Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 31 Aug 2020 20:01:22 -0400
From: Rich Rumble <>
Subject: Re: SIMD performance impact

On Mon, Aug 31, 2020 at 5:09 PM Vincent <> wrote:

> Hi,
> John can use instruction set specific optimizations to fully exploit
> technology like SIMD. I haven't got CPUs that support AVX-512 but I'm
> very interested in the possible performance gains. So my question is:
> can someone with the latest generation CPU run a 'john --test' with
> different instruction set binaries (for example SSE4.2, AVX, AVX2,
> AVX-512) on the same CPU?
You may find some of what you're after on the WIKI, not only in terms of
instructions optimizations, but threading as well as parallel/workload
splitting methods like MPI, HT and thread count. You could certainly force
JtR to build and favor those instructions, but I think by default it tries
to optimize on what is detected with ./configure and when built too I
believe it will try to figure out what is present.
I remember when JtR started using MMX and then SSE (can't remember which
came first) and I thought it was genius! With AMD (Ryzen etc...) I'm
curious myself to see how the threadrippers do vs Intel ;)

> I think it might also be interesting to match current Intel (Core i9 &
> Xeon) and AMD (Ryzen & Epyc) processors to see what the impact of
> implementation details is (like Xeon AVX-512 versus Epyc AVX2).
> Perhaps interesting to upgrade
> <> with all algorithms (since
> some do not seem to be SIMD-accelerated) and with results based on
> different instruction sets on the same CPU?
> Thanks!
> Cheers, V

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.