ECHO hash function home design security hard soft compare

This page gives an overview of the performances of the ECHO hash function on various platforms (the figures given below are in cycles per bytes). Performances comparison with other SHA-3 candidates are also available here and here.

quick overview

cycles/Byte
ECHO-DP (double-pipe) ECHO-SP (simple-pipe)
256-bit hash 512-bit hash 256-bit hash 512-bit hash
Intel AES-NI 64-bit (Core i5 540M) 6.812.65.88.4
NIST platform 32-bit (Core 2 Duo) 32.559.726.740.7
NIST platform 64-bit (Core 2 Duo) 28.350.324.435.3

detailed software performance figures

Clicking on the platform name downloads the corresponding implementation. You can also get our full eBASH package. The performance figures in the table link to the details from the eBASH database, or from our machines when not available. Note that Cell and GPU implementations are not yet publicly available.

32-bit mode 64-bit mode
256-bit hash 512-bit hash 256-bit hash 512-bit hash
recent
Intel
Core i5 uses AES-NI 8.315.3 6.812.6
Core i7/i5 30.656.6 25.747.5
Core 2 32.559.7 28.350.3
recent
AMD
Phenom I/II 34.964.6 28.452.5
Opteron K8 32.359.0 28.551.7
Athlon K8 35.365.3 28.452.5
legacy
(old)
Intel/AMD
Athlon K7 XP 40.775.3
Athlon TBird 41.676.0
Pentium D ×Pentium 4E Prescott has specific issues, see Matsui's "How Far Can We Go on the x64 Processors?" 54.2×Pentium 4E Prescott has specific issues, see Matsui's "How Far Can We Go on the x64 Processors?"101.0×Pentium 4E Prescott has specific issues, see Matsui's "How Far Can We Go on the x64 Processors?" 49.3×Pentium 4E Prescott has specific issues, see Matsui's "How Far Can We Go on the x64 Processors?"98.7×Pentium 4E Prescott has specific issues, see Matsui's "How Far Can We Go on the x64 Processors?"
Pentium M 41.274.5
Pentium 4 44.282.4
Pentium III 47.787.2
Pentium II 47.888.9
Atom 84.1158.4
non-ia32 ARM unoptimized C code, improvements expected 147272
SPARC IIi non optimized assembly code, improvements expected 95.0175.8
PowerPC G3 optmized C code, improvements expected 66122.1
IBM Power4 optmized C code, improvements expected 59.0101.0
Itanium II unoptimized C code, improvements expected 33.960.9
multi-core Cell Broadband
(PlayStation 3)
29.61reported in "Performance Analysis of the SHA-3 Candidates on Exotic Multi-Core Architectures"
Nvidia GPU
(GTX 295)
0.851reported in "Performance Analysis of the SHA-3 Candidates on Exotic Multi-Core Architectures"
optimized C code, possible improvements with Altivec simd on G4/5 series     uses the AES-NI, see there
non optimized assembly code, possible improvements unoptimized C code
× non-optimal code, Pentium 4E Prescott has specific issues, see Matsui 1 see Bos and Stefan

compiling and running

Packages we provide are samples from the implementations submitted to eBASH. All implementations (C or assembly) embed the NIST API (i.e. the Init, Update, and Final function), and can thus be directly used with NIST API compliant software.