• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

The Altcoin Thread

You think people will "come back" to LTC. I am not so sure and worry about it.

I think if dogecoin and the others burn out, they will just make others.

I think litecoin is going to get pulled down to the single digit dollar value of other alt coins.

I'm not too worried about LTC. But with that said, I'm also not mining it. Because right now there's a strategy of mining the newest alts at low difficulties and attempting to trade them into BTC and rates that are 4-5X or even more than mining straight LTC. It's becoming more and more difficult to do this. And when scrypt ASICs become common, it's going to be impossible to do this, so they only things worth mining for most users will be whatever coins are well established by then - LTC being at least one of them.

In the past few days I've mined DGB, NOBL, and PXL. I plan to trade PXL into BTC later today if possible.
 
As long as people that use Doge understand it will NEVER go to a $1, unless BTC skyrockets into the tens if not hundreds of thousands per coin.
 
About DOGE... Saw this on bitcointalk talk. :p



I believe everything that tried to "copy DOGE" is now all but dead.
 
I'm not too worried about LTC. But with that said, I'm also not mining it. Because right now there's a strategy of mining the newest alts at low difficulties and attempting to trade them into BTC and rates that are 4-5X or even more than mining straight LTC. It's becoming more and more difficult to do this. And when scrypt ASICs become common, it's going to be impossible to do this, so they only things worth mining for most users will be whatever coins are well established by then - LTC being at least one of them.

In the past few days I've mined DGB, NOBL, and PXL. I plan to trade PXL into BTC later today if possible.

I'm too lazy for all that hah. I'd rather work a couple days of OT and build another mining rig.
 
About DOGE... Saw this on bitcointalk talk. :p



I believe everything that tried to "copy DOGE" is now all but dead.

this is probably pretty accurate. The doge community wants it to be a universal tipping currencey,not a fiat replacement like BTC/LTC. I think it could work too, everyone is already tipping people through twitter and reddit with doge. If it could make it to facebook(very unlikely) then it could explode.
 
So where do you "experts" get your profitability estimates? :)
Between "Where to mine" and "CoinWarz", their estimates vs. actuals that I am generating can be off by as much or more than 10x. And are never within 2x.
I understand some uncertainty with all the mining HW/SW/pool variables that are in place, but a coin flip probably has a better chance of helping than the data I am seeing.
 
I think most leave it up to multipool/middlepool and the like to pick the coin.
 
For you ASIC doomsday types.

http://www.reddit.com/r/litecoinmining/comments/1vdihq/gridseed_confusion_cleared_up/

Check out the picture of the actual board, lol.

I assume the "lol" is in reference to the glue and tape in the picture, just like the idiots at reddit are laughing about?

Yeah - that's actually a GOOD thing. Take apart any typical consumer electronics and you'll probably find several solder points, particularly on flexible connection points, covered with hot glue to protect them from breaking. The tape is to prevent that part from shorting out other parts in the "sandwich" that the module becomes.

For a beta part - which this initial gridseed order is - that's pretty damn GOOD build quality. But I'm happy to see that I was correct in my post a while back on the pricing.
 
I assume the "lol" is in reference to the glue and tape in the picture, just like the idiots at reddit are laughing about?

Yeah - that's actually a GOOD thing. Take apart any typical consumer electronics and you'll probably find several solder points, particularly on flexible connection points, covered with hot glue to protect them from breaking. The tape is to prevent that part from shorting out other parts in the "sandwich" that the module becomes.

For a beta part - which this initial gridseed order is - that's pretty damn GOOD build quality. But I'm happy to see that I was correct in my post a while back on the pricing.

Ya, I don't use hotglue on my PCB designs. I use proper headers and block connectors, suppose I should go back to school and learn how to use hotglue.
 
Looking at the board and datasheet pin-out, maybe I missed it, but it's obvious to me the chip is using internal ram only.
Scrypt being highly ram dependent as it is, I'm not sure how much more they can cram on that small die.
That being said, I suppose they are depending on dropping more and more asics down to scale up the hash rate per assembly, but there are only 3 chip ID pins, giving you a total of 8 unique addresses, so for this asic spin, 8 devices may be the maximum load on one controller device.
To really ramp up the hash rate they are going to need to drastically increase the die size to include more ram, thus more costly and my guess is the internal ram is the bulk of the power used now.
Their next step up is not impossible by any means but it will be interesting to see if they can scale the hash rate and power used efficiently.
 
Ya, I don't use hotglue on my PCB designs. I use proper headers and block connectors, suppose I should go back to school and learn how to use hotglue.

In commercial production, it's completely acceptable. If you're prototyping, then yeah, that's a different story.

You also have to remember this is China. Saving a tenth of a penny here and there actually means something over there, especially considering these companies are expecting everything they do will always turn into volume orders. I'm not saying it's great, but it's definitely "acceptable". I can see a few design flaws anyway (like why header if you're just going to solder wires to it? just thru-hole the wires directly!).

(This is coming from someone who grew up in a family electronics mfg business that had to compete with off-shoring).
 
As long as people that use Doge understand it will NEVER go to a $1, unless BTC skyrockets into the tens if not hundreds of thousands per coin.

Heck if it goes to a penny I could buy a car AND a house. It won't go there.

But it will survive. People who are too serious just need to get over the "doge" name and realize that beyond the meme is actually a coin that is fun for beginners and that is the attraction.

Now we'll see what happens to it next year when it has the difficulty of litecoin and a fraction of the reward.
 
Banking on Worldcoin doing something big next month, just sold off my expenses for this month.

*sigh*, I hate that I can't use my r9 270x yet, losing 450kh/s there until next week :(
 
Banking on Worldcoin doing something big next month, just sold off my expenses for this month.

*sigh*, I hate that I can't use my r9 270x yet, losing 450kh/s there until next week :(

I feel your pain. I have a 7950 and a Tahiti LE sitting in the closet upstairs along with a motherboard, but daily life and bills have gotten in the way from picking up a PSU to power them for the last few weeks. That's probably a solid 3 weeks of ~1MHash/s I'm sitting on.
 
If you don't have the budget for a psu but you have the ability to solder and such, you can always do the ibm bladecenter psu mod.

It's a $25 power supply with an incredible 12v rail @ 92% efficiency. Needs 220v though.

You can also bridge a couple of PSUs like the lower cost corsairs.
 
In commercial production, it's completely acceptable. If you're prototyping, then yeah, that's a different story.

You also have to remember this is China. Saving a tenth of a penny here and there actually means something over there, especially considering these companies are expecting everything they do will always turn into volume orders. I'm not saying it's great, but it's definitely "acceptable". I can see a few design flaws anyway (like why header if you're just going to solder wires to it? just thru-hole the wires directly!).

(This is coming from someone who grew up in a family electronics mfg business that had to compete with off-shoring).

I have a tendency to over engineer to the point of lunacy. Hard to keep in mind this is the real world sometimes hah.
 
I have a tendency to over engineer to the point of lunacy..

The true description of a designer :)

The old adage of "Better is the sworn enemy of good enough" always applies in our trade.
 
Why would Worldcoin be doing something big next month? Is that just a hunch or is there some evidence of something coming up?

Worldcoin has been too pump and dump for me to put the real faith in but I'll probably stay sitting on some just in case.
 
Worldcoin has been too pump and dump for me to put the real faith in but I'll probably stay sitting on some just in case.

Likewise, I've just thought it to be "just another coin" up until now. It's exhibited no real signs of being anything but an average performer. Even after I stopped mining LTC directly a couple weeks ago (ya, I got my courage back), I went straight to DOGE and have been doing that ever since because it's more profitable than WDC a good 90% of the time.

But ya, it wouldn't hurt to be sitting on a few WDC if something should happen that is noteworthy and drives its value up even temporarily.

Guess I'll have to switch my miners over for a few days and at least bank a few WDC just in case.
 
Where can I find a link which shows which coin uses which algorithm. I am a bit confused about the whole thing. So for eg, Bitcoin uses x and therefore is compatible with coin A,B and C, whereas protoshare uses y and is the same as coin X. Y and Z....something like that.

Is there any USB miner for LTC?
 
So I have this one stupid card that runs 3-5 degrees warmer than the others, really annoys the heck out of me.

The real answer is probably to take it apart and re-do the thermal paste which I bet is not done right. I have a second one of the same card and it runs warm but not as hot as this one. But I am afraid of the downtime and accidentally breaking it.

What angle would you say blowing a fan on it would improve this card, I have one pointing it at the edge from the back and it doesn't really help much at all.

7950PCS-01.jpg


7950PCS-03.jpg


7950PCS-04.jpg
 
Two words, box fans. =P I have a card or two that regularly run 5c hotter than the rest. I stopped caring cause they are still in the 70's.

But, I say that blowing across the card of any angle is fine. I blow towards the bracket, then across the card to the back, which probably make no sense lol.
 
No picture, but I actually have a 120mm fan resting on the top-rear (opposite end from the connector plate) of two cards plugged into the MB. Resting on both cards like a bridge, just inside the 6 pin connectors of both.
This fan, I believe, is pulling air up between the cards and blowing it out the top.
I found this reduced the temps of both cards by 2-3 degrees, but most importantly it allowed the GPU fans to spin down to 50% or much less and still control the card temps to my set limits.
This fan is in addition to a large box fan that blows across the whole assembly.
 
OpoXd2tl.jpg



How I have mine setup. Cards run 70-75c. And ya it's ugly, but once I got it churning out coins I didn't want to touch it. I'll beautify it one day.
 
AH! Seeing that mobo sitting on that metal shelf just makes me all anxious. Obviously you have something under it like wood, cardboard or plastic?
 
How I have mine setup. Cards run 70-75c. And ya it's ugly, but once I got it churning out coins I didn't want to touch it. I'll beautify it one day.

Is that motherboard sitting on a metal rack? How is that even possible? :confused:

No room for box fan unfortunately (in small studio apartment).

I think I am switching the new usb risers so I can cool things down more since I am in warm climate.

So I am about to put up some ribbon risers for sale I think this weekend if anyone is interested, I've got a variety of different kinds. The last ones I got work great but they are too short to move the cards completely apart.
 
AH! Seeing that mobo sitting on that metal shelf just makes me all anxious. Obviously you have something under it like wood, cardboard or plastic?

Of course.

Edit: Cardboard is actually taped to the motherboard.
 
Where can I find a link which shows which coin uses which algorithm. I am a bit confused about the whole thing. So for eg, Bitcoin uses x and therefore is compatible with coin A,B and C, whereas protoshare uses y and is the same as coin X. Y and Z....something like that.

Is there any USB miner for LTC?

usb miner = asic, which does not yet exist outside of testing for LTC and other scrypt coins

the algorithm you are asking about is sha256 vs scrypt

many other alt coins are scrypt now except a couple like primecoin

if you already own usb asics there is this bunch of sha256 but nothing like bitcoin
http://www.coinwarz.com/cryptocurre...Check=true&scryptCheck=false#tblCoins_wrapper
 
I have a tendency to over engineer to the point of lunacy. Hard to keep in mind this is the real world sometimes hah.

It's all good. That's what a real engineer should do. But then it goes to the bean counters up front who solicit the bids and usually it comes down to who can make it cheaper. That's why my dad ultimately ended up closing down shop - besides, it was time for him to retire. He just laughed as his biggest client ended up off-shoring, and then a year later came crying back over high failure rates and poor quality control. You get what you pay for, quite literally. In the end, unfortunately, it all comes down to making money.

One thing I have to mention though, since you said it... those header connectors you mention? Yeah, those are by far one of the biggest areas for cost cutting on projects like that. We're talking easily a nickle or more plus labor per connector just for small headers (like 2x3 pin ones). And those are the cheap ones!

So I have this one stupid card that runs 3-5 degrees warmer than the others, really annoys the heck out of me.

The real answer is probably to take it apart and re-do the thermal paste which I bet is not done right. I have a second one of the same card and it runs warm but not as hot as this one. But I am afraid of the downtime and accidentally breaking it.

What angle would you say blowing a fan on it would improve this card, I have one pointing it at the edge from the back and it doesn't really help much at all.

One of my 6950's is like that. It just runs about 8 degrees warmer than the others no matter what or where it is. I pulled it apart and re-TIM'd it. I originally ran probably 10-12 degrees hotter, so that definitely helped a bit, but not too much after all was said and done.

What I figured out is the temps make a different how much air gets to the cooler. These are reference cards on a mobo (no risers), so they're stacked right next to each other. The cards get suffocated, so I figured putting this one down in the lowest slot so it could draw air without obstruction would help. It didn't change much. Then I figured out the top card ran cooler because it's backplate was exposed with nothing "hot" behind it (no other cards to warm it up). What I ended up doing was shoving a Q-Tip between the front edge of each card to separate them all by about 1/4" so air could get in to the blowers of the first two cards. This dropped the temps on ALL the cards, and most importantly dropped the temps on the "hot" card to right around 70c AND let me set the fans to a more manageable 60%. So far the hottest it's gotten on the warm days has been about 73c.

I eventually plan on migrating these cards to a mining rack build with risers so that they can have some "real" airflow.
 
Thinking about switching over my 360 kh/s (7850) card over to a switching pool vs straight mining of FTC/WDC. Would it be worth it since my hashing power is weak? If it is worth it should I go to hashcow, or multipool?

Thanks
 
Thinking about switching over my 360 kh/s (7850) card over to a switching pool vs straight mining of FTC/WDC. Would it be worth it since my hashing power is weak? If it is worth it should I go to hashcow, or multipool?

Thanks

I wouldn't want to solo mine and I'm running 6 GPUs for a total of about 3.8 to 4.2M/hash....

over time I'm sure the law of averages says I would end up with the same amount of coin either way, I just prefer a sure bet vs the "luck" of finding a block. Especially since some of the guys in theblocksfactory WDC pool are kicking out 100+MHash on their account.
 
OpoXd2tl.jpg



How I have mine setup. Cards run 70-75c. And ya it's ugly, but once I got it churning out coins I didn't want to touch it. I'll beautify it one day.

a whole rack with just 1 rig on it? fill it up!
 
Is that motherboard sitting on a metal rack? How is that even possible? :confused:

No room for box fan unfortunately (in small studio apartment).

I think I am switching the new usb risers so I can cool things down more since I am in warm climate.

So I am about to put up some ribbon risers for sale I think this weekend if anyone is interested, I've got a variety of different kinds. The last ones I got work great but they are too short to move the cards completely apart.

you need to devote all of the space to your rigs! Get your priorities straight bro and just sleep on the bottom shelf of a rack
 
I wouldn't want to solo mine and I'm running 6 GPUs for a total of about 3.8 to 4.2M/hash....

over time I'm sure the law of averages says I would end up with the same amount of coin either way, I just prefer a sure bet vs the "luck" of finding a block. Especially since some of the guys in theblocksfactory WDC pool are kicking out 100+MHash on their account.
Hes asking about moving between pools, not solo mining
 
I guess that's one way for the government to disrupt the bitcoin economy.
 
I wouldn't want to solo mine and I'm running 6 GPUs for a total of about 3.8 to 4.2M/hash....

over time I'm sure the law of averages says I would end up with the same amount of coin either way, I just prefer a sure bet vs the "luck" of finding a block. Especially since some of the guys in theblocksfactory WDC pool are kicking out 100+MHash on their account.


I'm not solo mining. Just wondering about switching pools. With my mining power being so low, is it better to just stick with switching between ftc/wdc pools or would I be better off switching to hashcow or multipool. And if so, which switching pool do people prefer
 
Back
Top