The problem is that since they seem to be pricing this model in USD, the cost of the units in BTC will fluctuate. For potential buyers that use BTC for their ROI guesswork (like me), fluctuations in the fiat value of BTC can affect the value proposition of the product. Plus a low fiat value of BTC means potentiality a longer ROI, or no ROI, in pure BTC terms.

Compare this to the S2 S3, which was priced in BTC. Sure people bitched about the price when it didn't go down between batches (because BTC's fiat value was going down), but at least you could look at a batch knowing the actual price, do your calculations, and decide whether to buy or not without having to worry that the ROI equation in BTC terms would change by the time you pulled the trigger.

My gut feeling is that Bitmain's careful calculations to sell a product that can just barely ROI (with them keeping the rest of the profits) have been thrown off by BTC's price tanking at an inconvenient time. They want to make a profit, clearly. But if this unit costs too much in BTC, it might not be as competitive as other products, and experienced buyers may pass on the purchase (he says, expecting the usual onslaught of "people on this forum will buy anything regardless of logic").

LOL, but it doesn't bode well that they would be late just providing us with the sales info...

Maybe end of the week for them means Sunday, not Friday.

Of course they seem to update their site during working hours in China, so maybe we have to break out a world clock app or site and set an alarm for 8 AM Sunday Beijing time

+1

The other day I asked for the unit's dimensions, which seems like a harmless request, and haven't gotten a response. If I'm going to get any of these, I have to plan where I'll put them.

Also, some of us (like me) don't have a lot of spare BTC laying around. If the S4 makes sense, I'll probably have to convert from fiat. That takes time, which means that if I don't have pricing info in advance of them taking orders, I won't be part of batch 1.

They should put up the info, including price, and hold off on orders until they're ready.

Bitmain, can you tell us the dimensions of this new unit? I'm interested in rack mounting it if possible.

Will it have "ears" for rack mounting, or at least an option? I'm not a big fan of installing shelves to hold boxes like this.

I'm of course interested in the price as well, but I assume you're not going to pre-announce that. However you must already know the dimensions of the box if you're shipping before the end of the month, so hopefully you'll answer my question.

Bitmain, can you tell us the dimensions of this new unit? I'm interested in rack mounting it if possible.

Will it have "ears" for rack mounting, or at least an option? I'm not a big fan of installing shelves to hold boxes like this.

I'm of course interested in the price as well, but I assume you're not going to pre-announce that. However you must already know the dimensions of the box if you're shipping before the end of the month, so hopefully you'll answer my question.

EDIT - Sorry! I'll ask this on the S4 thread instead

Thanks for the tip on the SP20. I hadn't seen that model for some reason. I'll keep my eye on it, but I agree it's too far out time wise. Mid to end of October is a long time away in this market.

I've also come to the conclusion that keeping things in BTC is the only reasonable way to calculate ROI. Anything else is more like currency speculation and wishful thinking than ROI calculation.

BTW, Bitmain announced B9 of the S3+ shipping on September 18th for 0.58 BTC. To put this price in perspective, this is only 0.089175 BTC less than my batch 3 units (after subtracting the refund that Bitmain offered).

Thanks for the feedback. I'm pretty much in agreement and am holding my S3s.

Sure if I sell today at 0.6 BTC a unit (assuming that I can still get that with B8 now shipping) minus about $30 for shipping, I'd be about 48% over ROI. Whereas if I hold them, I'm looking at just hitting ROI in about 24 days (with pessimistic guesses about the difficulty increases). It will take a while to match the profit of selling today. And of course the further out in the future you push it, the harder it gets to predict difficulty and the corresponding profitability.

But the issue for me is that if I sell the S3s, I'd be bored. Unlike you, I held on to my crappy early gen GridSeeds (5 chip and DualMiner units) and didn't update. As you might expect, those aren't making much BTC nowadays. If I didn't have free power, I doubt I'd keep them running.

Since there really is no mining hardware alternative that has a chance of an ROI, I think I'll stick with my S3's just to keep some skin in the game. Maybe if something new comes out, or if the L1 really gets delivered, I'll look into refreshing my hardware. Hopefully history will hold and the S3s will still have some value then.

I've kept log of S3 performance in a spreadsheet. I track the number of seconds running, local average hash rate, hardware errors, DiffA, and DiffR. From these numbers, I compute the error rate percentage, and the DiffA per second, which I think is an interesting pool side performance statistic that can be compared against different configs running for different amounts of time. After each configuration tweak or update, I update this log and compare the results.

Whenever I make a change, I like to accumulate at least a days' worth of data before determining the affect of the change. You really have to give any changes some time to settle in, unless you're seeing dramatically bad effects (like ASICs shutting down). Also gathering more data over time should help average out pool side variability.

I'm currently running the antMiner_S320140721.bin firmware on my B3 units (I saw no reason to update). The only changes from the stock configuration were to set the "--queue 1" option on the command line, and a frequency of 237.5 (which is as far as I can push it). I've been running like this since around August 11th (which is when I settled into this config).

After running for 6 days with the new cgminer (the one with the security fixes) and no other changes, I can state that with my two B3 units there is no statistically relevant performance difference.

I wasn't expecting any difference with this cgminer update. I did the update because of the security fixes.

I'm posting these results because there are some people who think that this version will either make their hardware hash faster or slower. That is not my experience. All it seems to do is free up some processing power that allows the web GUI to run a bit faster (but that observation is a bit subjective).

I recommend that people who are concerned about the security fixes, or that want a slightly snappier web GUI, should update to this version. Don't expect it to magically affect your hash rate, in either direction.

Seeing as how this thread needs a bit of discussion, here goes...

Suchmoon, you mentioned before that you had some success offloading script miners before they became worthless, then rolling the proceeds into more powerful mining hardware.

So have you sold your S3's yet? I'm around 68% of ROI (batch 3) after 45 days mined. Note that I have free power at a data center, which helps.

Anyway, I've been tempted for the last week or so to sell the S3s and cover the rest of the ROI (and then some).

The thing is that selling on EBay is a pain. In the end, laziness might win and I just coast into a slow ROI by doing nothing ;-)

You say that SSL is overkill for the problem, and it probably is from a general perspective. But overreaching as it may be it also sounds like SSL would in fact solve this problem, making it impossible for an attacker who does not have access to the SSL keys to send the redirect message in the first place.

Sorry for the off topic post. Maybe there's a better place to discuss this.

Me too. No rejects.

I doubt we'll see the Vertcoin fork earlier than November. Since when has any developer beaten their advertised schedule?

When PoolWaffle gets back, I still want to hear his opinion regarding the Script-N pool's future in a post-Vertcoin world.

+1

For me, the S3 was right on the edge. I was very close to not placing an order, but in the end took the risk, largely based on Bitmain's reputation.

I have no intention of ordering an L1 under the current conditions.

Bitmain is now pushing all of the risk onto its customers. The result is that Bitmain's reputation has gone down, and that will factor in my decision making process for future purchases.

In late July the VertCoin devs announced that Vertcoin will be changing its Proof of Work algorithm to Lyra2. The quote from the dev post is pretty clear:


When the fork happens (estimated at the end of November), I assume that no Script-N pool will be able to mine it. It looks like WafflePool's Scrypt-N pool mostly mines VertCoin (about 90% when I checked today). There's only 3 other coins being mined currently.

When VertCoin becomes unminable, will the WafflePool Script-N pool still be viable?

It's very disappointing that Bitmain is doing this as a preorder. In the past, they've announced existing products and shipped from stock (mostly; the S3 was more like build to order). While the specs look good today, I think they will be inadequate after 3 months of difficulty increase, especially if the monster miners (like KnC) ship first and chew up the difficulty.

That said, it is interesting that the L1 appears to support Scrypt-N, which is often quite a bit more profitable than Scrypt if you look historically at the various multipools.

However if you look at Wafflepool, for example, it's mining mostly VertCoin. I think I read somewhere that VertCoin is moving from Scrypt-N to a different algorithm (Lyra2?). If VertCoin is the staple of Scrypt-N and it moves on, then I doubt that there will be any value left in Scrypt-N mining.

So at the time of this writing, it's 12:39 PM Beijing time on September 3rd. Where's the big announcement?

I just noticed that cgminer on one of my batch 3 S3 units restarted. It had been running for nearly 8 days. It appears that only cgminer restarted, not the entire unit (which has an uptime of 20 days currently).

This has never happened before. I've typically run these units for weeks on end, especially after I worked out the config that I want to use. I haven't changed pools recently either.

I've poked around in the logs but can't find any reason for the cgminer restart. I guess I could ignore this since it's basically harmless at the moment. It came up and is hashing away.

But I would like to understand why it restarted. Does anyone have suggestions for debugging this? Maybe there are logs that are available from ssh and not the web GUI that might help?

Given Bitmain's reputation, here's hoping that they ship a scrypt miner made to order (not pre order) with a chance of an ROI. My guess is that it probably won't have the power of the upcoming mega monster miners (like the KnC's that haven't shipped), but will likely have a decent MH/s to BTC cost ratio.

I've been pretty happy with my S3's which are 50% of ROI from a BTC perspective after 30 days. If they can turn the same basic deal with scrypt mining, I'm very interested.

Enough with the Nicehash spam already!

If you have nothing to contribute but advertisements, I'm going to banish you with the Ignore link.

Ack! Bad luck for too long! Not enough air! Passing out...

Someone give my S3's a good home when I'm gone.

Yeah, I have no idea where that came from. I didn't think you led a double life...