Bowing to pressure, GHash agrees to limit share of hashrate to 39.9%
Image source: https://www.flickr.com/photos/gastev/
As a bitcoin miner, it’s hard not to have mixed feelings about GHash.io. On one hand, the pool’s massive amount of the total network hashrate means consistent and generally high payouts. The pool’s operators have clearly invested considerable resources in creating a powerful, easy-to-use dashboard and user experience. It’s a well-made, zero-fee system that works exactly as advertised.
On the other hand, the pool also blindly drove over the 51% cliff last month, crashing profits and making even non-miners worried about the security of the overall network from a true 51% attack. The 51% scenario could have been avoided easily simply by pointing some of the pool’s own miners elsewhere. GHash seemed to be asleep at the wheel, letting their own pool members take the hit when the price took a nose dive.
It’s hardly surprising that so many miners and smaller pools, like PetaMine, flee GHash in the hopes of keeping the network — and the exchange rate — stable.
GHash appears to have gotten the message. At a meeting of bitcoin mining bigwigs during Coinsummit London, the pool announced it would cap its operation to “not exceed more than 39.99% of the overall Bitcoin hashrate.” Although that’s still a huge amount of hashing power, and there’s no guarantee that GHash won’t effectively control far more simply by splitting up its hashrate among smaller pools, it is a start.
GHash, the Bitcoin Foundation and other related parties also committed to looking for a trustless, code-based solution to the 51% problem.
Industrial-scale bitcoin mining booming
Is bitcoin mining about to become big business? Until recently, the massive warehouses filled with thousands of ASICs and roaring fans struggling to cool them off seemed like the exception to the rule. Companies like MegaBigPower seemed more like the exception than the rule.
For most of bitcoin’s history, mining operations have been relatively small, both in terms of physical size and financial investment. Even a full-time professional miner might only have $20,000 or $30,000 worth of equipment.
In the last six months, however, large-scale bitcoin mining has become very serious business, with tens of millions of dollars being spent on creating massive new “hashing centers” across the country. According to a report on Data Center Knowledge, a trade publication for data center technology, the costs of converting a warehouse to a bitcoin mining facility is a fraction of the expense of building a Google or Facebook-like server farm.
“To build out one of these (warehouse) data centers isn’t as expensive,” Butterfly Labs COO Josh Zerlan told the publication. “You can put in $10 million instead of $100 million for this kind of facility.”
What’s more, although the power consumption can be quite high, with the proper cooling arrangement a hashing center can have an extremely dense and efficient layout.
Bitcoin mining servers can pack as much as 1 megawatt of equipment into 1,000 square feet of space, according to Eric Doricko, a veteran of Exodus Communications who now helps Bitcoin businesses find data center space. That’s a big change from the 8,000 to 12,000 square feet of space for 1 megawatt of traditional IT space.
Will this boom in industrial-level bitcoin mining squeeze out the hobbyist miner? Or will it encourage many of those miners to test out the cloud-based operations many of these data centers are experimenting with?
Miners breathe easy this week: 6% difficulty increase expected
Image source: BitcoinWisdom
For most of 2014, miners have been faced with an unprecedented growth in the all-important difficulty setting. Six months ago, the hashing power of the entire bitcoin network was a mere 20 PH/s, meaning that a miner with a 1 TH/s rig had a fighting chance at making a decent profit in the daily game of roulette that is bitcoin mining. Since early March, however, a flood of newer, faster, cheaper ASICs and giant cloud-based mining companies have added around 110 PH/s to the network, resulting in dramatic and painful jumps in overall difficulty.
Since the difficulty increases every 2016 blocks (generally three times a month, for the non-miners), a mining set-up can go from a somewhat profitable venture to a money-losing operation in a matter of days. This point was driven home for many miners in late June, when the difficulty spiked, growing almost 25% from 13,462,580,115 to 16,818,461,371.
Even relatively well-equipped miners were suddenly facing a situation where their equipment might cost more to power and operate than it was likely to make back from block rewards.
But there is good news: For the first time since early 2013, the next difficulty increase is expected to be in the single digits. BitcoinWisdom is expecting a mere 5.8% increase on Sunday, giving miners a much-deserved break from their constant struggle for profitability.
It might not last long, however. With many mining hardware makers on the verge of releasing next-generation, high-powered ASICs, this might simply be the calm before the storm.
PetaMine shareholders vote to dump GHash for P2Pool
Few mining pools have provoked as much resentment and outright anger as GHash. When the CEX.io-owned pool crossed 51% of the total bitcoin network hashrate last month, it did more than crash the price — it made many in the bitcoin community realize that the dreaded “51% attack” scenario wasn’t as implausible as some experts had claimed.
With this in mind, shareholder-owned mining collective PetaMine decided it might be time to point their 1.2 PH/s of hashing power away from GHash. Last week, PetaMine’s administrators announced the results of a survey seeking shareholder input on moving towards decentralized mining pool P2Pool, selecting another pool to join, or staying with GHash.
The results couldn’t have been more clear. With 68% of unit holders responding, 96% voted in favor of switching to P2Pool. A mere 3% voted in favor of staying with GHash, even though staying with the larger pool would mean their profits would be substantially less erratic on a daily basis.
PetaMine has since moved forward with their P2Pool adoption, although a compatibility problem with BitFury systems has caused some delays.
MegaBigPower launches franchise program
Inside MegaBigPower’s facility.
With over 5.16 PH/s at its command, MegaBigPower is easily one of the biggest direct mining operations in existence. The Washington-based company’s BitFury mining rigs fill a 20,000 square foot warehouse, earning millions of dollars worth of bitcoin and various alt-coins per month. As impressive as the operation is, it’s not quite enough for MegaBigPower. The company announced last week that it was launching a new franchise program that could add as much as 50 PH/s to the overall bitcoin hashing pool.
Speaking with CoinDesk, MegaBigPower owner Dave Carlson said that the move comes in anticipation of an oversupply of hardware on the market.
"I perceive that there is going to be a large surplus of manufactured hash power," Carlson said. "What I think is going to come of that is that the new commodity won’t be access to mining equipment. It’s access to power.”
If Carlson is right, his company will soon have a glut of cheap, powerful mining equipment, but limited space to operate it. Setting up a new facility and paying for power is too expensive for his current profit margin, prompting the idea of a franchise model for new locations. MegaBigPower supplies the mining rigs, software and expertise, franchisees provide the location and the electricity.
With a 50/50 split of mining revenue, the operation could prove to be hugely profitable for MegaBigPower. Success will, of course, depend on finding franchisees with easily cooled facilities and access to cheap electricity.
“What we want to do is build a large network presence,” Carlson told CoinDesk. “If we do it through this franchise network, we’ll do it faster and easier than if we were to build this power ourselves.”
Fine-print clauses KnCMiner’s Neptune-delay compensation agreement angers customers
In April, Swedish mining hardware maker KnCMiner decided to make a good-faith effort to appease customers who had pre-ordered their long-delayed Neptune bitcoin mining machine. Those who had ordered from batch one and two would be entitled to a free Neptune from the third batch. It was a popular move, as early adopters could count on doubling their hashing power if they were willing to be patient.
With the first batch of Neptunes preparing to ship next month, KnCMiner seems to have changed tactics, offering hosted mining services for those still waiting on their mining hardware. As CoinDesk reports, some customers have taken issue with the small print, which seems to renege on the April agreement.
“If the purchaser of the Products (referred to as the Customer) accepts KnCMiner’s offer and connects to KnCMiner’s Mining Services, the Customer automatically accepts that these General Terms shall apply for the provided Mining Services and that the Mining Services provided by KnCMiner shall constitute the Customer’s sole and only compensation due to the delay of the Products and shall supersede and extinguish all prior agreements relating to compensation for the delayed Products.”
Not surprisingly, many KnCMiner customers took to Reddit and other popular bitcoin forums to cry foul. Given that a new Neptune sells for $5,995, a maximum of six weeks of hosted mining at the same hashrate seems like a weak trade to many.
The company was quick to respond, and said it will allow users who have not yet redeemed their hosting offer to switch back to the previous agreement.
KnCMiner co-founder Sam Cole told CoinDesk: “It was never our intention to deceive people. … Really, it’s a lose-lose for us and either a little win or a big win for the customers.”
Mining hardware maker CoinTerra moves for out-of-court settlement in class-action suit
CoinTerra’s ill-fated TerraMiner IV
According to a new report by Ars Technica, bitcoin mining hardware maker CoinTerra’s dark days may be finally coming to a close. The company has been facing a class-action suit brought by disgruntled customers who purchased the TerraMiner IV on pre-order, only to wait through 10 months of development and shipping delays. With a courtroom battle expected to be slow and expensive, both parties now seem willing to consider a mediated solution.
According to the plaintiffs’ attorney, Edward Mullins:
“We hope to have it resolved sometime this summer,” Mullins said. “If it doesn’t work out then we will go forward [with the lawsuit.]”
With several similar cases plaguing other mining hardware makers, the CoinTerra case may offer an insight into how those cases will ultimately be resolved.
CoinTerra’s situation has become something of a cautionary tale in the mining hardware industry. Even with a highly experienced and well-funded team, the company still experienced huge delays. As Ars Technica explains:
The startup was founded by a team that appears to have extensive experience in the industry. The company’s CEO, Ravi Iyengar, was Lead CPU Architect at Samsung for two years and worked out of the Samsung Austin Research Center. The head of the company’s advisory board is Naveed Sherwani, an Intel veteran and current chair of the Global Semiconductor Association Technical Steering Committee. Sherwani literally wrote the book on very-large-scale integration (VLSI) semiconductor design and production.
But even they couldn’t seem to get their act together.
No date has yet been set for the CoinTerra mediation, but the company has formally asked the court for more time to allow for negotiations to take place.
Bitmain unveils Antminer S3
BM1382 chips for the Antminer S3
Seven months after releasing the BM1380 ASIC, the Bitmain has finished testing the latest iteration of their chipset, the BM1382. The updated design is a 28nm chip measuring 8mm square, and packing an impressive 63 cores. Bitmain claims the chips can deliver 15.75 GH/s at 0.59 J/GH.
The new chips will power the latest update to the company’s popular Antminer line. The Antminer S3 contains 32 BM1382 chips in a blade-style design, providing roughly 504 GH/s at 390W from the wall. The new design includes an improved heat sink and a two-fan cooling system, a clear improvement over the old open-air system seen in previous models.
Bitmain noted that the MB1382 chips are fully compatible with older S1 and S2 Antminers, allowing existing customers to upgrade at a reduced cost.
Pre-orders of the S3 are expected to ship in July. No shipping timeline has been released for the second batch, and at the moment Bitmain doesn’t even have public pricing data for the S3. The S3 appears to have similar specs to Rockminer’s RK-Box (460 GH/s), it may be aimed at a comparable price point of around $650.
KnCMiner announces launch of 20nm Neptune chips
After months of build-up, KnCMiner announced today that their new line of 20nm Neptune ASICs have arrived at their facility in Sweden. The chips are currently being tested, and units are expected to begin shipping later this week. The Neptune boasts five 1440-core chips, and KnCMiner claims that it can deliver more than 3 TH/s at 2.1 KW. If accurate in real-world use, this 0.7 watt per GH/s represents a 30% reduction in power cost versus the current KnCMiner heavy hitter, the Jupiter.
The company also claims that the Neptune system will increase in performance as future software updates maximize the new chips’ potential.
The Neptune puts KnCMiner in a small club of next-generation mining hardware makers, although many other developers are expected to have equally powerful hardware on the market in coming months. Early adopting miners can expect to pay $5995 for the third batch of the Neptune hardware, slated to ship early in Q4. The first two batches of Neptune rigs are already sold out via pre-order.
Hosting mining company HashPlex raises $400k in venture capital
Barry Silbert clearly sees potential in Seattle-based cloud-mining company HashPlex. The SecondMarket CEO and noted bitcoin evangelist recently joined a seed round investment of $400,000 in the company. Unlike other cloud-based bitcoin mining systems, however, HashPlex doesn’t sell set amounts of hash power to customers. Instead, the company manages your existing mining equipment for you, reducing the hassle and expense of running a small cryptocurrency mining operation.
It works like this: You send your mining hardware to HashPlex, along with the details of what pool the miner will point at and what wallets it will pay out to. They provide the host computers, passive cooling, internet connections, start/stop/control scripting and any basic maintenance on the hardware. You pay $99 per kilowatt used each month, with at $10 discount if you use more than 30 kw.
For even a semi-serious miner, those numbers are already looking pretty good. It’s easy to see why Silbert thought the idea was worth investing in. Senior Facebook engineer Jason Prado also contributed to the seed round.
Should HashPlex’s “1MW HashCenter” in central Washington prove successful, the company expects to open other sites in the region. Silbert has hinted that HashPlex may also develop its own “vertically integrated mining pool” for its customers.