[hashcash] Re: hashcash 1.04 released
- From: Adam Back <adam@xxxxxxxxxxxxxxx>
- To: hashcash@xxxxxxxxxxxxx
- Date: Fri, 10 Sep 2004 16:52:25 -0400
Hi, welcome to hashcash-list :-)
Following Jonathan's suggestion of MINGW, I hacked on it enough to get
it to compile on there. And then some more to get the assembler on
(signal handler API used by assembler to detect non-MMX cpus not
supported in MINGW API set).
And it works! (That's got to be the fastest hack-job ever:-)
I just uploaded 1.06 which now for first time in a really long time
has win32 exes.
You can get them from the binaries->windows->console menu on
http://www.hashcash.org
or if impatient:
http://www.hashcash.org/binaries/win32/hashcash.exe
and pgp sig:
http://www.hashcash.org/binaries/win32/hashcash.exe.asc
And this enables me to test on some more machines so here's a
1.4GHz celeron M:
2535393 AMD64/x86 MMX Standard 1x2-pipe *
Solution: 1:22:040404:foo::0123456789abcdef:0000000JaED
Iterations: 5129101
Time taken: 2.023
(that's MMX assembler code) and fastest C code:
1493622 ANSI Compact 1-pipe
Solution: 1:22:040404:foo::0123456789abcdef:0000000JaED
Iterations: 5129101
Time taken: 3.434
maybe as very rough estimate .6/1.4x the celeron M based on what
Jonathan said.
So 2.5 megahashes/sec * .6/1.4 = about 1 megahash/sec
1megahash = 20 bits.
So for 27 bits that would be 128secs ~ about 2mins.
(All highly rough given uncertainty about PIII 600 vs celeron M 1.4Ghz
scaling etc.)
Adam
On Fri, Sep 10, 2004 at 04:06:28PM +0000, panta-admin wrote:
> >Hi Adam, all !!
>
> I run an anonymous remailer with a hashcash requirement for usenet postings.
>
> Adam, I am in dire need of a windows binary to include in my GUI.
> If minting is now much faster it needs to be updated in my minter.
>
> I would also need it to check how long it will now take to mint tokens for
> panta
> (now at 27 bits) on my PIII 600.
> Could you benchmark a few 27 bit tokens for me ?
>
> > Still no windows compile. Oh well. If someone who is good at that
> > wants to compile and send patches that would be most welcome.
> > (Probably generic would be easier -- already Jonathan's C code is
> > faster).
>
> Please if someone could do that,
Other related posts:
