OpenSSL AES speed different with EVP and no AES-NI

Question

My computer does not have AES-NI, but when I run openssl speed with and without -evp I see a speed difference. Why is this?

openssl speed aes-128-cbc
Doing aes-128 cbc for 3s on 16 size blocks: 18132001 aes-128 cbc's in 3.00s
Doing aes-128 cbc for 3s on 64 size blocks: 4890318 aes-128 cbc's in 3.00s
Doing aes-128 cbc for 3s on 256 size blocks: 1246069 aes-128 cbc's in 3.00s
Doing aes-128 cbc for 3s on 1024 size blocks: 311859 aes-128 cbc's in 3.00s
Doing aes-128 cbc for 3s on 8192 size blocks: 38970 aes-128 cbc's in 3.00s
OpenSSL 1.0.1f 6 Jan 2014
built on: Thu Jun 11 15:30:15 UTC 2015
options:bn(64,64) rc4(16x,int) des(idx,cisc,16,int) aes(partial) blowfish(idx) 
compiler: cc -fPIC -DOPENSSL_PIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -m64 -DL_ENDIAN -DTERMIO -g -O2 -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -Wl,-Bsymbolic-functions -Wl,-z,relro -Wa,--noexecstack -Wall -DMD32_REG_T=int -DOPENSSL_IA32_SSE2 -DOPENSSL_BN_ASM_MONT -DOPENSSL_BN_ASM_MONT5 -DOPENSSL_BN_ASM_GF2m -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DMD5_ASM -DAES_ASM -DVPAES_ASM -DBSAES_ASM -DWHIRLPOOL_ASM -DGHASH_ASM
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
aes-128 cbc      96704.01k   104326.78k   106331.22k   106447.87k   106414.08k

And

openssl speed -evp aes-128-cbc
Doing aes-128-cbc for 3s on 16 size blocks: 90633962 aes-128-cbc's in 2.99s
Doing aes-128-cbc for 3s on 64 size blocks: 24880226 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 256 size blocks: 7574778 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 1024 size blocks: 1962765 aes-128-cbc's in 3.00s
Doing aes-128-cbc for 3s on 8192 size blocks: 247920 aes-128-cbc's in 2.99s
OpenSSL 1.0.1f 6 Jan 2014
built on: Thu Jun 11 15:28:12 UTC 2015
options:bn(64,64) rc4(8x,int) des(idx,cisc,16,int) aes(partial) blowfish(idx) 
compiler: cc -fPIC -DOPENSSL_PIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -m64 -DL_ENDIAN -DTERMIO -g -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -Wl,-Bsymbolic-functions -Wl,-z,relro -Wa,--noexecstack -Wall -DMD32_REG_T=int -DOPENSSL_IA32_SSE2 -DOPENSSL_BN_ASM_MONT -DOPENSSL_BN_ASM_MONT5 -DOPENSSL_BN_ASM_GF2m -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DMD5_ASM -DAES_ASM -DVPAES_ASM -DBSAES_ASM -DWHIRLPOOL_ASM -DGHASH_ASM
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
aes-128-cbc     484997.79k   530778.15k   646381.06k   669957.12k   679251.05k

Answer 1

evp doesn't actually control only AES_NI but rather a whole slew of CPU intrinsics and optimizations as it uses an entirely different implementation of the AES algorithm that attempts to take advantage of the hardware and compiler as much as possible. This answer from security.stackexchange explains it far better than I can.