Topic: Who ordered the AMD Radeon VII? Lets make a list of Miners / Hashrates / Coins - page 8. (Read 12978 times)

Running 6 cards on a onda riserless mobo, 1500w at the wall, 490mh/s.

1575 clock, 860mv, 950 mem phoenix miner.

What is your hashrate for the Radeon 7 on veil ?

I prefer NVIDIA rather than AMD, and that's indeed the advantages of choosing GPU because there was only a little decrease in its value after being used for some times.
Horrible price for second hand of ASIC when we are using ASIC to mine crypto but ROI will be even faster than GPU.
But as per POS transition of ethereum and that may become the worst thing for me to shift my mining rig to another coin.

I would also count the resale value of the GPUs in the roi calculations, as they will still hold their value pretty well 2-3 years down the line. That being said, AMD cards might not be a good long term investment seeing monero changing algo and eth changing algo and going pos in the near future.

Has anyone tried voskcoin's vii settings? He's getting 89mh/s at 250w

Lot of effort for under $4 profit a day, but well done for getting it profitable! On UK electric and GPU prices that's ROI in 800 days/2.1 years at current prices, just FYI. 

I'm thinking of purchasing some of these cards, if I ran say a 6 and 8 GPU RIG, how many watts roughly should it be using?

Would the ASROCK H110 PRO BTC+ Intel H110 still be any good? I am planning on "upgrading" my rig.... And would 2 1000W PSU be sufficient ?

Thanks for the info mate. Oh wow! 110c is safe ouch! That will definitely burn out the cards very quickly.
I've literally spent the last few hours trying to make changes, trying to reduce the heat on these gpus. The best I can achieve is 88c. I can't get it lower without impacting the performance.
At the moment it's running very smoothly @ 82 Mh/s on all 4 gpus. I think that's efficient, 910 watts in total. 


Hey, Nicehash Legacy v1.9.0.22 , running on Windows 10 64bit.

Check on your core volt.  setting it at 850mV doesn't actually make the card run at 850mV. 
Hwinfo would report something else.  It's tied to memory clocks/voltage too.  If you're running at 1000 Mhz+ especially.

To get good mhs outta vii the core needs to bumped up and vii loses efficiency.

What miner are you running? Also, are you using Windows OS or some other mining os such as minerstat?

Junction temp is the max of all sensors.  The gpu will start throttling around 110c, so really that is the threshold AMD considers ‘safe’, not 90.  That being said, lower temp = longer lifespan.

I’m uncomfortable w/ 80c, but these guys run hot - without drastically reducing performance, or altering the cards cooling solution, it is what it is.

It's been almost 2 days since I had the gpus. Progress is being made, I started with the rig running at very high temperature, around 105 Celsius. After, hours of tweaking and testing I have managed to reduce the "Junction" temperature drastically.

If you're using Afterbuner for overclocking/underclocking remember with the Radeon VII there will be two temperature to monitor :

GPU1 Temperature
GPU1 Temperature 2 (Junction temp)
 
Always remember "Junction" temp is very important as this is the temp inside the gpu, this must not exceed more than 90ish Celsius if you want the units to have longevity.

Using the settings below, I've been able to reduce temp from 105 Celsius to 80-82 Celsius which I think is huge improvement.

70% fan, 1600 core, 865 mV, 1000 memory, Power Limit -20

The above settings should give your gpus ;
70 C GPU and 80 Junction and around 900 watts

Average Mh/s @ 82

Remember my watts are based on 4 GPUs

What do you guys think about running gpus around 80 Celsius, is that too high?

I've got 82mhs with lowest 850mv voltage I was able to set. (~200watt)
Do you have lower voltage so your clocks are lower and hash is only 75?  What is the wattage?

I'm keeping the strategy "take the max clock from the min voltage", so I never run the card with lower gpu clock that it can handle with given voltage.

90-100 Mhs on Vii is possible not sweet spot in efficiency.

75 Mhs might be looking at sweet spot.

But i just aware if VII has tripled result on the hashrate power that already generated to mine ethereum use ethhash.

But that's enough and interesting, i will try to get some VII and try to create a comparison. Honestly, i wanna try to add more GPu and im feeling confused to choose which is better but im missing it if VII has more hash rate even it consumes doubled on its electricity compared with NVIDIA 1k series

Yeah, the 1600 comes with all 6+2s, and 6/6+2s.  Using the 6/6+2s on my vega 64s, I run the 6 to the riser, then split the 6+2 using this:

https://www.amazon.com/gp/aw/d/B07611QXG4?psc=1&ref=ppx_pop_mob_b_asin_title

Just make sure you get the pcie splitter version (and double check the wiring when you get them) - there are similar products which are wired for the cpu cables, and will cause problems.

One thing to note - while I’ve been running these for 1.5 yrs no problem, I don’t run my Vegas over 200w (I also have a seasonic PSU that suggests not running more than 225w through a single 6/6+2). You’ll likely be pulling more power w/ the vii, but technically the EVGA cables should be able to handle it.  The pins should be rated @ 11a, (or 8a at the very least), while I believe the cabling is 16g/13a (or again 18g/10a at least), which means a single pcie cable should be able to carry min 3 hots x 12v x 8a = 288w, or spec 3 x 12 x 11a = 396w.  Obv the adapters are an unknown quality - but you can at least check the wire gauge, then monitor wire/connector temp - slightly warm to the touch is fine, hot - not so much.

If you want something more reputable than these no-name Chinese adapters, I think you can get 1x8-2x8 splits from cablemod, but be prepared to pay for it.

---

I’m assuming that’s junction temp - and yes, that’s pretty hot.  I believe you’ll start throttling just over 110c.

My single vii runs at mid-70s junction around 70-80% fan, in a pretty tight atx enclosure w/ only middling exhaust, along w/ a 590 and a 8700k.  Obv your second vii is putting out a bit more heat than my 590, but you’re also open-air, so it’s prob a wash.  The main prob is your clocks/voltages.  As I mentioned earlier - 90mh/s is pretty inefficient (meaning a lot of extra heat as well)

Sweet! I appreciate the support mate.

Finally, I have managed to get 2x Radeon VII working, and yes it's stable on Nicehash.

I'm using the following settings ;
Core 950 mV, Power limit -10%, Core Clock 1750 MHz, Mem 1100 MHz, Fan Speed auto.

DaggerHashimoto @ 0.179 GH/s ; 89 Mh/s & 89 Mh/s
GPU temperatures ; 103 degrees & 105 degress

The only worrying thing now is the temperature. I've read on this thread from post temp hits 89-91. But mine they are all 100+ , should I be worried or is that normal?

Yes, use 2 of them. Each cable connects one slot from PSU into one slot of your GPU. The remaining 6 pin on that cable just ignore. If you dont have enough cables, just buy some of these cable mods for EVGA PSU's.

If you need more help, just open another thread especially for that problem. Its not really related to hashrate & co  

I wanted to ask. When connecting the 3rd Radeon VII GPU, instead of adapters, if I was to use 2 of these : https://images.evga.com/products/gallery/100-CK-1300-B9_XL_4.jpg , that would still work rite?
1 for each port on GPU as 2x 8-pin power is required, the 6-pin will not be used...

They dont fit, I'm nearly sure. But in case they will, you have to check the assignment of the pins. You can find that diagrams on their homepages or try to google them.

Nope, I dont have any of these Radeons, only tons of Vegas. Tested one and sent it back after testing.

If the corsair cables fits, would you say I'm good to use the cables?
Yeah mate, I always read comments before I purchase anything, but thanks for the heads up.

BTW, do you know any stable settings for Radeon VII to work with Nicehash?