john-dev - AVX bitsliced sha256

Date: Mon, 22 Jun 2015 17:09:06 -0400
From: Alain Espinosa <alainesp@...ta.cu>
To: john-dev@...ts.openwall.com
Subject: AVX bitsliced sha256

Hi.

Just test bitsliced sha256 with floating point AVX boolean (256 bits registers). In a core i3-2120 is 20% slower than bitsliced SSE2 (128 bits registers with integer boolean). Giving a look at instructions throughput SSE2 is 50% faster per bit than AVX (3 instructions each cycle for SSE2 vs only 1 for AVX).

Regards,
Alain

Content of type "text/html" skipped

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