Search content
Sort by
Showing 20 of 567 results by MaxDZ8
@Lucazane
Please remove link to my miner from 'getting started'. I haven't had any chance to work on in about one year and half and the project is completed. There won't be any new releases for this project and I would expect competition to have catched up by now.
Again, thank you for taking notice. But I'm just getting tired of the occasional emails I get in which I explain this over and over again so I guess I might write a line publicly.
Please remove link to my miner from 'getting started'. I haven't had any chance to work on in about one year and half and the project is completed. There won't be any new releases for this project and I would expect competition to have catched up by now.
Again, thank you for taking notice. But I'm just getting tired of the occasional emails I get in which I explain this over and over again so I guess I might write a line publicly.
If you are going to burn 7 digits and you decide to go GPU be sure to get in touch with some GPU developers. If there are game developers around odds are they can get you in touch with someone doing GPU stuff. I'm not following development those days but I'm still inclined to believe some algos hold surprises. With your budget you can probably get someone really badass.
Going to kill perhaps. Those are not on the shelves yet and availability might be a factor. I've seen Nv 750 launched at 180 EUR at some e-shops around here.
For pure compute I don't think they make GCN1.2 obsolete, main problem Tonga has always been overpriced, especially in its 2GiB variant (bleh)!
For pure compute I don't think they make GCN1.2 obsolete, main problem Tonga has always been overpriced, especially in its 2GiB variant (bleh)!
Quote
To me it looks more like a die-shrunk 285/380
Nope. Node shrink + new architecture iteration. It is known they have new instructions.Quote
The 1070 however does about 27MH/s at the same bus type and speed, at 125W
A 70% price increase for 16% lower consumption? They will have do take down the upfront cost considerably. How expensive are 1070s for you?Didn't the OP basically describe Ethereum?
Nope. Eth runs his own PoW something along the lines of 'dagger-hashimoto', at a certain point it'll become PoS.In general, maintaining a data structure isn't 'wasting power' albeit this is indeed taken to extremes with blockchains.
The idea of running non-cryptographic PoW has been around for a while. There are a lot of technical issues to deal with and a few mindset shifts as well. I am positive GridCoin sorta deals with it in the viable way.
Main problem is GridCoin isn't running the computation itself but running on another system (BOINC). In the real world, we could discuss this is mostly the same thing... but no, that's not the 'non cryptographic PoW' at work. Does that make any difference? I cannot tell.
The idea of running non-cryptographic PoW has been around for a while. There are a lot of technical issues to deal with and a few mindset shifts as well. I am positive GridCoin sorta deals with it in the viable way.
Main problem is GridCoin isn't running the computation itself but running on another system (BOINC). In the real world, we could discuss this is mostly the same thing... but no, that's not the 'non cryptographic PoW' at work. Does that make any difference? I cannot tell.
Can you try mine? I haven't got any feedback from Tonga users yet and I've been told it has nice instructions for AES!
Beware it's little more than a toy and for the time being I'm not even actively maintaining it. It's a strange beast and users have reported performance is quite odd.
Also try running some games in the background! I'm interested in keeping them smooth!
Beware it's little more than a toy and for the time being I'm not even actively maintaining it. It's a strange beast and users have reported performance is quite odd.
Also try running some games in the background! I'm interested in keeping them smooth!
Well this works too :
System information
Computing devices (OpenCL 1.2)
AMD Accelerated Parallel Processing
by Advanced Micro Devices, Inc.
OpenCL 2.0 AMD-APP (2004.6)
full profile GPU, Graphics Core Next 1.x
Tonga
28 cores @ 985 Mhz RAM: 4096 MiB
Cache: 16 KiB
Local: 32 KiB Using
- , 66437 KiB
CPU,
AMD Phenom II X4 840 Processor
4 cores @ 3214 Mhz RAM: 8191 MiB
Cache: 64 KiB
Local: 32 KiB(emu) Unused
Mining performance
Hashing qubit @ 4184 KH/s
Performance of used devices
Device used Hashrate [KH/s] Nonces Since last result Diff solvedper second
Min Last Avg5 Max Found Bad! Low Stale
0, 3995 4184 4184 4217 34 34 0 0 2 0.000
Not as bad but the load on the card is the same like the copied tahiti kernel and sgminer ..
Regards
Did they (AMD) forgot the ROPS on again?
What do you mean by that?
Can we disable the GPU parts which are not used in the computing so as to save energy?
But I don't know if you disable them 'forever' to next reboot or what. I certainly wouldn't do that for cards which are not dedicated to compute only as shutting down the ROPs will make the hardware unable to run any graphics API in existence.
It was about time dude.
Can you try mine? I haven't got any feedback from Tonga users yet and I've been told it has nice instructions for AES!
Beware it's little more than a toy and for the time being I'm not even actively maintaining it. It's a strange beast and users have reported performance is quite odd.
Also try running some games in the background! I'm interested in keeping them smooth!
Beware it's little more than a toy and for the time being I'm not even actively maintaining it. It's a strange beast and users have reported performance is quite odd.
Also try running some games in the background! I'm interested in keeping them smooth!
The GPU uses some coalescing to minimize latency and maximize utilization of memory bandwidth. Isn't that likely to merge accesses and thus amortize?
Excuse me we are getting in area of hardware that I am not well researched.
Coalescing patterns are variable across implementations a common point across most hardware is occupancy ratio or hardware-based, context-switch-free multi-threading of sort. Intel calls it HyperThreading®, it's guaranteed to work on two threads whereas GPUs accomodate as many threads they can fit. In my experience is considerably more consistent and easy to use than exploiting coalescing patterns.Excuse me we are getting in area of hardware that I am not well researched.
Of course.
Note that GCN1.2 (Tonga) improved performance as well (it has an instruction to accelerate AES byte swaps) but apparently not many people bought those and it might require new kernels so IDK if the improvements materialized in the real world.
Note that GCN1.2 (Tonga) improved performance as well (it has an instruction to accelerate AES byte swaps) but apparently not many people bought those and it might require new kernels so IDK if the improvements materialized in the real world.
A fact that was when you use an intergrated GPU and buff up the amount of RAM, that it will get you better performance.
It is a fact under certain situations. Nowadays, if you bump up from 4GiB to 8 sure your games will run better.Is this also happening when you are CPU mining?
No. Mining consumes negligible amount of memory (often under a single texture). Some algo consumes more but they're still relatively rare.Which cards would you recommend me? It's better to get a bigger amount of cheaper cards or getting les cards but more expensive?
I would suggest against buying at all (be sure to run your numbers first, ROI can get in the years super easily).Most users around here mine upper midrange. As a side note, NV cards have regularly improved kernels and I'm not sure what's the current situation with AMD. In general big card is better than small, just avoid top models.
That's absolutely cool man. Too bad it won't be considered hardcore because the HW isn't hardcore by itself but you have my respect!
The plasma speaker is badass stuff too.
The plasma speaker is badass stuff too.