Probably dropped the core voltage by a few percent to do that. Decreasing chip temps would be a compounding effect, and then shaving off a few more watts with less fan.

And that's why we reinforce each one with epoxy.

Wonder how many months they've been running tuned in their own farms.

Yeah nobody here is going to lie to you about the numbers. It's a toy, not a business case. Lot of people will solo mine with them because, for a bitcoin lottery ticket, you can't really beat a nearly silent 2TH miner using under $10 of power per month.

The Pi4 doesn't even blink, so a Pi3 should handle just fine.

Resellers only, same as we've done for the last several years. Mine is a small shop so we focus on R&D and manufacture and let resellers handle advertising, distribution and customer support.

I've only tested 1 at a time on my Pi, but we're running five per USB bus (10 per machine) on the old Dell 755 testers. If a current model Pi has comparable horsepower to half an office PC from like 2010 you'd be fine. Given a Pi will run something like 12-14 Compac F, 3-4 R909 shouldn't be difficult.

By the way, 1.5TH is the baseline spec. The last batch off testers were running 1.6-1.7TH and still had headroom. Kano and I have both tested units at 2TH and up. Of course the top-end of your units will be dependent on the hardware behaving nicely but I bet you could get 3.6TH just from the R909s.

If Microsoft was honest, its antivirus software would delete Windows 10.

Antivirus software has been known to flag cgminer not on its own merits but because malware bots would run it on people's computers to harvest free hashes.

It's a brand-new product and we busted our butts to get the assembly line up to speed in time for the first batches to be out for Christmas, which is still 19 days off. Reseller stock requests are being filled right now, so they'll be "in stock" with resellers in one to two weeks.

The BM1397 will go to 750MHz. The Compac F will go to 750MHz, know of a guy that has a few running on that setting for multiple months. You're more likely around 1.6V core, and closer to 40W/TH efficiency than 30. Stock setting on S17 models is around 35W/TH at the chip, T17 around 45W/TH. 15W into a Compac F will get you about 350GH.

I've yet to see an S17 with 9 chips in parallel. Remember the total chip count is spread across three boards. What I'm most familiar with are the 195-chip version, with 65 chips per board in 13 sets of 5 and a domain around 1.5V

Bitfury's last couple generations of chip used a very different level shifting and data multiplexing setup than the daisy-chain of their original chips and what most other vendors use now (including Bitmain).

Fuzzy, the ASICMiner Tube used TPS53355 regulators for every I think three ASICs. The Prisma, using the same heatsink profile, took in unregulated 12V into a series string. I believe Bitfury was already capable of series but I don't know much about early generation BF gear, since it was expensive (boasting 2W/GH at a time when most gear was 6+W/GH). Bitmain's second gen S7 (45 chip models, not the 54 chip models) used a regulated string. Novak and I were experimenting with the idea of a regulated string about 9 months before Bitmain brought it to market, which was fun.

Sounds like a Core problem. Also sounds like a problem that's already been discussed elsewhere and doesn't really need to be here too.

This miner is built with the same physical dimensions as the R606 so anything built around them would still fit this unit. Airflow ducting is the same, same 80mm fan. However, you'll notice ports and jacks are located at the exhaust end instead of accessible from the side.

This makes the unit more readily stackable, as you no longer have to keep three sides clear - two for airflow and one for cabling.

It also allowed us to expand the circuit board by 15%, making room for a better main regulator and improved thermal management. Heatsink area over the chips is increased by 40% over the R606, and we've added more board mounting legs for improved stability especially during shipping.

The miner runs on the S17's BM1397 ASIC, as used in our Compac-F USB stick miner. Kano has been hard at work improving the driver, and has added several features that will be very attractive to tuning enthusiasts, though the stock settings will already be fairly excellent.

The miner is guaranteed to run at 1.5TH from under 100 DC watts at stock voltage and do it nearly silently. Better performance isn't guaranteed, but is probably attainable. Take for example the test unit I had running 1.8TH at 70 watts, or the above-photographed machine which pulled down almost 2.1TH from 88 watts (on stock voltage, no voltage tuning) for over a week:


The Terminus R909 needs an external controller running cgminer, with USB connection to the miner (a 6-foot USB cable will be included with purchase). The driver has been tested extensively on linux PC and Raspberry Pi. Like previous Terminus models, the miner runs on 12V power from either a standard 2.1/5.5mm barrel jack (rated for 8A) or a PCIe 6-pin jack.

The "stock" speed will be set to 400MHz, expected to see 1.6TH, but higher frequencies are possible. Kano will explain the details, but he's been working on adding tuning features which include setting individual chip frequencies. Core voltage is adjustable in the range of 1.4V-1.6V per chip, same as on the Compac F, and though the R606 had a fancy push-button interface, we've rolled back to a basic knob you turn with a screwdriver on this one. That may be improved on future versions but I won't make promises just in case.

This miner also features Blinky Lights, but they're not as intrusive as previous models of pod or stick. The rate of blinking is proportional to hashrate. The base light color is a delightful pink-magenta.

The BM1397 chips have a significant temperature coefficient, which means that they pull noticeably more power as they get warmer. Nobody at GekkoScience will be offended if a tuner or enthusiast upgrades the fan, as more airflow could mean reducing power consumption by 5-10% just from maintaining lower temperatures. These units take a standard 80mmx25mm 3/4-wire 12V case fan.

And like all GekkoScience products, the Terminus R909 is Made In The USA.

Approximate dimensions: 3.5 x 4.2 x 6 inches.

---

So far, two resellers have confirmed to stock this machine, and should have stock in time to start delivering orders by Christmas. Production is already well underway on the first batch. This post will be updated as more sellers are added.

bitcoinmerch.com shipping from California
419mining.com shipping from Ohio
bitshopper.de in Europe/Germany (forum user MacEntyre)

The Terminus R909 will include a 6ft USB cable. It'll be up to resellers to provide any extras, power supplies or anything.

Dang, I might have to hire one of you guys for some embedded dev on a new project for Gekko.

There was a change in the USB connector from the original batch to all successive batches (2022 models). In testing, I've had *fewer* issues with mechanically reliable connections with the new jack than the old one, and if there's an issue in testing it gets replaced. I have no experience with the issues some are seeing, and no solution except using a different meter.

Bitfury are rolled cores, so multiple clocks per calculation. Bitmain's always used unrolled cores, which take up a lot more die space but do the whole hash in a single cycle. BM1384 has 55 cores, 1387 has 114 cores.

Fuzzy, the S17 chip has 672 cores. Interestingly, the chip actually has four blocks of 168 cores each built in series within the die.

Probably every USB meter except the one you're using, given yours is the first time I've heard of one not working. There was no change to the USB hardware in the last year or so that would affect physical interfacing.

As a test of a redesigned regulator, I set up a Compac F on our regular tester (air cooled) at 600MHz and ran it 12 hours no problem. I've talked to a guy that runs a few at 750MHz I'm pretty sure with air cooling, pulls about 25 watts per stick. Between the huge die and copper plating, the package cools very efficiently.

At that point, the fixed load of fans and controller probably start to become a more significant part of the overall power draw. It's kinda like how at low speeds in a car, rolling resistance will dictate your energy use more than wind resistance though wind resistance dominates at high speed.