They plan on using a single 120mm rad to dissipate 650W? hah
At wall consumption. But still that's a lot of heat to dissipate, even for a thick push/pull 120 mill rad. I still think its possible, noise will definitely be a factor. similar discussion about the hashfast units. But its been shown that the corsair H80i can handle at least 250W of heat in the babyjet boards and supposedly will be able to handle up to 750W in the yolievo boards AFAIK, the most heat you would see from a high-end overclocked CPU is 400-550W, so im pretty interested to se how such a small liquid cooling solution handles at >600W |
|
|
|
tbh at $1500 the price wasnt that bad, and was fairly well in-line with other manufacturers. the biggest problem is that with BTC in the $400s it is hard to price the hardware low enough to look like a good deal. (particularly if the price jumped to $600, because most companies would then be called out on adding 20% to thier prices in order to stay close to the ideal price/btc/hshrate. That said, im not buying these yet. They are not standalone and appear to not have much of a chasis for mounting or stacking. Hopefully the next version will be similar to the babyjet, with a few of these boards in a case with an Rpi ontroller |
|
|
|
This has now been enacted.
good call. the two had been getting more and more alike in the last few months |
|
|
|
I think overclocking is going to be extremely limited, with four TPS53355 per board, or one for 16 chips. On the S1, it's one for 8 chips.
Thats the detail I've been waiting for  *really hoping for a PCB photo still* The S1 could consume around 45W per group of 8 chips overclocked (overclocking pushed the regulator right to its limits, but the real limiting factor was the voltage being set to 1.1V - some people have apparently pencil-modded the S1 to almost 215GH/~470W, or about 60w/8 chips) 45W/1.1V=40.9A on a 30A-rated component. (bitfury h-boards faced very similar amperage limits when overvolted around 0.95V) Its likely these chips run at around 0.8V and 1000W/640chips is about 25W per cluster of 16. That leaves a LOT of headroom on the regulators, but some of this is lost to the difference in voltage (more amperage). 25W/0.8V= 31.25A stock. Presumably these should be able to overclock without any issues from the regulator. I imagine all other components are beefy enough also. I imagine you will hit hardware errors at 15-20% overclock as with the S1. These may even be capable of up to 1.5TH is voltage was increased to ~0.9V 0.75V unloaded, so a little higher than that on the current. Still some room and I'd expect people should be able to get 20-40% overclocks, but not the 100+% ones we might have been able to do with 8 chips per vrm. I'm quite sure those puny heatsinks will not be enough when overclocked, 200Mhz+ will fry the chips unless some hefty cooling is applied! If you want to overclock the BM chip, the S1 was made just for that!! thats a losing attitude - anything can overclock! |
|
|
|
I think overclocking is going to be extremely limited, with four TPS53355 per board, or one for 16 chips. On the S1, it's one for 8 chips.
Thats the detail I've been waiting for *really hoping for a PCB photo still* The S1 could consume around 45W per group of 8 chips overclocked (overclocking pushed the regulator right to its limits, but the real limiting factor was the voltage being set to 1.1V - some people have apparently pencil-modded the S1 to almost 215GH/~470W, or about 60w/8 chips) 45W/1.1V=40.9A on a 30A-rated component. (bitfury h-boards faced very similar amperage limits when overvolted around 0.95V) Its likely these chips run at around 0.8V and 1000W/640chips is about 25W per cluster of 16. That leaves a LOT of headroom on the regulators, but some of this is lost to the difference in voltage (more amperage). 25W/0.8V= 31.25A stock. Presumably these should be able to overclock without any issues from the regulator. I imagine all other components are beefy enough also. I imagine you will hit hardware errors at 15-20% overclock as with the S1. These may even be capable of up to 1.5TH is voltage was increased to ~0.9V 0.75V unloaded, so a little higher than that on the current. Still some room and I'd expect people should be able to get 20-40% overclocks, but not the 100+% ones we might have been able to do with 8 chips per vrm. thanks for the actual value Still leaves a lot of hardware overhead, espescially looking at the PCB. improved cooling an an extra 600W PSU could allow some nice overclocking results. (how are the PCIe plugs wired - is it seperate 12V for each board or are they all sharing a common voltage? |
|
|
|
what the heck is the password.. root root will not let me in
Its an RPi in there correct? Should be u: pi p: raspberry |
|
|
|
I think overclocking is going to be extremely limited, with four TPS53355 per board, or one for 16 chips. On the S1, it's one for 8 chips.
Thats the detail I've been waiting for *really hoping for a PCB photo still* The S1 could consume around 45W per group of 8 chips overclocked (overclocking pushed the regulator right to its limits, but the real limiting factor was the voltage being set to 1.1V - some people have apparently pencil-modded the S1 to almost 215GH/~470W, or about 60w/8 chips) 45W/1.1V=40.9A on a 30A-rated component. (bitfury h-boards faced very similar amperage limits when overvolted around 0.95V) Its likely these chips run at around 0.8V and 1000W/640chips is about 25W per cluster of 16. That leaves a LOT of headroom on the regulators, but some of this is lost to the difference in voltage (more amperage). 25W/0.8V= 31.25A stock. Presumably these should be able to overclock without any issues from the regulator. I imagine all other components are beefy enough also. I imagine you will hit hardware errors at 15-20% overclock as with the S1. These may even be capable of up to 1.5TH is voltage was increased to ~0.9V |
|
|
|
The board is definitely the first-gen hashfats. AFAIK the yoli evo will be a different chip design in order to get higher speeds So there are gen 1 and gen 2 HF GN chips? after a closer look, i guess not. It looks like hashfast could basically push the chip from 400GH/250W to 500Gh to 800GH/720W. I jumped the gun a bit assuming it was a totally different chip revision that was able to double the speed http://hashfast.com/hashfast-announces-fastest-bitcoin-mining-chip-in-the-world/ so it is entirely possible with the right board design the prospector1 could actually push well past 500GH... ps: i was half-wrong on the H80i. apparently some new high-end CPUs actually do reach >550W when overclocked heavily. Presumably it COULD cool up to a 720W chip |
|
|
|
Switching to GHash.io just until I can get the average hashrate. Will post screenshots of it and then when I know they are working fine move them back to Eligius for Public Viewing.
you should keep on eligius - other than the annoying 22.5min delay it is the best pool IMO. either way, looking forwards to reports on the 1min, 5min and 15min from ghash |
|
|
|
the 747GH hashrate is an error, i would not be surprised if it was some issue with the reported difficulty being 256 vs 512 or other similar 2x issue. (ie: actual hashrate of 375GH) The board is definitely the first-gen hashfats. AFAIK the yoli evo will be a differnt chip design in order to get higher speeds (an h80i or similar cooling solution would seriously struggle to move more than 500W of heat)^i stand corrected. the same chip is in both the 400Gh baby jet and the 800Gh yoli evo. and the H80i can presumably cool an overclocked top-end CPU that reaches 550W or power draw |
|
|
|
Is it good or bad?
Plain English!
747Gh/s average so far.  Hah before we say that's what it is, what the pool reports is what matters. Is there often a big discrepancy between the local report and the pool report? there can be. his screenshot shows a 5% error rate which is pretty high. (should be 1% or less) his eligius page shows that it was in the 375-400GH range, but went offline for the last 10min or so. maybe tweaking up the hashrate a bit. |
|
|
|
They used to have a limit on their S1's. It kind of irks me they don't have any limits on the S2. Probably because of the turbulent ride BTC has been having lately, they couldn't afford to stick to ideals over profit margins this time around. can't say I blame them much, either...
That said, I have 1 S2 Batch 2 coming soon, so I too would like to know some of the questions others have been asking:
1) How loud is it?
2) What were the kinks, if any that faced in setting it up
3) Does it make the room an oven?
Mine is going out in a shed with half assed temp controls and airflow, so I'm curious what sort of accomodations I'll have to make. Drill another vent in the wall and add a giant fan on it, perhaps?
number 3 is entirely relative. its 1000W of heat no matter what. If 3x 360W antminer S1 was making it an oven, so will this. |
|
|
|
Lastly, rumor has it Bitmain is culling it's S1 inventory and discontinuing S1 production in favor of the S2.
thats my assumption. We are at the point where the size and weight of the S1 to run at 2w/gh is unneccessary. The S2 is much denser hashing speeds and is rack-mountable (the most common S1 complaint). presumably/hopefully there is an S3 coming in May that will be able to use a smaller chip design to further improve speeds and density |
|
|
|
i am certainly not as impressed as giga with my first shipment. none of the 6 hashed out of the box and only 1 could i even connect to the mgmt interface. prolly gonna be a long night of troubleshooting. each one of my 50+ S1's hashed out of the box - bring back the S1's...........
sounds like some sort of general network issue - you sure they are on the right IP subnet? |
|
|
|
For your enjoyment!  Super sexy 4u case with rack mounting brackets!  64 chips per board. 10 boards. Beaglebone powered.  64 * 10 * 1.6Gh/s per chip == 1024Gh/s any chance of a photo of the PCB? There are 10 S1 units worth of bitmain chips in there, meaning that if there are sufficient regulators on the PCB these things could likely be overclocked to 1.5TH/2kW with some cooling improvements and perhaps some pencil modding |
|
|
|
It's pretty logical that the ActiveMining scammers chose to associate with the HashFast scammers.
Both companies really don't care about their customers; everything is fine as long as they can get a few coins.
Otherwise I wouldn't even be surprised if kslaugther was selling the boards at a loss. But we probably will never know.
It is quite likely ActiveMining/VMC was a customer of Hashfast and just now they received the order placed in the last year. Looking at the begin of this fraudulent scheme it is possible to infer that ActiveMining/VMC fooled investor and customers with the prospect to create they own hardware, when in reality they intention was just sell re-branded hardware from Hashfast or Avalon. Therefore all promises to produce they own chip and boards was just a façade to hide the real plan, which relied in the promises of another enterprises. I dont think that was the original intention. however, incompetence and delays have brought ACTM to a point where re-branding hashfast hardware is the best possible business model at the moment. I imagine that the full cost of these boards must be in the range of $1000-1250. Ken desperately needs to sell the boards or get them mining ASAP if there's much margin to be made |
|
|
|
10 gridseeds and 1 R9 290 Windforce running on the PC.
2 antminer S1s each with their own 500w PSU, just in case one dies I only lose 1 miner and not both.
Couldn't help but laugh at the smoke alarm A good precaution but I just pictured that all going up in flames, haha. Actually that is rly smart, did not see anybody done that . What a smart man  would be really clever to wire it up to a relay for your power bars/outlets. Detector smells smoke, all your power to the miners is cut off to prevent fueling the fire (and further melting/burning wires). Of course a false alarm would cause a halt to mining until resolved |
|
|
|
Anyone in Canada received their S2 yet? Really like to know what happens at customs, and what value it's declared at.
UPS wanted $71 and change from me. thats what i paid for 3x S1 a week ago - pretty good cheap fee |
|
|
|
|
my last post got removed (presumably because i quoted someone else's criticism_ - so I wanted to re-ask it in a more concise manner:
1) How much does it cost to produce one of these fasthash boards? (all costs considered - pcb, chips, components, assembly). If you can't answer this for whatever reason, just say so.
2) You said in the VMC thread that you 'puchased the PCB'. This should say 'we had the pcb fabricated ourselves', since what you wrote sounds like the PCB was purchased from someone else who fabricated it in advance. Seeing as how you have bought the chips from Hashfast, it seems reasonable to ask whether you bought pre-fabricated PCB from them as well, particularly since you acknowledged that one of your assembly lines is the same assembly line that hashfast uses. please clarify if the PCB was bought or if you contracted out a fabrication for it using the hashfast design files.
|
|
|
|
|
Show Posts
