Re: PARASITIC COMPUTING - WHAT'S NEXT

From: aleph1@securityfocus.com
Date: 10/23/01 

Date: Mon, 22 Oct 2001 22:22:28 -0600
From: aleph1@securityfocus.com
To: "brian.carvalho@verizon.net" <brian.carvalho@verizon.net>
Subject: Re: PARASITIC COMPUTING - WHAT'S NEXT
Message-ID: <20011022222228.L19534@securityfocus.com>

* brian.carvalho@verizon.net (brian.carvalho@verizon.net) [011023 03:40]:
>
> Surely there is a more efficiant way of performing computations
> than to forge tons of TCP header info and deal with tons of
> unnecessary traffic and dropped packets due to failed
> "answers".

As they clearly point out in their paper this is simply a proof of
concept. They clearly state the TCP checksum implementation does not
provide a good ROI (the computation they can extract from remote CPUs
via this method is minimal compared to the cycles they spend generating,
sending, and receiving the packets). Nonetheless, there might be other
areas that can be exploited that have a better ROI making parasitic computing
payoff. In particular they say SSL/TSL may be an interesting area to explore.

> BC 

-- 
Elias Levy
SecurityFocus
http://www.securityfocus.com/
Si vis pacem, para bellum