Implementation notes: x86, gcc16, crypto_dh

Computer: gcc16
Architecture: x86
CPU ID: AuthenticAMD-00100f23-178bfbff
CPU cycles/second: 2194490999...2194496999 (x86cpuinfo)
SUPERCOP version: 20120310
Benchmark dates: 20100918...20120311

crypto_dh

Time	Relative time	Primitive	Implementation	Compiler
32472798	1.00	`claus`	`crypto_dh/claus/gmp`	`g++ -m32 -march=nocona -Os -fomit-frame-pointer (4.3.2)`
55309057	1.70	`claus`	`crypto_dh/claus/openssl`	`gcc -funroll-loops -m32 -march=i386 -O -fomit-frame-pointer`
		`claus`	`crypto_dh/claus/cryptopp`	`g++ -m32 -O3 -fomit-frame-pointer`
		`claus`	`crypto_dh/claus/ntl`	`g++ -m32 -O3 -fomit-frame-pointer`
6447342	1.00	`curve2251`	`crypto_dh/curve2251/mpfq`	`gcc -funroll-loops -m32 -march=pentium-m -O -fomit-frame-pointer (4.3.2)`
2317286	1.00	`curve25519`	`crypto_dh/curve25519/ref`	`gcc -m32 -march=native -mtune=native -O3 -fomit-frame-pointer (4.3.2)`
4419764	1.91	`curve25519`	`crypto_dh/curve25519/mpfq`	`gcc -funroll-loops -m32 -march=pentium-mmx -O3 -fomit-frame-pointer`
		`ecfp256e`	`crypto_dh/ecfp256e/v01/var`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256e`	`crypto_dh/ecfp256e/v01/w8s1`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256e`	`crypto_dh/ecfp256e/v01/w8s2`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256e`	`crypto_dh/ecfp256e/v01/w8s4`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256e`	`crypto_dh/ecfp256e/v01/w8s8`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256h`	`crypto_dh/ecfp256h/v01/var`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256h`	`crypto_dh/ecfp256h/v01/w8s1`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256h`	`crypto_dh/ecfp256h/v01/w8s2`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256h`	`crypto_dh/ecfp256h/v01/w8s4`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256h`	`crypto_dh/ecfp256h/v01/w8s8`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256i`	`crypto_dh/ecfp256i/v01/var`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256i`	`crypto_dh/ecfp256i/v01/w8s1`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256i`	`crypto_dh/ecfp256i/v01/w8s2`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256i`	`crypto_dh/ecfp256i/v01/w8s4`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256i`	`crypto_dh/ecfp256i/v01/w8s8`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256q`	`crypto_dh/ecfp256q/v01/var`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256q`	`crypto_dh/ecfp256q/v01/w8s1`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256q`	`crypto_dh/ecfp256q/v01/w8s2`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256q`	`crypto_dh/ecfp256q/v01/w8s4`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256q`	`crypto_dh/ecfp256q/v01/w8s8`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256s`	`crypto_dh/ecfp256s/v01/var`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256s`	`crypto_dh/ecfp256s/v01/w8s1`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256s`	`crypto_dh/ecfp256s/v01/w8s2`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256s`	`crypto_dh/ecfp256s/v01/w8s4`	`gcc -m32 -O3 -fomit-frame-pointer`
		`ecfp256s`	`crypto_dh/ecfp256s/v01/w8s8`	`gcc -m32 -O3 -fomit-frame-pointer`
		`gls1271`	`crypto_dh/gls1271/ref4`	`gcc -m32 -O3 -fomit-frame-pointer`
4470814	1.00	`hector`	`crypto_dh/hector/ref`	`gcc -funroll-loops -m32 -march=barcelona -O -fomit-frame-pointer (4.3.2)`
		`nistp256`	`crypto_dh/nistp256/nawaz`	`gcc -m32 -O3 -fomit-frame-pointer`
5415057	1.00	`sclaus1024`	`crypto_dh/sclaus1024/gmp`	`g++ -m32 -march=core2 -msse4.1 -O -fomit-frame-pointer (4.3.2)`
		`sclaus1024`	`crypto_dh/sclaus1024/cryptopp`	`g++ -m32 -O3 -fomit-frame-pointer`
27365517	1.00	`sclaus2048`	`crypto_dh/sclaus2048/gmp`	`g++ -m32 -march=k8 -O2 -fomit-frame-pointer (4.3.2)`
		`sclaus2048`	`crypto_dh/sclaus2048/cryptopp`	`g++ -m32 -O3 -fomit-frame-pointer`
4987434	1.00	`surf127eps`	`crypto_dh/surf127eps/mpfq`	`gcc -funroll-loops -m32 -march=athlon -O3 -fomit-frame-pointer (4.3.2)`
4798404	1.00	`surf2113`	`crypto_dh/surf2113/mpfq`	`gcc -funroll-loops -m32 -march=k8 -O -fomit-frame-pointer (4.3.2)`

crypto_scalarmult

Time	Relative time	Primitive	Implementation	Compiler
1149814	1.00	`curve25519`	`crypto_scalarmult/curve25519/athlon`	`gcc -m32 -march=core2 -msse4 -O3 -fomit-frame-pointer (4.3.2)`
4760429	4.14	`curve25519`	`crypto_scalarmult/curve25519/ref10`	`gcc -funroll-loops -m32 -march=nocona -O -fomit-frame-pointer`
20981787	18.25	`curve25519`	`crypto_scalarmult/curve25519/ref`	`gcc -funroll-loops -m32 -march=pentium -Os -fomit-frame-pointer`
		`curve25519`	`crypto_scalarmult/curve25519/amd64-51`	`gcc -m32 -O3 -fomit-frame-pointer`
		`curve25519`	`crypto_scalarmult/curve25519/amd64-64`	`gcc -m32 -O3 -fomit-frame-pointer`
		`curve25519`	`crypto_scalarmult/curve25519/costigan-schwabe/cbe`	`gcc -m32 -O3 -fomit-frame-pointer`
		`curve25519`	`crypto_scalarmult/curve25519/donna`	`gcc -m32 -O3 -fomit-frame-pointer`
		`curve25519`	`crypto_scalarmult/curve25519/donna_c64`	`gcc -m32 -O3 -fomit-frame-pointer`