roycewilliams

------------------------------------------------------------------------------
You can drop the power on NVIDIA cards and improve performance per watt.

Credit for the idea to atom on the 1080s:

https://docs.google.com/spreadsheets/d/1B1S_t1Z0KsqByH3pNkYUM-RCFMu860nlfSsYEqOoqco/edit#gid=1307275979

On the GTX 970s, you can drop to 150 and still get 99.96% of the same 
performance! This saves power (and heat?). And there is evidence that the
performance difference varies per hash and per attack (masks, rules, etc.)

roycewilliams

hashcat (v3.40-93-g6c98fac1) starting in benchmark mode...

OpenCL Platform #1: NVIDIA Corporation
======================================
* Device #1: GeForce GTX 1080, 2028/8113 MB allocatable, 20MCU
* Device #2: GeForce GTX 1080, 2028/8114 MB allocatable, 20MCU
* Device #3: GeForce GTX 970, 1009/4037 MB allocatable, 13MCU
* Device #4: GeForce GTX 1080, 2028/8114 MB allocatable, 20MCU
* Device #5: GeForce GTX 970, 1009/4037 MB allocatable, 13MCU
* Device #6: GeForce GTX 970, 1009/4037 MB allocatable, 13MCU

roycewilliams

# Devices: 
# - 3x NVIDIA 1080 Founders Edition, no overclock
# - 3x EVGA 04G-P4-2974-KR, factory overclocked but otherwise unmodified
# NVIDIA driver: 375.39
# Hashes are in order from slowest (better for the defender) to fastest (better for the attacker)

hashcat (v3.40-93-g6c98fac1) starting in benchmark mode...

OpenCL Platform #1: NVIDIA Corporation
======================================

roycewilliams

# Standard hashcat benchmark.
#
# As a supplement to this benchmark, there is also a one-record-per-hash summary, 
# including total aggregate speed, sorted in order from slowest to fastest hash:
#    https://gist.github.com/roycewilliams/844be514d6286fa963b0c9d88e8bb739

hashcat (v3.5.0) starting in benchmark mode...

OpenCL Platform #1: NVIDIA Corporation
======================================

roycewilliams

# This is a summary of total aggregate speed for the raw hashcat benchmarks here:
# https://gist.github.com/roycewilliams/0d6a837a5c47de36f9aa9c7953384017
#
# Devices: 
# - 3x NVIDIA 1080 Founders Edition, no overclock
# - 3x EVGA GTX 970 SC (04G-P4-2974-KR), factory overclocked but otherwise unmodified
#
# hashcat v3.5.0
# NVIDIA driver: 375.39
#

roycewilliams

# hashcat benchmark for irongiant - 2017-04-06
#
# Devices: 
# - 4x NVIDIA 1080 Founders Edition, no overclock
# - 2x EVGA GTX 970 SC (04G-P4-2974-KR), factory overclocked but otherwise unmodified
#
# hashcat v3.5.0
# NVIDIA driver: 375.39
#
# Hashes are in order from slowest (better for the defender) to fastest (better for the attacker)

roycewilliams

CAA Adoption Notes

CA/B mandate

The CA/Browser Forum has voted to mandate that CAs must start checking CAA records to control certificate issuance (to verify what CAs are allowed to issue certificates for a domain) by September 8th, 2017.

https://blog.qualys.com/ssllabs/2017/03/13/caa-mandated-by-cabrowser-forum

Full language of the ballot is here.

roycewilliams

# Multiple
    https://github.com/magnumripper/JohnTheRipper/tree/bleeding-jumbo/run

# itunes backups
    https://github.com/philsmd/itunes_backup2hashcat

# in-memory axsuite
    https://github.com/Fist0urs/AxSuite

# netntlmv1 - still needed?

roycewilliams

# hashcat benchmark for irongiant - 2017-05-02
#
# Devices: 
# - 6x NVIDIA 1080 Founders Edition, no overclock
#
# hashcat v3.5.0
# NVIDIA driver: 375.39
#
# Aggregate benchmark: https://gist.github.com/roycewilliams/cd5c4fcc482785f98146a154115ab217

roycewilliams

# hashcat benchmark for irongiant - 2017-05-02
#
# Devices: 
# - 6x NVIDIA 1080 Founders Edition, no overclock
#
# hashcat v3.5.0
# NVIDIA driver: 375.39
#
# Per-card benchmark: https://gist.github.com/roycewilliams/a4375257af749b33ccfb681b3d868d99
#