Slight issue with review testing ,, opinion

[DeathtoGnomes]

I suggested a change to mining tests to use Nicerhash since its what most beginner miners use to get started. When I think about this, I think addressing mining testing should be like testing for gaming. In gaming you test across several games, I am sure there are server gui based miners that could be used to perform comparisons as well. IDK if @W1zzard likes this idea or not, but if we are going to argue about mining in reviews, lets atleast agree what needs to be done to try to make "most" everyone satisfied with the results.

 

[W1zzard]

I don't have a lot of time to test mining, and I'm certainly not gonna start modding cards. Especially not for new releases where no tools are available.

Can you link me to the download of Nicerhash? Google just gave me some marketplace

 

[dorsetknob]

Next thing you know your all be demanding w1zz test all the different alt coin's Performance
what you lot are asking for is so time consuming for a tester/reviewer
How the heck is w1zz supposed to set a baseline when there is a mirrard of different baseline hardware

 

[TheoneandonlyMrK]

https://www.nicehash.com

Again I mearly meant software tuned, no mods just tuned to mine , please see my other posts.

 

[W1zzard]

Ah, now I understand. But wouldn't the software update between reviews? And test different blocks at different difficulty levels?

 

[TheoneandonlyMrK]

Ah, now I understand. But wouldn't the software update between reviews? And test different blocks at different difficulty levels?

Yes but that's just the block reward that changes not the hash rate ,it makes less difference then you would expect.
The drivers do change, the mining software is occasionally optimised a bit yes but not so much it would still be referential , like driver optimization.

Given the vega example all i would expect is a run at 850 gpu clock at .9 of a volt( typically stable i go lower) ,and 1100 memory with power at -25% ,all by wattman for 42 Mh/s.
Nvidia ,typically minimum gpu clocks allowed with oem oc software, max memory clock before errors start occouring, ie Nvidias tend to reset at failpoint.

 

[W1zzard]

Given the vega example all i would expect is a run at 850 gpu clock at .9 of a volt( typically stable i go lower) ,and 1100 memory with power at -25% ,all by wattman for 42 Mh/s.

Yeah that's easy to request months after launch .. Imagine that it's 3 days to NDA lift. Vega sample just arrived, there is zero information in the public about it other than people on Reddit going crazy: "beats GTX 1080 Ti".
Need to fully bench the card and photos and do overclocking, power, heat, noise, and somehow not miss NDA.

Would you sit down for like 10 hours and try to find mining settings?

 

[TheoneandonlyMrK]

Yeah that's easy to request months after launch .. Imagine that it's 3 days to NDA lift. Vega sample just arrived, there is zero information in the public about it other than people on Reddit going crazy: "beats GTX 1080 Ti".
Need to fully bench the card and photos and do overclocking, power, heat, noise, and somehow not miss NDA.

Would you sit down for like 10 hours and try to find mining settings?

No worries I get that point but it could be made easier than that ie forget power use ie max power allowed and just higher memory Oc.
Or after the first review add mining results.
Or cut them or do folding at home instead maybe.

All just ideas though w1zzard.