ietf-asrg
[Top] [All Lists]

Re: [Asrg] Re: "HashStamp" == hashcash? (Re: Stamping)

2003-03-24 12:10:26
On Sat, 22 Mar 2003 16:43:17 MST, Vernon Schryver 
<vjs(_at_)calcite(_dot_)rhyolite(_dot_)com>  said:

What is the minimumm speed CPU that you will allow to send mail?  A
25 MHz 486 is at least 200 and perhaps 500 or 1000 times slower than
a 3 GHz P4.  A 25 MHz 486 needs several minutes to generate 60
GHz*seconds of hashcash.  You might fix the problem for mailing lists

Now, Linux BogoMips are called that because they're bogus for anything except
calibrating kernel spin loops, but anyhow, here's some numbers seen in the last
week or so on the linux-kernel list.  Note that ALL the boxes are "currently in
use"(*).

Somebody's linux-based firewall system:
model name      : 486
bogomips        : 12.44

My laptop:
model name      : Intel(R) Pentium(R) 4 Mobile CPU 1.60GHz
bogomips        : 3145.72

A dual 2.6GHz P4 Xeon (hardly a "large" system by any means) (yes, 4CPU due
to Intel's Hyperthreading stuff - actual performance is likely 60% of this):
Total of 4 processors activated (20696.26 BogoMIPS).

The current high-end Linux code is being extended to support 64-CPU well, I
don't have any numbers handy for the high-end NUMA/Summit systems....

So I have to agree that "in the field" there's over a 1,000X factor in
relative performance...

/Valdis

(*) There's a report of a 4.77mz 8086 clocking in at 0.02 BogoMIPS, but
I don't classify this box as "current in use".

Attachment: pgpIrDmVIDcIU.pgp
Description: PGP signature