harder to dump when the difficulty gets too high, we already know the minting rate isn't gonna go up so watch the value rise in the future. Well yes, the coin rate is constant, however a GPU binary is now available. Some of us that at least put effort on compiling the provided source code, perhaps are also into mining for long enough to know better than dumping at any price. Maybe ...  |
|
|
|
If these things are so efficient and powerful why won't they just plug the machines and mine crypto themselves? Who said they don't, for "testing"  ? Then we're at the stage where who profits are the shovel sellers. |
|
|
|
|
I was really hoping that there was an unnofficial 750Ti with 960 cuda cores...
Come on NVidia, release better Maxwell cards for mining already!
|
|
|
|
|
Too much hassle. Cryptsy doesn't yet support auto-sell of NXT for BTC.
|
|
|
|
|
I have a things to say:
- I'm GPU mining ~17 Mh/s.
- I'm selling, but not dumping at any price and will not go below the current ~0.000026. I rather hold and/or move to GHash.IO and get 2xLTC. That's because I don't want to destroy coins for the quick satoshis, neither I like zero-sum gains (i.e. dumping on bag holders and laughing). Furthermore, at such low HVC prices, the Ghash.io 2xLTC offer becomes competitive in my situation
- "CPU-only" is gone, but the coin PoW is indeed GPU resistant. Yes it was a bold claim which ended up being false, however it's hard to see such a mix of brilliance, programming skill, contorcionism and brainfuck as on heavy140312.cl . I'm sure the OpenCL and Cuda programmers sweated a bit to pull this off.
|
|
|
|
There has been a lot of talk about this coin. Should i mine it? What hash rate should i expect? I get about 3.7mhs on scrypt.
After today's mine and dump and the 800Mh/s whale (damn you!) at hvc.1gh.com, the window of opportunity is closing or just closed. Ghash.io 2xLTC may be better because HVC value is going downhill on exchanges, but see for yourself. |
|
|
|
|
Try --gpu-engine 1050 --temp-target 68 --temp-overheat 75 --temp-cutoff 80 --gpu-fan 70-100
Could be thermal throttling, overclock too high and fan speed too low.
|
|
|
|
1.25x LTC is a still a great offer, and the pool is in Beta. OP, seriously ... |
|
|
|
|
This may be interesting ...
|
|
|
|
I've used 1GH for ~2 months now, and had to post to say I just love this pool. Thank you so much for providing it! This is what I have to do to mine on a generic MPOS: Click "Signup"
Fill username, password, repeat password, email, repeat email, pin, click the "Accept terms of service", click "Signup"
Go to email, click the message, copy your activation token, paste on the browser and activate your account
Click "login"
Fill the email and password. Click "login"
Click on "Workers" and fill the entries as many times needed
Click "Getting started" and copy the address to your config file
Click "Edit account" and press unlock.
Go to email, click the message, copy your unlock token, paste on the browser and unlock
Fill your payout address, auto-withdraw threshold, etc...
Go to Firefox options and allow a cookie from the address. Bookmark the pool
To mine on 1GH: Copy the pool address to the config file
Search my payout address
Bookmark the resulting page
|
|
|
|
|
OP, thanks for your comparison. Great work!
|
|
|
|
I guess I'm one of the few then who is really satisfied with GHash's new pool.
It's clean, functional and got a reply from support within half an hour.
Sure beats the various MPOS pools out there.
Oh, the pool itself is great, so it is the layout but frankly MPOS is the worst on simplicity and convenience. MPOS is okayish _after_ you fill the register form, login, set up workers, unlock the account, go to your email, paste the unlock token, fill the payout address and whatever other garbage one needs to do to just freaking mine. The MPOS software per se is great and I respect it, what is not is having 90% of the pool operators picking it and indirectly forcing so much worthless bullshit down one's throat. Then, if other pools are actually proxying into it to extract the extra LTC and bail-out when the promotion ends, then ghash has actually re-earned part of my respect despite changing the terms. I'm waiting for hours to get the confirmation email to withdraw ltc, but that's another matter |
|
|
|
So i managed to withdraw and all so look like its 100% legit.
But the pool just sitting under the 25gh = 2x jumped instantly to 48gh. I wonder where that 23gh came from, but look like this mean its down to only 125% profitability for the rest of the promotion =/
I've read that wafflepool is proxying there, and so screwing the existing ghash miners. This means wafflepool is out of my config files |
|
|
|
|
Card may be throttling due to heat.
Test with cgminer --scrypt -I 18 -g 1 -w 256 --thread-concurrency 21712 --gpu-engine 950 --gpu-memclock 1250 --gpu-vddc 1.05 --temp-target 68 --gpu-fan 70-100
and see if the hash rate is stable at around 600 Kh/s or whatever.
Voltage is very very high on your cards.
|
|
|
|
That's why it is MANDATORY to undervolt by modding the bios. There are plenty threads about this, even a famous one by drakoin here on the forum. Bios modding isn't mandatory. But it is recommended in case you get Elpida or HynixMFR that you get one of Stilts modded bios, otherwise, Afterburner for voltage and clocks is just as good. What is "mandatory" is just undervolting the cards, not the actual method to do so. I should have been clear though. |
|
|
|
- PCI-e ##x does not really matter what the number is as long as it is compatible with the graphics card. Most GPUs are at 16x, but you can always use a PCI-e extender and use PCI-e 1x with a GPU that requires 16x.
- PCI-e transfer rate (the numbers before the x) does not really affect mining speeds.
- Motherboards required to support multiple GPU's need CROSSFIREX(for ATI cards) or SLI(for nVida cards) compatibility, which allows you to use multiple GPUs. I saw that someone post that they mined WITHOUT it, but I rather not question or try that.
- Do not skimp on the power supply. The PSU should be able to support your card(s), roughly, anything worth buying 7950, r9s should use about 200-400watts. PSU efficiency can lower the wattage needed.
- CPU does not really matter, because all of the work will be done with the PSU.
- For super pro-ness, you need to adjust custom drivers and find the most optimal settings for cgiminer
PCI-e version and speed usually does not matter. However be aware that some slots may have indirect links to the CPU, i.e. with a northbridge in between doing translation and managing peripherals through a few PCI-e lanes. This is more common on Intel platforms. In these situations, you may lose hashrate. Don't skimp on the PSU, indeed. Consider 200W per 280x or 7950, minimum and after undervolting the cards. Don't use a PSU at above 80% capacity 24/7 specially at high ambient temperature. Single rail PSUs are easy to work with, multi-rail have some extra precautions but can be a pain in the ass to balance the load. If you can get an EVGA Supernova G2 1300W for a decent price, go for it. Some guys use 2 PSUs per rig, but dual PSU with powered risers is only recomended for experienced miners. CPU does not matter for scrypt, that's for sure. I suggest though that you pick a dual-core Sempron 180/190 for $5 more than a Sempron 145 or a PentiumG for Intel platforms. You'll use CGWatcher, Teamviewer, MSI Afterburner, etc... mine coins that need more CPU side validation. That said, I don't really want to argue the point. Adjusting settings isn't really pro-ness, it's a necessity after you throw 2000€ into hardware and pay 0.1 ~ 0.25€ / KW/h. - I think you can use multiple types of GPU's to mine? I do not think it matters, they do have to be the same if you are using the GPUs to improve graphics, since SLI/Crossfire doesn't work with different cards. But since we are using the cards to mine, it does not matter. Right? lol
- How the F-, can you calculate how much wattage/Hour or day that are using? I see all these calculators that determine profit, but I do not understand how we can insert wattage into the fields. Like everyone advertises or says that it uses #####w. But does that mean per day? Overall? Hour? Or is it a universal thing I just dont know about?
- Even with optimal cooling, a lot of people are running cards at 100% efficiency, 24/7 foreverrrrrrrrrr, wouldnt the cards be too strained and just die out? Or is there a rest time?
- Should your GPU's fan be at 100% or is lower better? I assume having other things to cool the GPU's and keeping the fan UNDER 100% would make it last longer and lessen the strain. Or is it better for good outside cooling + 100% fan?
- Can you use riser cables with pcie extenders?
- Meh, thats all for now, I thought I had more... But thats it.
You can use multiple types of GPUs to mine, yes. I suggest you stick to R9-270, R9-280 and R9-280x. The R9-290's are harder to mod and undervolt. Sometimes it can be convenient to have a rig with N cards and a half( ), e.g. 4 280x and a single R9-270, due to the PSU you're using. For example, you have enough power for more than 4 280x but you're not comfortable with 5. To calculate wattage, you get a device like a KillAWatt. Otherwise to get a general idea, use HwInfo software, watch the GPU and VRAM power usage and multiply by ~1.3 Cards WILL get strained and die. The temperature reported by the miner software isn't the only one, VRMs can reach 110ºC and the ram temperature isn't really known. That's why it is MANDATORY to undervolt by modding the bios. There are plenty threads about this, even a famous one by drakoin here on the forum. The fans shouldn't need to go above 70%. If they do, you need to undervolt more and get better cooling. Make sure you get those PCI-e powered risers with USB cabling. Connect the videocards, move them away from the motherboard and provide ample ventilation. I suggest the videocard driving the display should be connected on the PCI-e 16x slot, for convenience and stability, but it isn't mandatory to do so. |
|
|
|
|
Yeah, same issue. On one card, I actually use a fan plugged into the motherboard and attached to the card heatsink
|
|
|
|
|
I get that it's not easy to come up with a substantial amount upfront. If one can't mine it back, then it's kinda pointless. Otoh, demanding something and setting low ball "donations", i.e. "I want it, and at the price I determine" is at best spamming and wasting everyone's time.
I'm mostly saying this to prevent that the most talented people of the community end up becoming fed up and leave or slow down to a crawl. I've seen it a few times, either someone that optimized scrypt OpenCL, the dude that modded ATI bios to lower the vram latency, etc...
|
|
|
|
|
You guys are delusional ...
If you want to convince him (i.e. to distribute the cuda HVC miner), you must come up with offers that do not insult intelligence and cover 0.25BTC/day which he is already making (and more if he grabs more cards or rents Amazon EC2 Gpu instances), not your shitty 0.01 BTC or (LOL!) $30.
Oh, and it's not a "donation". You must pay for the service and development time at hour rate.
|
|
|
|
|
Show Posts
