Topic: Official FutureBit Apollo LTC Image and Support thread - page 48. (Read 49878 times)

spec sheet  shows a problem to me


EVGA   
EVGA 450 BR, 80+ BRONZE 450W, 3 Year Warranty, Power Supply 100-BR-0450-K1
Part Number: 100-BR-0450-K1

The EVGA 450BR Power Supplies are here. Designed with both beginner and enthusiast needs in mind, EVGA power supplies are the best choice to power next generation computers. Combining 450W of continuous power with exceptional performance, stunning efficiency, and reliable quality, the 450W BR power supply is what you need to build a solid foundation for your next system.
FEATURES
80 PLUS Bronze certified, with 85% efficiency or higher under typical loads
Single 12V. Rail
Active Power Factor Correction (PFC)
Heavy-duty protections, including OVP (Over Voltage Protection), UVP (Under Voltage Protection), OCP (Over Current Protection), OPP (Over Power Protection), SCP (Short Circuit Protection), and OTP (Over Temperature Protection)
3 Year Warranty
DC-DC Converter improves 3.3V/5V stability
Quiet and Intelligent Auto Fan for near-silent operation
CABLE LENGTH
ATX Cable   1x 550mm
EPS Cable   1x 600mm
PCIe Cable   1x 550mm, 670mm
SATA Cable   2x 450mm, 570mm, 690mm
Four-Pin Peripheral Cable   1x 450mm, 570mm, 690mm + 810mm (Floppy)
Floppy Adapter   1x
AC Power   1x 1200mm (m)
Modular Cable Type   No
CONNECTOR QUANTITY
24 Pin ATX   1x
EPS (CPU)   1x 8pin (4+4)
PCIe   2x 8pin (6+2)
SATA   6x
Four-Pin Peripheral   3x
Floppy   1x
INPUT
Input Voltage   100 - 240 VAC
AC Input   8/4A
Input Frequency Range   50 - 60 Hz
Efficiency   85% Typical
OUTPUT
Rail   +3.3V   +5V   +12V   +5Vsb   -12V
Max output   20A   20A   37.5A   3A   0.3A
450W
120W   450W   15W   3.6W
Total   450W @ +30C
ADDITIONAL DETAILS
Operating Temperature   0° to 30° C
Japanese Capacitors   No
ECO Mode   No
MTBF   100,000 Hours
Net Weight   4 lbs
Size   85mm (H) x 150mm (W) x 140mm (L)
Approvals   TUV, CB, CE, FCC, RCM, cTUVus, WEEE, RoHS
Compliance with ErP Lot 6 2013 Requirement
Fan Size / Bearing   120mm Long Sleeve Bearing
UPC   843368051815
EAN   4250812423542
EVGA Warranty
PRODUCT WARRANTY
This product is covered under EVGA's 3 year limited warranty which covers parts and labor.
Further warranty extension is available upon registration within 90 days of purchase.
For more details please visit http://www.evga.com/warranty/


I bolded the problem

1 cable with a dual plug

so you will be doing more then 220 watts on one cable.

that is not good.

EVGA 450 BR 450W 80+ Bronze PSU this https://www.evga.com/articles/01212/EVGA-700-600-500-450-BR-Power-Supplies/

which bronze psu?  some are not well made.

does yours have 4 separate   pcie cables

or will it have 1 dual head pcie cable for each piece

If it uses 2 pcie cables and if they are 18 gauge it should work

if the pcie cables are 20 gauge you could be in trouble.

Thought so. Thanks!!

Yes, the eco mode of a single Apollo requires about 130-140W from the 12V rail.

Quick question - could I run 2 apollos in eco mode off a 450w bronze PSU?

philipma1957 is 100% on target. He very tactful told you that you were asking the wrong question. Nobody faults you for not knowing why or how because you are still learning, but your question was like asking "will blending ethanol with gasoline make my car go faster?" Increasing a unit's hashrate post manufacturing through a software update rarely happens. When it does happen, the increase is typically anecdotal at best.

(1) Improvements to the voltage/efficiency of the ASIC chips could theoretically improve hashrate if the limiting factor is thermal. However, I believe that we are likely hitting frequency barriers inherent to the ASIC's architecture, but it is possible I am wrong.

(2) It is understood that functionality equates to features. This could lead to increased hashrate through increased frequency if that is the feature updated, but I would again state the likely frequency barriers of the architecture argument from (1).

philipma1957 is probably the most user friendly forum member willing to help newcomers. Having personally read a great many of his almost 30k posts over the last 6 or 7 years, I believe his statements regarding his gear and setup were meant to lend credence to his advice not gloat. You would do yourself a service to avoid expressing prejudice by listening a little more carefully and speaking a little less harshly. In case I am unclear, stop being rude.

Has there been consideration on using the 512MB variant of the MCU board for Batch 2 units?

Yea we are dealing with the memory limitation and is one of the reasons we are testing different approaches to enabling full node functionality on the Apollo. Were looking into ways to do a full node download instead of syncing from scratch.

Such central data point design may not even be necessary. The Apollo UI uses Minera which already supports multiple devices in one interface based on screenshots. In order to avoid needing a central data point, it could be possible to extend Minera in two ways: 1) add an http endpoint for data export of the "primary miner" using some industry standard data format (json), 2) add the option of introducing miners to Minera using that data provided by (1). Such solution would most likely require less new implementation than a completely new solution for data storage. As a result, it would be possible to configure one Apollo UI to include the data on the fly (=client side) from other units so no data storing would be required except for the endpoint configuration of the other units, all data would be visible from one interface.

PS. Minera also appears to provide some data history charts you were looking after.

I was thinking to avoid the system to swap during initial sync of the blockchain. Regarding the poor memory on the board, https://i.imgur.com/GWmgOri.png, dphys-swapfile will be used and thus lot of read-write to SD card.
I made full LTC and full ETH on Raspberry Pi and used an external SSD with full system migrated on it + swap partition.

jstefanop would really need to come up with some strange implementation in order to block such possibility.

The blockchain is just a collection of files in some directory. Since access to the mcu and its filesystem isn't blocked, reallocating a directory to a connected external disk will be possible even if such option doesn't get implemented in the user interface. I don't however think that having the blockchain on the sd card could cause an issue with card wear as the blockchain progresses with a clear pattern instead of random writes. Random reads on the other hand don't really matter as long as the read performance of the sd card doesn't become the bottleneck. The benefit on having the blockchain on an external drive (or usb stick) would be to keep the option making the Apollo software upgrade with full sd card overwrite without having to start from scratch with the blockchain.

On the UI - I like it, easy to use and intuitive. If there were any big updates to the UI it would be productive to see graphs in addition to just values. For example, the miner temp is 66.31 deg - nice, but did it spike sometime while I was not watching? Are those rejected counts over a linear period of time? Are those 0.7% hardware errors a gradual accumulation or did 0.6% happen in a sudden spurt? Seeing trends are productive.

It would also be nice to see overlays from other Apollos but that would require a central point of data accumulation - One Apollo computer set up as a primary with other units feeding data as secondary with the user display that of the primary. Adds lots more complexity. Might be a nice marketing incentive to buy more units to see more graph overlays, lol.

Would be great for full node to have the possibility to sync the blockchain to an external drive so that it preserves SD card

I've got 2 as well and use this https://i.imgur.com/PDY5ruN.jpg
It's using one usb port of my ASIC and cooling down really well both MCUs thus Miners. In that way I can run both in balance mode with 10% fan speed.

Hey,

any work done around the GUI? Would be great to have the capability to manage with one ApolloUI instance the rest of the ASICs.
Cheers

Thanks whiteogre! - pretty much what I thought. I have purchased 2 so far and still learning about crypto stuff. I know efforts, as mentioned by jstefanop, to bring full node to Apollo. I understand what that means but not sure how that would impact mining. For example, it seems like less network latency time would be realized so there should be less stale shares. Just guessing.

I did put a usb fan in front of each which helps as they purr perfectly in balance mode.

Kind of fixed. There's some random element involved in the quality of the ASICs you have but that can't be changed after manufacturing. As a result, I'd bet the operation modes currently provided in the Apollo software are safe (or worst case) averages to ensure the software works without issues with almost all units. The variables that can still be changed are:

• voltage
• frequency
• temperature
• bfgminer
• other parts of the Apollo software

Playing with the voltage and frequency values could provide some unit specific efficiency or speed gains, as was the case with the Moonlanders. Temperature is obviously also part of the equation if you go for maximum speed. Since bfgminer isn't static, some new version could theoretically improve something. The other parts of the Apollo software are mentioned mainly from efficiency point of view as the MCU also has some consumption but it really doesn't matter when compared to what the ASICs need.

I'd guess some future version of the software could include some kind of autotune for the voltage and frequency values if speed or efficiency is the target at some specific hash rate but I don't see that as a priority and there hasn't been any discussion regarding such being even in development.

My question was legitimate. The first couple of lines of your reply is obvious and the stats of the Apollo are a given. To rephrase my question - are the hash speeds a fixed limitation of the hardware?

I prefer you don't answer philipma - please don't force others to read how wonderful your setup is nor your gloating. It was my understanding this was an Apollo thread - not a boasting with non-Apollo hardware one (or you have 1?).

BTW - you said "do not knock voltage and efficiency gains bro" - well gee bro, I never said I knocked them and in my first line said they are always welcome.

BTW2 - jstefanop's past entries have stated their market target - and it didn't include you, so back off and stop attacking beginning Apollo users just because you don't like how a question is phrased.

BTW3 - Yeah, I'm small scale as are many others here. Scales are relative so just remember your a minnow dweeb compared to the really big fish miners.

 to be fair making some money is part of the game. So I understand your concern.

So the scoop is
This is the most efficient miner there is for LTC so watt per hash no better.
This is not the cheapest miner there is $ per hash for LTC.

I know this miner can run quietly and be used in many spots.
Mine will mine 24/7/365 even though it most likely will lose money
I purchased it to support the designer.
The designer gave out valuable  software for the bitmain L3+
I save 16 x 100 = 1600 watts of power 24/7/365 on my solar array hosted L3+ gear.
So I tossed him some LTC and I purchased one of these.
It is solo mining Ltc and Doge  I hit 2 doge blocks so far.
If I hit a LTC block I will give the op a LTC coin.

As for a node I run a node and a full ltc wallet  but I have 30 pcs and 100 pcs of different mining gear.

You look to be smaller scale.
So it is a low end machine for power cost it is very efficient.
it is stable.  you can actually earn some coin with it.  all good things.

do not knock voltage and efficiency gains bro = most imporant numbers at the moment.