CCminer(SP-MOD) Modded GPU kernels.

[Velgelm]

SP Private 2 miner awesome !

tpruvot 1.7.4 get 1860 mhs SP#2 get 2013 mhs (960x2 cards)
thanks SP!

[bensam1231]

Yesterday I managed to push decred up to 537MHASH(+3.8%) on the asus strix 750ti.(1260 core) But it use more cpu power, so my celeron based rigs are struggleing with 6 cards.. (same problem as in release #1) The cpu is used for precalculation of the first round. Release #2 has no precalc of the first round.

I will try some stuff in the weekend, and hopefully release version 3.

Consider making two modes. One for high cpu utilization and one for low as has been done in the past. Some people have more powerful CPUs so it's not a issue, but I agree most machines are pretty anemic.

I'm currently having issues with mining Ethereum on one of my older A4-4000 systems I originally bought when I entered cryptos. Seemed like a great idea to save $10 at the time, now not so much. Most of the time Eth doesn't use any CPU cycles, but occasionally it starts parsing stuff or something and causes the system to freeze after sucking out all the cycles and there aren't enough to run basic tasks. If I run it at below normal priority sometimes it errors out.

[Ayers]

SP Private 2 miner awesome !

tpruvot 1.7.4 get 1860 mhs SP#2 get 2013 mhs (960x2 cards)
thanks SP!

for what algo? if this is decred it's a one hell of an achievement, but i can not see this possible

[sp_]

SP Private 2 miner awesome !
tpruvot 1.7.4 get 1860 mhs SP#2 get 2013 mhs (960x2 cards)
thanks SP!

for what algo? if this is decred it's a one hell of an achievement, but i can not see this possible

Decred 2x cards. 1007MHASH per card. And more hash is coming in release 3

[Vaccomondus]

how much decred is doing with 970 with future release? can you reach 1500 1600? i may pay for it

[sp_]

how much decred is doing with 970 with future release? can you reach 1500 1600? i may pay for it

release #2 is doing 1515@1366
release #3 beta(unreleased) is doing 1560@1366 (gigabyte gaming default clock) (single card)

[sp_]

After I am done with decred mod #3. I think I will do Vanilla. Anybody interested? Around 5% boost.

[chrysophylax]

After I am done with decred mod #3. I think I will do Vanilla. Anybody interested? Around 5% boost.

im always here ...

and interested ...

#crysx

[Grim]

LOL 



From cryptominingblog:

The last part of our initial impressions that did not leave us very happy is the fact that there was quite a bit of dust buildup inside the miner, way more than there should be from a normal testing of new hardware before shipping to customer for example. The outside of the miner makes an impressions that it is new (cleaned well), but on the inside – the fans and the heatsinks of the blades with chips have dust that is hard to cleanup with just compressed air without taking the device apart. The amount of dust we should normally be associated with at least of few weeks of usage in normal everyday environment (non server room) and we don’t think that these devices are tested in a very dusty room. So it seems that they have been “tested” for some more time before being shipped to customers…

[pallas]

LOL 

From cryptominingblog:

The last part of our initial impressions that did not leave us very happy is the fact that there was quite a bit of dust buildup inside the miner, way more than there should be from a normal testing of new hardware before shipping to customer for example. The outside of the miner makes an impressions that it is new (cleaned well), but on the inside – the fans and the heatsinks of the blades with chips have dust that is hard to cleanup with just compressed air without taking the device apart. The amount of dust we should normally be associated with at least of few weeks of usage in normal everyday environment (non server room) and we don’t think that these devices are tested in a very dusty room. So it seems that they have been “tested” for some more time before being shipped to customers…

hehe!
if someone was still not convinced that they lie when they say "we do not mine" ;-)

[antantti]

sp_ have you tested your dcr mod on 980?

[sp_]

Release #3 does 2370 on the 980ti
1.7.4 buildt with cuda 6.5 does 1940 on the same  card.
I din't tune it for compute 5.2 yet. compute 5.0 is the priority now.

[sp_]

Release #3 sendt to the donators

release 2: asus strix Gtx 750ti 517MHASH(1260 core)
release 3: asus strix Gtx 750ti 528MHASH(1260 core)

Also bether on gtx 970

release 2: gigabyte g1 gaming 1515MHASH (1366 core)
release 3: gigabyte g1 gaming 1540MHASH (1366 core)

This version Doesn't use much CPU and should work on celeron rigs..

Buildt with cuda 7.5 32 bit this time. (Use less power.)
release #2 was buildt with cuda 7.5 64 bit.

[sp_]

People are asking for a linux version. I might give the sourcecode of version #3 to trusted members for a 0.3 BTC donation.

[AliMan]

Release #3 sendt to the donators

release 2: asus strix Gtx 759ti 517MHASH(1260 core)
release 3: asus strix Gtx 759ti 528MHASH(1260 core)

Also bether on gtx 970

release 2: gigabyte g1 gaming 1515MHASH (1366 core)
release 3: gigabyte g1 gaming 1540MHASH (1366 core)

This version Doesn't use much CPU and should work on celeron rigs..

If you're able to do over 2500 on the 980 ti's, then I'll be interested.

[sp_]

980ti: 2,5GHASH 300W
750ti: 0.5GHASH (30W) (tdp/OC)

2x more bang for the buck with the 750ti, sp-modded kernals & configuration.

[sp_]

Consider making two modes. One for high cpu utilization and one for low as has been done in the past. Some people have more powerful CPUs so it's not a issue, but I agree most machines are pretty anemic.

I might add a "--cpu-mining" in Release #4. Gains a few MHASH but use alot of power.

[bensam1231]

After I am done with decred mod #3. I think I will do Vanilla. Anybody interested? Around 5% boost.

Vanilla is a pretty poor earner now, even with a 5% boost. Time better spent on Neo, Lyra, Quark, Decred...

LOL 



From cryptominingblog:

The last part of our initial impressions that did not leave us very happy is the fact that there was quite a bit of dust buildup inside the miner, way more than there should be from a normal testing of new hardware before shipping to customer for example. The outside of the miner makes an impressions that it is new (cleaned well), but on the inside – the fans and the heatsinks of the blades with chips have dust that is hard to cleanup with just compressed air without taking the device apart. The amount of dust we should normally be associated with at least of few weeks of usage in normal everyday environment (non server room) and we don’t think that these devices are tested in a very dusty room. So it seems that they have been “tested” for some more time before being shipped to customers…

Yup... Saw that. Sure those things have been in use for at least for six months. If it's in a datacenter maybe even longer then that due to the air quality.

[joblo]

LOL 

From cryptominingblog:

The last part of our initial impressions that did not leave us very happy is the fact that there was quite a bit of dust buildup inside the miner, way more than there should be from a normal testing of new hardware before shipping to customer for example. The outside of the miner makes an impressions that it is new (cleaned well), but on the inside – the fans and the heatsinks of the blades with chips have dust that is hard to cleanup with just compressed air without taking the device apart. The amount of dust we should normally be associated with at least of few weeks of usage in normal everyday environment (non server room) and we don’t think that these devices are tested in a very dusty room. So it seems that they have been “tested” for some more time before being shipped to customers…

hehe!
if someone was still not convinced that they lie when they say "we do not mine" ;-)

There are two parts to that comment. Do they have their own farm with their own mining equipment and do they mine with
their customer inventory. The latter is assumed to be false while the former is likely the target of such a claim.

Miners don't seem more likely to be passed off as new when used than other electronics. It was probably a production
sample that had been stress-tested before shipping to be reviewed. Assuming they knew who they were shipping to
they would make sure the unit wouldn't die during the review tests. Either way the reviewer and the producer should
have been explicit about the unit's status.  And not cleaning it was simply sloppy and unprofessional.

[bensam1231]

LOL 

From cryptominingblog:

The last part of our initial impressions that did not leave us very happy is the fact that there was quite a bit of dust buildup inside the miner, way more than there should be from a normal testing of new hardware before shipping to customer for example. The outside of the miner makes an impressions that it is new (cleaned well), but on the inside – the fans and the heatsinks of the blades with chips have dust that is hard to cleanup with just compressed air without taking the device apart. The amount of dust we should normally be associated with at least of few weeks of usage in normal everyday environment (non server room) and we don’t think that these devices are tested in a very dusty room. So it seems that they have been “tested” for some more time before being shipped to customers…

hehe!
if someone was still not convinced that they lie when they say "we do not mine" ;-)

There are two parts to that comment. Do they have their own farm with their own mining equipment and do they mine with
their customer inventory. The latter is assumed to be false while the former is likely the target of such a claim.

Miners don't seem more likely to be passed off as new when used than other electronics. It was probably a production
sample that had been stress-tested before shipping to be reviewed. Assuming they knew who they were shipping to
they would make sure the unit wouldn't die during the review tests. Either way the reviewer and the producer should
have been explicit about the unit's status.  And not cleaning it was simply sloppy and unprofessional.

It takes a lot of time to build up dust, especially in a good environment such as a datacenter more so if it's already been pressure dusted and there is residue that wont come off. This isn't flipping it on and testing it for a week.

I'd definitely say it's more likely for miners to be used before being sold compared to other electronics. Cryptos are a completely different market.