Here is a comparison between x22i and x16rt, tested on the same rig (4 x R9 Nano, stock ROM, Win7, Crimson 16.3.2) with the same settings on the cards (900MHz/500MHz/-72mV) and the same settings in the miner (--opencl-threads 2 --opencl-launch=18x0). Total power draw for the entire rig is measured at the wall, i.e. includes CPU (70W in idle).
x22i gives a constant 5MH/s per GPU at constant 650W total power draw.
x16rt gives 6MH/s to 12MH/s per GPU at 500W to 800W total power draw.
With x16rt the power draw changes dramatically every time the hash order is changing. Consider this and use lower intensity to not run into the max current breaker of the PSU if your PSU isn't oversized.
For x22i wildrig-multi-0.14.0-beta was used.
For x16rt wildrig-0.15.0.8-beta was used.
Topic: WildRig Multi 0.35.1 beta 2 multi-algo miner for AMD & NVIDIA - page 46. (Read 92105 times)
January 01, 2019, 04:52:30 PM
January 01, 2019, 08:13:49 AM
Does the miner WildRig Multi 0.15.0.8 mtp x16rt support mining solo ?
I tried but no success
January 01, 2019, 06:09:24 AM
Does the miner WildRig Multi 0.15.0.8 mtp x16rt support mining solo ?
I tried but no success
December 31, 2018, 02:22:03 PM
Won't explain everything, but some hard hashfunctions uses memory more and that's why you get lower temperature
Okay, that makes very much sense. I was suspecting something like this but it's another thing when the dev confirms it. Keep up the good work.
December 31, 2018, 02:17:43 PM
Yes, because as for x16r - it depends on current hashorder of job(you can see that colorized string, more red chars - lower hashrate, because harder hash functions are need to be executed).
Thanks. When the harder hash functions are to be executed, shouldn't the GPU temp remain at nominal? I see the temperature dropping during that time, implying that the GPUs are working too little on those harder functions. Or is my reasoning flawed?
December 31, 2018, 02:13:08 PM
Yes, because as for x16r - it depends on current hashorder of job(you can see that colorized string, more red chars - lower hashrate, because harder hash functions are need to be executed).
Thanks. When the harder hash functions are to be executed, shouldn't the GPU temp remain at nominal? I see the temperature dropping during that time, implying that the GPUs are working too little on those harder functions. Or is my reasoning flawed?
EDIT: Haven't used x16r before. I used this miner only for x22i for two months now and I didn't see the mentioned behaviour on that algo.
December 31, 2018, 02:01:02 PM
Is it to be expected that the hashrate when mining x16rt is varying quite a lot?
In my case it varies between 8 and 16 MH/s (for each GPU) over a few minutes, with all GPUs on a rig showing the same ups and downs synchronously.
Is this a property of this new algo or rather a bug in this very early release?
EDIT: When the hashrate goes down, so does the temperature and fan speed on the cards. So they obviously don't work as much as they should/could.
Yes, because as for x16r - it depends on current hashorder of job(you can see that colorized string, more red chars - lower hashrate, because harder hash functions are need to be executed). In my case it varies between 8 and 16 MH/s (for each GPU) over a few minutes, with all GPUs on a rig showing the same ups and downs synchronously.
Is this a property of this new algo or rather a bug in this very early release?
EDIT: When the hashrate goes down, so does the temperature and fan speed on the cards. So they obviously don't work as much as they should/could.
December 31, 2018, 01:57:44 PM
The last version this year with x16rt support and some improvements to mtp(still not GPU-only, but anyway). Happy New Year everyone! 
Thank you & Happy New Year. Any chance of linux release for x16rt?
My hero! You rock sir.
December 31, 2018, 01:52:25 PM
Is it to be expected that the hashrate when mining x16rt is varying quite a lot?
In my case it varies between 8 and 16 MH/s (for each GPU) over a few minutes, with all GPUs on a rig showing the same ups and downs synchronously.
Is this a property of this new algo or rather a bug in this very early release?
EDIT: When the hashrate goes down, so does the temperature and fan speed on the cards. So they obviously don't work as much as they should/could.
In my case it varies between 8 and 16 MH/s (for each GPU) over a few minutes, with all GPUs on a rig showing the same ups and downs synchronously.
Is this a property of this new algo or rather a bug in this very early release?
EDIT: When the hashrate goes down, so does the temperature and fan speed on the cards. So they obviously don't work as much as they should/could.
December 31, 2018, 01:38:26 PM
The last version this year with x16rt support and some improvements to mtp(still not GPU-only, but anyway). Happy New Year everyone!
Thank you & Happy New Year. Any chance of linux release for x16rt?
December 31, 2018, 01:04:13 PM
The last version this year with x16rt support and some improvements to mtp(still not GPU-only, but anyway). Happy New Year everyone!
Thank you & Happy New Year. Any chance of linux release for x16rt?
December 31, 2018, 10:40:39 AM
ADD SOLO mining pliz.
December 31, 2018, 09:33:10 AM
The last version this year with x16rt support and some improvements to mtp(still not GPU-only, but anyway). Happy New Year everyone!
December 31, 2018, 08:09:14 AM
andrucrypt, do you think x16rt will be ready to go Jan 1? Looking forward to it.
looks good
December 30, 2018, 02:02:51 PM
What is the hashrate for MTP on a RX580 ?
I have about 700-900 h/s on RX470/570/480/580 on different rigs with overclocking similair to Ethereum mining (GPU Core is slightly increased).Thank you
December 30, 2018, 01:59:58 PM
andrucrypt, do you think x16rt will be ready to go Jan 1? Looking forward to it.
December 29, 2018, 01:58:16 PM
What is the hashrate for MTP on a RX580 ?
I have about 700-900 h/s on RX470/570/480/580 on different rigs with overclocking similair to Ethereum mining (GPU Core is slightly increased). December 29, 2018, 01:41:56 PM
Well, Mr Andrucrypt, you've got work to do again since the release of CryptoDredge... There's approx 100% performance to be gained to be about on par with Pascal cards (at least for Vegas that is)... Come on, enough xmas turkey for you, get back to work! 
No more plans for MTP this year Probably only fee lowering to default scheme and release with x16rt support.Out of interest, do you believe there is more potential for AMD on MTP to come anywhere close to Dredge/krnlx's nvidia variants? Something tells me that with that power draw on MTP, we're unlikely to see anything earth-shattering from AMD for this algo at all.
Let's keep fingers crossed for you))) Wish you all the best!!! Maybe it can help https://www.codeproject.com/Articles/872477/Assembler-for-AMD-s-GCN-GPU
December 29, 2018, 01:08:01 PM
Well, Mr Andrucrypt, you've got work to do again since the release of CryptoDredge... There's approx 100% performance to be gained to be about on par with Pascal cards (at least for Vegas that is)... Come on, enough xmas turkey for you, get back to work!
No more plans for MTP this year Probably only fee lowering to default scheme and release with x16rt support.Out of interest, do you believe there is more potential for AMD on MTP to come anywhere close to Dredge/krnlx's nvidia variants? Something tells me that with that power draw on MTP, we're unlikely to see anything earth-shattering from AMD for this algo at all.
December 29, 2018, 08:50:42 AM
Well, Mr Andrucrypt, you've got work to do again since the release of CryptoDredge... There's approx 100% performance to be gained to be about on par with Pascal cards (at least for Vegas that is)... Come on, enough xmas turkey for you, get back to work!
No more plans for MTP this year Probably only fee lowering to default scheme and release with x16rt support.Out of interest, do you believe there is more potential for AMD on MTP to come anywhere close to Dredge/krnlx's nvidia variants? Something tells me that with that power draw on MTP, we're unlikely to see anything earth-shattering from AMD for this algo at all.
Jump to: