So what would happen to Sia (price wise) after these Obelikses and A3`s hit the mining pools? The price would go down, logically, right? If those numbers are correct one unit would mine 300k sia monthly Why would the price of the coin go down? Net profitability will go down as more ASIC's come online, which means your 300k number will drop to 200k, 100k and below super fast. But the price of the coins will stay the same, may even increase as has happened in the past when Dash and LTC ASICs came out and generated a lot of new interest.
|
|
|
Also worth to mantion that The RJ45 cable between SRR and Extension board can be from 10-100 meters long 100 meter is more less total lenght that RJ45 (ethernet cable) could have. Its becasue SRR power supply is 24 Volts and both SRR and Extension Boards will work if the input voltage will be in range of 12-24 Volts. So if 24 Volts will be at the beginning of ethernet cable that is connected to SRR and at the second end it will rop to 14 Volts then Extension Board will still be working So limitation is Voltage. Also better ethernet cable = less voltage drop = cable could be longer. 100 meters is safe lenght but i probably could be even 300mters also od good quality copper cable And Second thing. Extension Board doesnt require addonitional Power Supply. But if the voltage drop is too high on very long distance then you also have option to connect second PSU directly into extension board Still i gues those informations are not needed for 98% users. Hi Tytanick, so these bards are daisy chained? Each board will passthru the voltages to the next board, correct? Lets say the first extension board is 10mtr away from the SRR, the next extension board which is an additional 10mtr away would plug into the first board and so on? This is a welcome change, I was expecting each extension board to need its own power supply. Looking forward to stocks of SRR2 and the extension boards
|
|
|
Any old timers here get any good results with legacy GPUs.
7970/280x I get about 620 H/s 290 I get about 700 H/s R9 270 with Hynix I get about 530 H/s
Anyone play around with their memory straps to get faster speeds?
You get The Stilt's straps for those, hes the original bios memory strap modded. Check out the overclocke.net forums. Yeah already running the Stilt bios for the 290. Doesn't seem to make much difference from stock with XMR. It helped during Litecoin days. Aah well, cryptonite benefits more from looser timings and higher clocks, than dem tight ass timings.
|
|
|
Any old timers here get any good results with legacy GPUs.
7970/280x I get about 620 H/s 290 I get about 700 H/s R9 270 with Hynix I get about 530 H/s
Anyone play around with their memory straps to get faster speeds?
You get The Stilt's straps for those, hes the original bios memory strap modded. Check out the overclock.net forums.
|
|
|
Guys, can you give me your opinion on the following please? I'm going to help a friend build his first mining rig. He got the Gigabyte GA-H110-D3A board. I gave him the entire list of components (including an HDMI dummy plug). Unfortunately I didn't realise that board did not have a HDMI video output... but a D-SUB.
So ideally I would like to use the IGFX as signal output to not use any of the mining GPUs for that task. Obviously the HDMI dummy plug won't work. Amazon UK doesn't seem to feature any D-SUB dummy plugs. I've got old VGA dummy plugs that I built myself back in 2014 using capacitors to short the pins and emulate a monitor... but will they fit a D-SUB output?
Any ideas?
But why do you need a dummy plug on an H110 motherboard? AFAIK dummy plugs were only needed on the H81's, on all other hardware they are optional (to overcome the 480p display cap on older boards) and even then if you force enable the onboard gfx on the H110 board, display scales 720p.
|
|
|
Been running 8 RX570 for over 3 weeks now , rock stable both windows and Linux , pulling around 800W for ETN and 1000-1050W for ETH . keeping it on ETN for the past two weeks because i have no forced ventilation yet .
not all amd, mine is mixed: i have 2x470s and a 460 on my d8p (other cards are 2x1080ti, 1070ti, 1070) runs fine, fwiw
I have the onda 6 board not the 8 but it has been running flawless for about 2 months now. I have 4 x vega56 and 2 x rx570 on it. Only thing I had to do was patch the drivers on the 570's after bios flash. 9000h/s @ 700w.
Thanks for your reports guys, sounds pretty encouraging. I guess I will order a batch of these boards then.
|
|
|
Hey hows peoples experience with running AMD cards like the 570/580/Vegas on the Onda D8P motherboards, all I see is them running Nvidia. I want to know since AMD cards, unlike Nvidia, pull a lot of juice through the PCIe slots, so if anyone's done an 8 card build and its running stable, I'd love to hear about it, maybe see some pictures.
|
|
|
The juicy part is the expansion coverage on the ASUS B250 Mining Expert. There are a total of 19 expansion slots of which 18 are PCIe Gen 3.0 x1 and a single slot is PCIe Gen 3.0 x16. ASUS has officially stated that the latest GPU drivers from AMD and NVIDIA only support up to 8 GPUs when running in multi-GPU setting. A new driver is said to launch in late 2017 by AMD that will enable up to 19 GPUs to be supported by the motherboard. Till then, users can populate up to 16 graphics cards by mixing 8 AMD and 8 NVIDIA GPUs on the motherboard. It still makes it the world’s first board to support 16 GPUs so far.
Doesn't support 16 GPU's, there's a hardware limit at 13 GPU's, to go beyond that you need to get Mining cards.
|
|
|
Its faint but I promise its there, just need to turn up the volume. Hard to catch on my phone but very audible in person.
Yup, this is GPU fans dying. I've had several GPU fans die on me so far, no biggie. Just look out for any GPU overheating, that's a good indicator of which fans dying.
|
|
|
Those voltages are pretty good too ... what are your settings and memory strap?
|
|
|
Anyone else using 1050 TI's? they are one of the only cards that's actually available lately. . lol
Great Video Vosk, you have really grown in knowledge over the last year And yeah, the 1050ti's are the only ones really available at somewhat reasonable prices, but I really want to know what memory are you folks getting? Are they all Samsung memory? Because the last time I bought GTX 1060's in a big batch, ALL of them turned out to be Hynix, Zotac, Inno3D, Palit, MSI, Asus - the lot, everything was hynix. The Current GPU Crunch, and the upcoming slowdown in production schedules owing to the Chinese New Year celebrations, mean the situation is not going to improve, so I want to stock up on GPU's.
|
|
|
For the life of me I can't get past 4 GPU's on this motherboard. I've tried all the recommendations throughout this thread.
I have confirmed that the GPU's and risers all work fine. I've plugged them in with different combinations of 4 x GPU's, and they all work. Just not when a 5th is added. All GPU's are RX580 8GB's. All work fine in Windows, up to 4 max.
-Latest ASUS drivers installed. -Latest BIOS installed. -4G enabled. -DMI/PEG/PCIe Speed all set to Gen2. I've played with many different variations of this.
After the BIOS screen, it just goes black and I can't do anything.
$50 to anyone who has experience with this motherboard, and can troubleshoot and get me up to 6 GPU's.
Change PCIe speeds to Gen1 Disable all Sata ports and USB ports not in use Ill take Ethereum
|
|
|
(or w/e built that for you and charged you PREMIUM COSTS, hence the ad)
Nobody charged him premium costs. The post was just an advert of his small business: posted by Marcinprv (where "prv" I guess stands for "first" in polish) and you can find the same name on the advertised website (firstever.eu). BTW two risers per sata is probably OK; more than 2 is a problem. Two risers per sata is ok IF using sata powered risers, the main issue is the sata-molex adapters that ship with the risers crap out and melt.
|
|
|
Cool setup. Btw, the RM1000x have 3 Molex cables each, so you can connect all 6 risers directly instead of using those sata-molex adapters that you are using in the image, its safer that way.
|
|
|
This looks super cool. Please keep updating the thread with photographs as you go along, we love us some Buildout-Porn
|
|
|
Hi,
I have Sapphire Nitro+ Radeon RX 580, factory Overclocked.
I have couple of questions on GPU tweaks:
1) The stock Memory Clock is 2000. What is the safe max limit/range of Memory Clock setting in which there is no/minimal harm to the GPU?
2) The stock Core Clock is 1411. What is the safe min limit/range of Core Clock setting in which there is no/minimal harm to the GPU?
3) What is the safe min limit/range of Core Voltage setting in which there is no/minimal harm to the GPU?
4) Does BIOS modding inflict any physical damage (in the longer run) to the GPU?
5) Do I need to further Overclock it provided it is already factory overclocked.
6) On Claymore, my GPU temp is maintaining at 61C and fans are 53%. Are these safe numbers for longer run? Are my fans running bit fast?
Looking forward for your valuable input.
Thanking you in anticipation.
Regards,
Rizwan
I would recommend you START READING the dozens of already present threads, instead of creating new threads for EVERY question you have. If you continue to spam the forums with unnecessary threads, we would have to ask the mods to start deleting them.
|
|
|
My asus motherboard cannot start gpu that I connect to the PCI Express x1 However I tried each card on the 16x and they work fine. I got the assus h270 plus. I would be glad if someone could help me work this out.... The motherboard has only two pairs of 16x and the rest four are 1x...
The Riser needs to be powered ...
|
|
|
If I need set core clock as +100 and mem clock as 500 then
Globalcore 100 Globalmem 500
Is this correct or should I add the numbers to my GTX1070 clock limit and give the total?
Below I found in manifacture’s site Graphics Clock 1506MHz Boost Clock 1683MHz Memory Clock 8000MHz
Can someone please help me understand ?
globalcore globalmem params uses for all rigs on config file. if all of your cards are same, its usefull. otherelse if you want to +100 use +100 on config file. cor RIGUNIQID +150 +150 +150 +150 +150 +150 (for 6gpu write 6 times) mem RIGUNIQID +500 +500 +500 +500 +500 +500 Wow so we can use the + and the incremental value. Thanks for the clarification. I was struggling with that set up as I just moved from win10 to ethOS. Just another quick question, all of my GPUs are same, so can I use the incremental value like +100 in global core, global mem to apply it to all GPU/rigs ? Voltage and clock sensitivity varies from GPU to GPU (even if all of them are the same make/model) so while a liberal increment/offset value applied to all cards would work, you will have to tune them individually to extract the maximum out of them.
|
|
|
Some new straps for polaris 20 cads ? old straps can't exceeds 30 for new batches or maybe it is drivers issues ? I returns 12 x Gigabyte RX 580 Gaming 8G because I can't exceeds good Hashrate
You returned perfectly good polaris cards because you were not able to exceed 30mh on them? What were you expecting exactly? 40mh per card? What made u have such absurd expectations? Do you have any mining experience? Do you understand that a better memory strap can help you push the last drop of mining performance out of a card, but it wouldn't change the card itself? Do you understand how graphics chips, or memory chips, or any kind of chips work? Do you know the term silicon lottery?
|
|
|
|