Bitcoin Forum
December 08, 2016, 06:12:34 PM *
News: To be able to use the next phase of the beta forum software, please ensure that your email address is correct/functional.
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 [4] 5 »  All
  Print  
Author Topic: Sound card mining?  (Read 19969 times)
shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
June 27, 2011, 06:50:21 AM
 #61

Stop filling the tropic of stupid posts.

Quote
As far as I understand it, you can give the DSP access to up 2megs of data in your normal memory, the data however can only be read/written in 32 16-bit parts per cycle of the dsp.
The emu10k1 got 32 registers to access data on the xtram, for both reading and writing. You can change the adresses of those registers for the next cycle only.
So generally it should be possible to use the 10k1 for data processing that has nothing to do with sound.

1 minute of google.

OP post about that guy is its bs, but the concept is its not.

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
1481220754
Hero Member
*
Offline Offline

Posts: 1481220754

View Profile Personal Message (Offline)

Ignore
1481220754
Reply with quote  #2

1481220754
Report to moderator
1481220754
Hero Member
*
Offline Offline

Posts: 1481220754

View Profile Personal Message (Offline)

Ignore
1481220754
Reply with quote  #2

1481220754
Report to moderator
1481220754
Hero Member
*
Offline Offline

Posts: 1481220754

View Profile Personal Message (Offline)

Ignore
1481220754
Reply with quote  #2

1481220754
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1481220754
Hero Member
*
Offline Offline

Posts: 1481220754

View Profile Personal Message (Offline)

Ignore
1481220754
Reply with quote  #2

1481220754
Report to moderator
bcpokey
Sr. Member
****
Offline Offline

Activity: 462


View Profile
June 27, 2011, 07:17:26 AM
 #62

Even at the theoretical maximum (which is unlikely to be achievable) the processing power of a soundcard is fairly low for it's cost. I believe somewhere weeks ago I posted something about how a soundcard would be like a crappier version of an ASIC if anyone developed the ability to hash on them. Low power consuming, low hashrate producing, high cost items. I suppose for those who are really batass crazy about power consumption and don't care about initial costs can find this appealing, but as no one on the boards seems to have the technical expertise to rip apart soundcard coding, there doesn't seem to be much to discuss.
shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
June 27, 2011, 03:42:33 PM
 #63

you have any stimated hashrate?

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
bcpokey
Sr. Member
****
Offline Offline

Activity: 462


View Profile
June 27, 2011, 04:35:51 PM
 #64

EMU20K1 estimated processing power is placed at 10,000 MIPS

A radeon 4850 (quickest thing I found with my googlefu) is estimated at about 500,000 MIPS. For a sense of what this means, a 4850 gets about 85MHash/sec according to the hardware comparison wiki. So 85 / 50 would lead one to expect about 1.7Mhash/sec if one managed to make a soundcard hash.

You could have just read this thread and found this. It's on the first page. Hence why I want this thread to just die.
shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
June 27, 2011, 06:17:11 PM
 #65

But you can compare processing power just like that? that will hold true only if the sound card was using OpenCL.

I was wondering, if the EMUx chips can read/write 32 chunks of 16bit of data per cycle it doest mean it can get all IO in a single cycle?

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
bcpokey
Sr. Member
****
Offline Offline

Activity: 462


View Profile
June 27, 2011, 06:21:20 PM
 #66

No, obviously you can't do a 1:1 comparison, but it's the best for now since there exists no way for soundcards to hash at all.

Just a raw comparison of computing power. And it looks abysmal. It seems fairly unlikely that someone will manage to figure out a way to improve beyond 100fold some optimization for soundcards, which is what would be needed for them to be not totally worthless.
xenon481
Sr. Member
****
Offline Offline

Activity: 406



View Profile
June 27, 2011, 06:25:07 PM
 #67



True story

That does have the all important Rotate command that nVidia cards don't have.

Tips Appreciated: 171TQ2wJg7bxj2q68VNibU75YZB22b7ZDr
shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
June 27, 2011, 06:40:56 PM
 #68

No, obviously you can't do a 1:1 comparison, but it's the best for now since there exists no way for soundcards to hash at all.

Just a raw comparison of computing power. And it looks abysmal. It seems fairly unlikely that someone will manage to figure out a way to improve beyond 100fold some optimization for soundcards, which is what would be needed for them to be not totally worthless.

check this site...
http://emu10k1.tripod.com/_index.htm

Quote
Yes, a cool new driver set is scheduled to be released before this year is out, and here's a list of features which are already either fully supported or partially supported: Wave Recording & Playback, Midi In/Out, DirectSound 2D, DirectSound 3D (partial), Soundfonts, Synth Engine, ASIO Playback (with 5ms latency), DSP Microcode Loading/Unloading, Mixer with full AC97 I/O control, and last but not least, an Application Programming Interface (API).

Quote
DSP *updated and revised*

Programming and uploading custom DSP routines to the Emu10k1 is now within our reach. Below are some thoughts on the subject.

The original Creative Labs SBLive Linux driver source v2.0 files offer little info on any of the DSP-related functions or registers (see 8010.h), but the linux guys have done quite a bit more work on this end. Their latest stuff can be found here. Be sure to check out the Utils link.

The linux guys have also developed a microcode compiler - shouldn't be too hard to port to windows if needed.

*New* Daniel Bertrand (one of the linux DSP gurus) has done a superb job on documenting  the Emu10k1's architecture and DSP features. Don't miss his AS10k1 site - here.

The ALSA (Advanced Linux Sound Architecture) project has some DSP-related code too. Check out the rest of their independent Emu10k1 linux driver code here.

Valentin Yakovenkov's original apslive ac97 utility source code contains some Emu10k1 register dump utilities, check it out here. Unfortunately Valentin appears to be too busy these days to work on further development. *Updated* FxViewer picks up where Valentin left off!

*New* DSP-uploading via the APS FxEngineServer COM functions is now possible.
We now understand the workings of most the FxEngineServer functions.

FxToy calls FxEngine fuctions, including the following, for Fx program manipulation:

HRESULT  Load(long sizeIn/*[in]*/, unsigned_char* pgmText/*[in]*/, unsigned_long* ppgmID/*[out]*/) = 0; // [-1]
HRESULT Unload(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Start(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Stop(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Rampup(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Rampdown(unsigned_long pgmID/*[in]*/) = 0; // [-1]

See the FxToy page and the APS / APSLive page for more FxEngine info.



*New* Max M. has just released FxViewer, an absolutely brilliant utility application which allows us to observe the contents of the various Emu10k1 registers in realtime, as well as a providing an RIFX  (Emu10k1 DSP program file) "translation" window. Get FxViewer here, and check out the FxViewer page for more info. Hats off to Max M. for all the hard work that went into this one. :-)

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
bcpokey
Sr. Member
****
Offline Offline

Activity: 462


View Profile
June 27, 2011, 07:00:16 PM
 #69

No, obviously you can't do a 1:1 comparison, but it's the best for now since there exists no way for soundcards to hash at all.

Just a raw comparison of computing power. And it looks abysmal. It seems fairly unlikely that someone will manage to figure out a way to improve beyond 100fold some optimization for soundcards, which is what would be needed for them to be not totally worthless.

check this site...
http://emu10k1.tripod.com/_index.htm

Quote
Yes, a cool new driver set is scheduled to be released before this year is out, and here's a list of features which are already either fully supported or partially supported: Wave Recording & Playback, Midi In/Out, DirectSound 2D, DirectSound 3D (partial), Soundfonts, Synth Engine, ASIO Playback (with 5ms latency), DSP Microcode Loading/Unloading, Mixer with full AC97 I/O control, and last but not least, an Application Programming Interface (API).

Quote
DSP *updated and revised*

Programming and uploading custom DSP routines to the Emu10k1 is now within our reach. Below are some thoughts on the subject.

The original Creative Labs SBLive Linux driver source v2.0 files offer little info on any of the DSP-related functions or registers (see 8010.h), but the linux guys have done quite a bit more work on this end. Their latest stuff can be found here. Be sure to check out the Utils link.

The linux guys have also developed a microcode compiler - shouldn't be too hard to port to windows if needed.

*New* Daniel Bertrand (one of the linux DSP gurus) has done a superb job on documenting  the Emu10k1's architecture and DSP features. Don't miss his AS10k1 site - here.

The ALSA (Advanced Linux Sound Architecture) project has some DSP-related code too. Check out the rest of their independent Emu10k1 linux driver code here.

Valentin Yakovenkov's original apslive ac97 utility source code contains some Emu10k1 register dump utilities, check it out here. Unfortunately Valentin appears to be too busy these days to work on further development. *Updated* FxViewer picks up where Valentin left off!

*New* DSP-uploading via the APS FxEngineServer COM functions is now possible.
We now understand the workings of most the FxEngineServer functions.

FxToy calls FxEngine fuctions, including the following, for Fx program manipulation:

HRESULT  Load(long sizeIn/*[in]*/, unsigned_char* pgmText/*[in]*/, unsigned_long* ppgmID/*[out]*/) = 0; // [-1]
HRESULT Unload(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Start(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Stop(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Rampup(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Rampdown(unsigned_long pgmID/*[in]*/) = 0; // [-1]

See the FxToy page and the APS / APSLive page for more FxEngine info.



*New* Max M. has just released FxViewer, an absolutely brilliant utility application which allows us to observe the contents of the various Emu10k1 registers in realtime, as well as a providing an RIFX  (Emu10k1 DSP program file) "translation" window. Get FxViewer here, and check out the FxViewer page for more info. Hats off to Max M. for all the hard work that went into this one. :-)

What does all that have to do with the weak processing power of soundcards? Ignoring the lack of OpenCL and whatnot. Sounds like it's basically akin to another CPU-miner really.
brunoshady
Sr. Member
****
Offline Offline

Activity: 264

Dubs Get


View Profile
June 27, 2011, 07:35:19 PM
 #70

No, obviously you can't do a 1:1 comparison, but it's the best for now since there exists no way for soundcards to hash at all.

Just a raw comparison of computing power. And it looks abysmal. It seems fairly unlikely that someone will manage to figure out a way to improve beyond 100fold some optimization for soundcards, which is what would be needed for them to be not totally worthless.

check this site...
http://emu10k1.tripod.com/_index.htm

Quote
Yes, a cool new driver set is scheduled to be released before this year is out, and here's a list of features which are already either fully supported or partially supported: Wave Recording & Playback, Midi In/Out, DirectSound 2D, DirectSound 3D (partial), Soundfonts, Synth Engine, ASIO Playback (with 5ms latency), DSP Microcode Loading/Unloading, Mixer with full AC97 I/O control, and last but not least, an Application Programming Interface (API).

Quote
DSP *updated and revised*

Programming and uploading custom DSP routines to the Emu10k1 is now within our reach. Below are some thoughts on the subject.

The original Creative Labs SBLive Linux driver source v2.0 files offer little info on any of the DSP-related functions or registers (see 8010.h), but the linux guys have done quite a bit more work on this end. Their latest stuff can be found here. Be sure to check out the Utils link.

The linux guys have also developed a microcode compiler - shouldn't be too hard to port to windows if needed.

*New* Daniel Bertrand (one of the linux DSP gurus) has done a superb job on documenting  the Emu10k1's architecture and DSP features. Don't miss his AS10k1 site - here.

The ALSA (Advanced Linux Sound Architecture) project has some DSP-related code too. Check out the rest of their independent Emu10k1 linux driver code here.

Valentin Yakovenkov's original apslive ac97 utility source code contains some Emu10k1 register dump utilities, check it out here. Unfortunately Valentin appears to be too busy these days to work on further development. *Updated* FxViewer picks up where Valentin left off!

*New* DSP-uploading via the APS FxEngineServer COM functions is now possible.
We now understand the workings of most the FxEngineServer functions.

FxToy calls FxEngine fuctions, including the following, for Fx program manipulation:

HRESULT  Load(long sizeIn/*[in]*/, unsigned_char* pgmText/*[in]*/, unsigned_long* ppgmID/*[out]*/) = 0; // [-1]
HRESULT Unload(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Start(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Stop(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Rampup(unsigned_long pgmID/*[in]*/) = 0; // [-1]
HRESULT Rampdown(unsigned_long pgmID/*[in]*/) = 0; // [-1]

See the FxToy page and the APS / APSLive page for more FxEngine info.



*New* Max M. has just released FxViewer, an absolutely brilliant utility application which allows us to observe the contents of the various Emu10k1 registers in realtime, as well as a providing an RIFX  (Emu10k1 DSP program file) "translation" window. Get FxViewer here, and check out the FxViewer page for more info. Hats off to Max M. for all the hard work that went into this one. :-)

What does all that have to do with the weak processing power of soundcards? Ignoring the lack of OpenCL and whatnot. Sounds like it's basically akin to another CPU-miner really.

probbably.. but maybe with much less power consuming..
shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
June 27, 2011, 11:22:09 PM
 #71

not to mention i can easily put 5 cards intro an Pentium MMX mb...

Anyway, its like the FPGA, paying 50USD for 10 mhash... why we cant consider sound card mining???

All the info and drivers capable of DSP programing are the KX drivers. there is info about in their forums.

From i see so far it seen very similar to a FPGA, you load a microcode to the processor.

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
phillipsjk
Legendary
*
Offline Offline

Activity: 1008

Let the chips fall where they may.


View Profile WWW
June 30, 2011, 05:25:59 AM
 #72

My Case fan puts out 18mh/s

True story

That does have the all important Rotate command that nVidia cards don't have.

Brushless fans like that do have some smarts for detecting fan position. That one even has limited I/O in the form of power+tachometer.

However, I suspect the controller does not have enough memory to actually achieve 18 millihashes/second. Wink

James' OpenPGP public key fingerprint: EB14 9E5B F80C 1F2D 3EBE  0A2F B3DE 81FF 7B9D 5160
syb3ria
Jr. Member
*
Offline Offline

Activity: 56


'buntusiast


View Profile
June 30, 2011, 07:01:30 AM
 #73

Imagine you put... say 15k brishless fans on a stands near wind turbines. You get 18 milihashes * 15 000 and you can plug all the fans to produce electricity and sell it for greater benefit  Grin
film2240
Legendary
*
Offline Offline

Activity: 994


Professional filmmaker/Freelance videographer


View Profile WWW
June 30, 2011, 11:37:49 PM
 #74

I will be the first to admit I don't know that much about sound cards but I'm still going to call shennanigans on 390Mhash/sec off a soundcard. Or any mhash/sec really

I didn't know you could attempt to mine on a sound card.I still think a GPU is more suited for the job though.Hmm maybe a sound card can supplement my GPU on both my computers (PC and 1 is MBP) even though the MHash/s may not be mind blowing :-)

[This signature is available for rent]
[This signature is available for rent]
[This signature is available for rent]
[This signature is available for rent]
yetis
Member
**
Offline Offline

Activity: 112


View Profile
July 01, 2011, 04:54:51 AM
 #75

It would be great to test this dspminer on my EMU20K2 (Titanium Fatality Pro)...  Smiley

shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
July 01, 2011, 11:56:11 PM
 #76

Well Kxproject drivers already has everything to do so... SDK and documentation.

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
ne1
Member
**
Offline Offline

Activity: 85


View Profile WWW
July 02, 2011, 01:53:20 AM
 #77

has anyone looked into the uad architecture? 
http://www.uaudio.com/uad-plug-ins/uad-2-quad/uad-2-quad-omni-v57.html
I know it's pretty specific and proprietary, but I have an extra uad-1 lying around.  I'd be awsome to put it to work.

film2240
Legendary
*
Offline Offline

Activity: 994


Professional filmmaker/Freelance videographer


View Profile WWW
July 20, 2011, 01:25:09 PM
 #78

I have a few questions to ask:
1.Is it actually possible to mine with sound cards? (or is this an elaborate joke)
2.if it's possible,show me the app that enables it and link to it for me to test out?
3.If mining off a sound card is possible,show me a list of sound cards that this can be done on?
4.Can mobos onboard sound cards like on my mobo Asus P6T SE and my MBP mid 2010 model do this?

Thank you.

[This signature is available for rent]
[This signature is available for rent]
[This signature is available for rent]
[This signature is available for rent]
TiagoTiago
Hero Member
*****
Offline Offline

Activity: 616


Firstbits.com/1fg4i                :Ƀ


View Profile
July 20, 2011, 10:53:53 PM
 #79

I have a few questions to ask:
1.Is it actually possible to mine with sound cards? (or is this an elaborate joke)
2.if it's possible,show me the app that enables it and link to it for me to test out?
3.If mining off a sound card is possible,show me a list of sound cards that this can be done on?
4.Can mobos onboard sound cards like on my mobo Asus P6T SE and my MBP mid 2010 model do this?

Thank you.


1. Yes. Might have started as a joke, but turns out there are cards with the capability.
2. I don't think anyone has made one yet.
3. I think those would be the cards supported by the kX drivers
4. I dunno, but i wouldn't expect them to.


edit: how efficient that would be is a whole'nother matter

(I dont always get new reply notifications, pls send a pm when you think it has happened)

Wanna gimme some BTC for any or no reason? 1FmvtS66LFh6ycrXDwKRQTexGJw4UWiqDX Smiley

The more you believe in Bitcoin, and the more you show you do to other people, the faster the real value will soar!

Do you like mmmBananas?!
shivansps
Sr. Member
****
Offline Offline

Activity: 346


View Profile
July 21, 2011, 12:11:14 AM
 #80

I think the only sound cards capable of doing something are the ones with an EMU10KX and EMU20Kx chips... those where capable of processing music effects by hardware. but im not sure, i have 3 SB Live Value of 1999, with the EMU10K1.

The problem is, that cards with that chip arent cheap at all, the Audigy SE, doest has a EMU chip...

Basicly it should work as an FPGA with the proper programing.

1NEvmZTwLhqBgr3h7LwZap1sEHsyf3QFuv
Pages: « 1 2 3 [4] 5 »  All
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!