vanityhash has become an accidental benchmark.
git clone https://github.com/rfinnie/vanityhash
vanityhash/vanityhash benchmark
- Benchmark mode takes 60 seconds to complete.
- Unless mentioned for fo0bar's machines: amd64, Ubuntu 18.04. Others' are whatever.
- This is nowhere near a scientific comparison. For example, Python 3.6 is about 25% faster than 3.5 with vanityhash, simply due to Python optimizations.
| Host | Workers | Hashrate | Arch | Person | Py3.6+ | System |
|---|---|---|---|---|---|---|
| lubbock | 128 | 48.53 Mhash/s | amd64 | fo0bar | ✗ | Dual-socket EPYC 7501 (borrowed) |
| c5.18xlarge | 72 | 34.17 Mhash/s | amd64 | fo0bar | ✗ | EC2 guest on dual-socket Xeon Platinum 8124M, 2.49c/Ghash |
| fairouz | 64 | 29.93 Mhash/s | amd64 | fo0bar | ✗ | Dual-socket EPYC 7351 (borrowed) |
| nibbler | 16 | 10.51 Mhash/s | amd64 | fo0bar | ✗ | Ryzen 2700X |
| covfefe | 12 | 9.29 Mhash/s | amd64 | fuzzie | i7-8700K | |
| feh | 16 | 5.95 Mhash/s | amd64 | fo0bar | ✗ | E5-2620 v4 (SuperMicro 5018R-M) |
| hex | 8 | 5.84 Mhash/s | amd64 | fuzzie | i7-7700K | |
| chmmr | 8 | 5.06 Mhash/s | amd64 | Octal | i7-6700 | |
| linda (new) | 8 | 4.20 Mhash/s | amd64 | fo0bar | ✓ | i7-8565U (ThinkPad T490), Ubuntu 19.04 |
| chroot-dev | 4 | 4.03 Mhash/s | amd64 | fo0bar | ✗ | KVM guest on 2700X |
| a1.4xlarge | 16 | 3.74 Mhash/s | arm64 | fo0bar | ✗ | EC2 guest on AWS Graviton, 3.03c/Ghash |
| ndnd | 6 | 2.52 Mhash/s | amd64 | fo0bar | ✓ | i5-8500B (2018 Mac Mini), MacOS Mojave |
| laptop | 4 | 1.80 Mhash/s | amd64 | teferi | i7-6600U | |
| linda | 4 | 1.52 Mhash/s | amd64 | fo0bar | ✗ | i5-5300U (ThinkPad X250) |
| rpi4 | 4 | 610.64 Khash/s | armhf | fo0bar | ✓ | Raspberry Pi 4, Raspbian buster |
| freebsd-dev | 1 | 550.24 Khash/s | amd64 | fo0bar | KVM guest on 2700X, FreeBSD 12.0 | |
| intelcs | 4 | 411.69 Khash/s | amd64 | fo0bar | ✓ | Atom Z3735F (Intel Compute Stick), Ubuntu 19.04 |
| zoidberg | 4 | 321.74 Khash/s | armhf | fo0bar | ✓ | Raspberry Pi 3 B+, Raspbian buster |
| a1.medium | 1 | 264.21 Khash/s | arm64 | fo0bar | ✗ | EC2 guest on AWS Graviton, 2.68c/Ghash |
| michelle | 4 | 211.52 Khash/s | arm64 | fo0bar | ✗ | Allwinner H5 (ALL-H3-CC H5) |
| cluster⁰ | 4 | 159.31 Khash/s | armhf | fo0bar | ✓ | Raspberry Pi 2, Raspbian buster |
| inez | 1 | 89.11 Khash/s | ppc | fo0bar | ✗ | G4 1.4GHz (2005 Mac Mini), Ubuntu 16.04 |
| s390x-dev | 1 | 32.61 Khash/s | s390x | fo0bar | ✗ | Qemu s390x guest on 2700X |
| ppc64el-dev | 1 | 19.33 Khash/s | ppc64el | fo0bar | ✗ | Qemu ppc64el guest on 2700X |
| amdk6 | 1 | 9.25 Khash/s | i386 | fo0bar | ✗ | 300MHz AMD K6 586, Debian jessie |
| rpi1 | 1 | 6.31 Khash/s | armhf | fo0bar | ✗ | Rasbperry Pi Zero W, Raspbian stretch |
| cobalt | 1 | 2.10 Khash/s | mipsel | fo0bar | ✗ | RM5231 (Cobalt RaQ2), Debian wheezy |
- ⁰ This is the coordinator machine for a cluster, not the full performance of the cluster itself.