Topic: PhoenixMiner 6.2c: fastest Ethereum/Ethash miner with lowest devfee (Win/Linux) - page 309. (Read 784965 times)

I got this error
2019.10.24:07:22:23.857: main CUDA version: 10.0, CUDA runtime: 8.0
2019.10.24:07:22:23.986: main No OpenCL platforms found
2019.10.24:07:22:23.986: main Available GPUs for mining:
2019.10.24:07:22:23.986: main GPU1: P106-100 (pcie 1), CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.986: main GPU2: P106-100 (pcie 2), CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.986: main GPU3: P106-100 (pcie 3), CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.986: main GPU4: P106-100 (pcie 5), CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.986: main GPU5: P106-100 (pcie 6), CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.986: main GPU6: P106-100 (pcie 7), CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.986: main GPU7: P106-100 (pcie , CUDA cap. 6.1, 5.9 GB VRAM, 10 CUs
2019.10.24:07:22:23.998: main NVML library initialized
2019.10.24:07:22:24.104: main Nvidia driver version: 431.60
2019.10.24:07:22:24.106: main NVAPI error in NvapiWrapper.c:101 : -6
2019.10.24:07:22:24.106: main Unable to load NvAPI: error 3
2019.10.24:07:22:24.606: main Eth: the pool list contains 2 pools (2 from command-line)
2019.10.24:07:22:24.606: main Eth: primary pool: ssl://eu1.ethermine.org:5555
2019.10.24:07:22:24.606: main Starting GPU mining
2019.10.24:07:22:24.606: main GPU1: unable to init NvAPI with bus ID 65536
2019.10.24:07:22:24.607: main GPU2: unable to init NvAPI with bus ID 131072
2019.10.24:07:22:24.607: main GPU3: unable to init NvAPI with bus ID 196608
2019.10.24:07:22:24.607: main GPU4: unable to init NvAPI with bus ID 327680
2019.10.24:07:22:24.607: main GPU5: unable to init NvAPI with bus ID 393216
2019.10.24:07:22:24.607: main GPU6: unable to init NvAPI with bus ID 458752
2019.10.24:07:22:24.607: main GPU7: unable to init NvAPI with bus ID 524288
2019.10.24:07:22:24.752: wdog Starting watchdog thread

Has anyone been able to get any further stability at 60mhs with Rx5700's? I haven't tried in a few weeks.

Hi,
Sorry for the bad English.
I am using 12x P106-090 Manli brand. (6GB 192 Bit Samsung GDDR5 Memory With 5SM GPU)
The Manli website states that it is 21-22MH / s, but I can't exceed 16MH / s with Windows 1903 and HiveOS, and if I overclock, my speed doesn't increase above 16.5MHs. I have tried the most up-to-date nVidia 430.50 WHQL and Manli with the original stock driver. I downloaded P106-100 REV 1 driver from Gigabyte and the problem persisted. I completely uninstalled the drives with DDU, tried with Phoenixminers 4.7c and old version.
disabled driver signatures. Where's the problem? Please help me and thank you very much.

The following happened: the gpu(s) share the virtual mem address space with the host. Gpus already support using host-side mapped memory. Your allocation(s) were split across gpu mem pages and host mem pages by the driver. Many of your reads from the gpu was actually performed over PCIe, returning data from host mem. This is the cause of your crap hashrate.

Normally you see this behavior when you’re juuuust pushing past the 4GB or 8GB limit, ie the full vram capacity on the gpu. The driver still allows the allocation but part(s) of the buffer will hit host side mem over PCIe, not proper gpu vram. I’ve seen it a ton of times in CN mining when you push the nr allocated pads too hard. Seeing the driver allocating 60% of the DAG off-gpu is very surprising, but assuming you’re not BSing, that is what happened. Still won’t produce any useful hashrates so not really useful in the end anyway.

Look I get it. But it worked, in April, with the dag being over 2gb. It's quite possible that last bit of what you stated is what was going on. Card gave a very low hashrate, about 1.2mhs. I had set some settings in the bios I cant recall them except Legacy when installing the OS. Still working to figure out what I had done in April. 

the case you describe here: dag is being generated - it gets written into the GPU memory, but then you need do free up all the memory to allocate another(different) DAG. there're many factors that can fail -  intensive mem writes on overclocked and already heated GPU,  memory free up and allocation can also fail easily, driver can be an issue and not only the miner app... so...

it all however doesn't mean that you can mine eth on 2gb gpus - it's total nonsense. very simple: you need DAG(over 2 GB size) for algo to work, if somehow you manage to do segmentation and put only "relevant"() parts of the dag to the current calculations - you'll kill all possibilities of producing any meaningful results by constantly bottlenecking around reloading the dag parts into memory.

in any case you are highly welcome to substantiate your incredible claims.

Also plenty of people with 4gb polaris cards are hitting dag limits. Im one of them. My 6 card rigs have been online for years until now. Issue started around dag 285 on ETC. ETH gave the same issues but later because the dag is slightly behind. I run windows 10 with Claymore connected to Nicehash dagger/hash pools. Virtual mem is set to 98gb. When the rigs were being fed ETH jobs they would mine no issue but as soon as the pool would switch to ETC it could not load the dag and would sit restarting the miner until Nicehash would switch back to ETH or whatever dagger/hash coin they are pushing. Did a couple Windows 10 reinstalls just to be sure I wasn't having any corruption or OS issues.

Yea it was me. I'm sorry about how unorganized or legitimate it may seem. Its mostly a realization after Claymore 14.7 hit 299 and my 4gb Polaris cards started having dag issues. I tried each of my 4gb rigs with a 580 8gb thrown in and it started working. Back in April I built a desktop build with a r7 200 2gb and a 570 8gb for my self. It no longer exists because I dismantled it after 2 months. Now I cant get the r7 200 to work the way it did again. I'm still troubleshooting on my free time see if I can replicate the results. I want to show it off. I feel stupid saying it worked and no proof. But the build was a Msi z270a with a Kaby Celeron and 8gb of ram. Windows was installed in Legacy mode. I found that be crucial. UEFI mode is better for newer boards, pcie accessories, and mining but I I was forced to use legacy because the 570s did not like working together with the r7 200 in UEFI. It worked like a charm in Legacy. The r7 took about 50 seconds to load the dag and it gave a PITIFUL 1.2mhs. But it worked. As much hate as I'm getting over this I'm trying to replicate the results with my little free time. When I do I will be posting on Reddit and here with screen shots. Hope this helps clear a little confusion. Cheers.

cannot tell you anything about "what they would have to gain" but that's really not that critical. it all sounds like confusing horseshit starting from "realization" of 4gb gpus "not generating the dag any longer"

@Bittap So you are saying that it's you who started this crazy claim... cool. Care to elaborate on the details? maybe screenshoots of 2gb gpu hashing eth? lol.

If you check out the comment I was responding to it mentioned that they are able to get 2 gb AMD cards to work provided they have one card in the mining rig with 8gb for the dag to populate on. I took them for their word and tried it with Nvidia but apparently it only works with AMD cards. I suppose I could find a 2gb AMD card and mix it with a 8gb card and see if what they say is true, but I do not really see what they would have to gain from making something like that up.

2Gb,4Gb DAG fix, need atleast 1 GPU large enough to generate DAG.

Recently realized in my farm that my 470 4Gb rigs, which were all 6 card setups at the time, were not generating the dag any longer for ETC and then soon after ETH dag approached whatever level it was that pushed my system over the limit. After a day of wondering what to do, I realized a 470 rig I have with 1 8Gb 580 in it was working fine with the other 5 cards being 470 4Gbs. I took apart one of my 580 8Gb rigs and replaced a card on one of the 470 rigs. Mined perfectly fine. Did this to the whole farm and split the gpus essentially so every 4Gb rig has atleast 1 8Gb card in it. As for now this works no issues. I have a p104-100 rig with 1 470 4Gb in it and same thing, generated dag and mines.

I tested this on 2Gb R7 200 gpus and works perfectly fine. Card generates the dag as long as the system has one available card that is able to fit the dag on it. This is identical to how the 470 rigs are operating.

The same goes for 470 4Gbs with rx 5700s and the same with the R7 200s with the rx 5700s.

It sounds like 4GB isn't enough for the DAG since the current DAG is 3.27GB for ETH or 3.34GB for ETC.
RedPandaMining recently posted a video on his Youtube Channel.
https://www.youtube.com/watch?v=xRIX5mpVntk

your discussion is seriously weird. it is impossible to load >3gb array into <3 GPU memory whatever tricks you may perform.

DAG  - directed acyclic graph

"The Ethash algorithm expects the DAG as a two-dimensional array of uint32s (4-byte unsigned ints), with dimension (n × 16) where n is a large number. (n starts at 16777186 and grows from there.) Following the magic number, the rows of the DAG should be written sequentially into the file, with no delimiter between rows and each unint32 encoded in little-endian format."

"...calculating the PoW (Proof of Work) requires subsets of a fixed resource dependent on the nonce and block header. This resource (a few gigabyte size data) is called a DAG. The DAG is totally different every 30000 blocks (a 100 hour window, called an epoch) and takes a while to generate."