phc-discussions - Re: [PHC] "Why I Don't Recommend Scrypt"

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20140313181538.GA31638@openwall.com>
Date: Thu, 13 Mar 2014 22:15:38 +0400
From: Solar Designer <solar@...nwall.com>
To: discussions@...sword-hashing.net
Subject: Re: [PHC] "Why I Don't Recommend Scrypt"

On Thu, Mar 13, 2014 at 05:14:54PM +0000, Brian Matthews (brmatthe) wrote:
> A service I help run has had peak authentication rates of 1200/sec, so in the worst case that's 1.2TB of RAM. Maybe the servers are extremely RAM-limited, but they don't have 1.2TB each. :-)

I wish this were the problem, but it's actually worse than that: you
can't fill this much memory quickly enough.

We can't even fill 16 MB (the minimum specified in scrypt paper) quickly
enough to meet the request rate capacity stipulated for a deployment.

That's why the "ROM" for use cases like that.

Alexander