I created an g4dn.xlarge instance on AWS,
# /hashcat/hashcat.bin -II
hashcat (v6.2.6) starting in backend information mode
CUDA.Version.: 12.2
Backend Device ID #1 (Alias: #2)
Name...........: Tesla T4
Processor(s)...: 40
Clock..........: 1590
Memory.Total...: 15102 MB
Memory.Free....: 14999 MB
Local.Memory...: 64 KB
PCI.Addr.BDFe..: 0000:00:03.6
OpenCL Info:
============
OpenCL Platform ID #1
Vendor..: NVIDIA Corporation
Name....: NVIDIA CUDA
Version.: OpenCL 3.0 CUDA 12.2.138
Backend Device ID #2 (Alias: #1)
Type...........: GPU
Vendor.ID......: 32
Vendor.........: NVIDIA Corporation
Name...........: Tesla T4
Version........: OpenCL 3.0 CUDA
Processor(s)...: 40
Clock..........: 1590
Memory.Total...: 15102 MB (limited to 3775 MB allocatable in one block)
Memory.Free....: 14976 MB
Local.Memory...: 48 KB
OpenCL.Version.: OpenCL C 1.2
Driver.Version.: 535.104.12
PCI.Addr.BDF...: 00:03.6
The benchmark result shows hashcat can test 136K hashes per second:
# /hashcat/hashcat.bin -m 1800 -b --optimized-kernel-enable
hashcat (v6.2.6) starting in benchmark mode
Benchmarking uses hand-optimized kernel code by default.
You can use it in your cracking session by setting the -O option.
Note: Using optimized kernel code limits the maximum supported password length.
To disable the optimized kernel code in benchmark mode, use the -w option.
nvmlDeviceGetFanSpeed(): Not Supported
CUDA API (CUDA 12.2)
====================
* Device #1: Tesla T4, 14999/15102 MB, 40MCU
OpenCL API (OpenCL 3.0 CUDA 12.2.138) - Platform #1 [NVIDIA Corporation]
========================================================================
* Device #2: Tesla T4, skipped
Benchmark relevant options:
===========================
* --optimized-kernel-enable
--------------------------------------------------------------------
* Hash-Mode 1800 (sha512crypt $6$, SHA512 (Unix)) [Iterations: 5000]
--------------------------------------------------------------------
Speed.#1.........: 136.0 kH/s (47.12ms) @ Accel:2048 Loops:512 Thr:32 Vec:1
Started: Thu Nov 16 01:17:06 2023
Stopped: Thu Nov 16 01:17:17 2023
However, when I load 8 hashes to hashcat, the performance is only 46029 hashes per second,
# /hashcat/hashcat.bin \
--status --status-timer 3 \
-o cracked.txt \
-m 1800 \
-a 3 \
-1 ?l?u?d \
hashes.txt ?1?1?1?1?1?1?1?1
...snip...
Session..........: hashcat
Status...........: Running
Hash.Mode........: 1800 (sha512crypt $6$, SHA512 (Unix))
Hash.Target......: hashes.txt
Time.Started.....: Thu Nov 16 01:47:20 2023 (35 secs)
Time.Estimated...: Tue Jun 2 13:34:21 3226 (1202 years, 199 days)
Kernel.Feature...: Pure Kernel
Guess.Mask.......: ?1?1?1?1?1?1?1?1 [8]
Guess.Charset....: -1 ?l?u?d, -2 Undefined, -3 Undefined, -4 Undefined
Guess.Queue......: 1/1 (100.00%)
Speed.#1.........: 46029 H/s (8.87ms) @ Accel:256 Loops:256 Thr:32 Vec:1
Recovered........: 0/8 (0.00%) Digests (total), 0/8 (0.00%) Digests (new), 0/8 (0.00%) Salts
Progress.........: 1630208/1746720844679168 (0.00%)
Rejected.........: 0/1630208 (0.00%)
Restore.Point....: 0/3521614606208 (0.00%)
Restore.Sub.#1...: Salt:3 Amplifier:13-14 Iteration:2048-2304
Candidate.Engine.: Device Generator
Candidates.#1....: karierin -> khshanan
Hardware.Mon.#1..: Temp: 46c Util: 99% Core:1305MHz Mem:5000MHz Bus:8
Why is that?