'm wondering if it's an illusion.  Still getting 720-725 according to ccminer (except the one stubborn card that's lower), but simplemulti reports me at just over 3 m/hash which, divided by five, is more like 620 k/hash.  Hmmm...

These are the EVGA OC cards, on default core speed but increased memory.  It's been a while since I tweaked those settings... it's been in zombie mode mining X11 for months, waiting for something better to come along (like the forked Vertcoin).

PS:  Further overclocking the cores to +40 and the memory to +475 jumps them - according to ccminer - into the realm of 725-735.  I haven't pulled out the kill-o-watt, but I highly doubt the power involved in the overclocking is worth the modest additional hashrate.

appreciate it, I really thing most we can pull is 620 MH/s on the 750 also please note this:

cuda_6.5.19_linux_64.run  drops about 10 MH/s on a card (just tested it)

use

cuda_6.5.14_linux_64.run

at least on Unix