The wolfSSL embedded TLS library and its crypto engine wolfCrypt are both highly configurable to give users the best cryptographic performance. Our users appreciate that they’re able to customize their builds to suit their specific needs. If your needs are a crypto or SSL/TLS solution optimized for performance, we’ve compiled a list of the top 5 wolfSSL build options that you should consider.
Intel Assembly
For use with Intel/AMD processors, enabling the intelasm option for wolfSSL will utilize enhanced assembly instructions of the processor that can dramatically enhance cryptographic performance for most algorithms. The instruction sets leveraged when the configure option is enabled include AVX1, AVX2, BMI2, RDRAND, RDSEED, AESNI, and ADX. This option also automatically enables our Intel AES-NI support, which on its own can lead to direct AES encryption that’s over 3.3 times faster than using software-based AES. Performance increases in total can be as large as 5,800% when intelasm is used.
Enabled with: ./configure –enable-intelasm
Macro: WOLFSSL_ARMASM
ARM Assembly
Whether you’re doing cryptography on an ARMv8, ARMv7 or even an ARM64, armasm is the quickest way to speedup your cryptographic operations, and by doing so speedup your TLS that makes use of these algorithms. By using the cryptographic instructions built into the chips, we get a significant boost in performance over straight C. We recently did another round of tuning on our ARM64 code and got the crypto running up to 9.5 times faster than it already was.
Enabled with: ./configure –enable-armasm
Macro: WOLFSSL_ARMASM
RISC-V Assembly
The newest addition on this list, wolfSSL now has support for RISC-V hardware acceleration. The RISC-V assembly implementation of AES includes standard/scalar cryptography/vector cryptographic instructions for ECB/CBC/CTR/GCM/CCM and can be up to 50 times faster than the software implementation.
Enabled with: ./configure –enable-riscv-asm
Macro: WOLFSSL_RISCV_ASM
Single Precision Math Assembly
Unlike some of the other build options on this list, SP ASM is hardware agnostic. SP is Single Precision Math and it is a wolfSSL developed math library that is extremely well optimized for cryptographic math calculations. SP ASM is the assembly component of the SP math library. Enabling this option leads to faster RSA, ECC and DH operations.
Enabled with: ./configure –enable-sp –enable-sp-asm
Macros: WOLFSSL_SP_ASM (along with WOLFSSL_SP_HAVE_RSA/ECC/DH)
AES GCM (4-bit Table)
A hardware agnostic feature for AES GCM (authenticated cipher) to improve the symmetric performance of the GHASH authentication calculation. The performance gain is about a 3x speedup.
Enabled with: ./configure –enable-asmgcm=4bit
Macros: WOLFSSL_SP_ASM (along with WOLFSSL_SP_HAVE_RSA/ECC/DH)
For more information on the wolfSSL library and its performance, please visit our benchmarks page or contact us at facts@wolfSSL.com or +1 425 245 8247.
Download wolfSSL Now
Questions? Contact Us- Support Forums
