Boost.SIMD

Esterie, Pierre; Falcou, Joel; Gaunard, Mathias; Lapresté, Jean-Thierry

doi:10.1145/2568058.2568063

Cited by 18 publications

(7 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Having to design and support functions for several different vectorized instruction sets is cumbersome. We could ask whether we can achieve similar results with higher-level C++ libraries such as the Generic SIMD Library [Wang et al 2014] or Boost.SIMD [Estérie et al 2014]. Programming languages like Swift or Rust also have generic SIMD packages that could be put to good use.…”

Section: Resultsmentioning

confidence: 99%

Faster Base64 Encoding and Decoding Using AVX2 Instructions

Muła

Lemire

2018

ACM Trans. Web

View full text Add to dashboard Cite

Web developers use base64 formats to include images, fonts, sounds and other resources directly inside HTML, JavaScript, JSON and XML files. We estimate that billions of base64 messages are decoded every day. We are motivated to improve the efficiency of base64 encoding and decoding. Compared to state-of-the-art implementations, we multiply the speeds of both the encoding (≈ 10×) and the decoding (≈ 7×). We achieve these good results by using the single-instruction-multiple-data (SIMD) instructions available on recent Intel processors (AVX2). Our accelerated software abides by the specification and reports errors when encountering characters outside of the base64 set. It is available online as free software under a liberal license.

show abstract

Section: Resultsmentioning

confidence: 99%

Faster Base64 Encoding and Decoding Using AVX2 Instructions

Muła

Lemire

2018

ACM Trans. Web

View full text Add to dashboard Cite

show abstract

“…Dynamic programs are implemented using a generic programming paradigm employing expression templates and compile-time polymorphism; generic recursion equations capturing the details of the structural ensemble and free energy model are translated via template metaprogramming into a separate vectorized executable for calculating each physical quantity in Table . Single-threaded single instruction multiple data (SIMD) vectorization is implemented using the Boost.SIMD library . The convex optimization problem (eq 14) is solved in the dual form using an efficient trust region method using the Armadillo linear algebra library for matrix operations …”

Section: Methodsmentioning

confidence: 99%

A Unified Dynamic Programming Framework for the Analysis of Interacting Nucleic Acid Strands: Enhanced Models, Scalability, and Speed

2020

View full text Add to dashboard Cite

Dynamic programming algorithms within the NUPACK software suite enable analysis of nucleic acid sequences over complex and test tube ensembles containing arbitrary numbers of interacting strand species, serving the needs of researchers in molecular programming, nucleic acid nanotechnology, synthetic biology, and across the life sciences. Here, to enhance the underlying physical model, ensure scalability for large calculations, and achieve dramatic speedups when calculating diverse physical quantities over complex and test tube ensembles, we introduce a unified dynamic programming framework that combines three ingredients: (1) recursions that specify the dependencies between subproblems and incorporate the details of the structural ensemble and the free energy model, (2) evaluation algebras that define the mathematical form of each subproblem, (3) operation orders that specify the computational trajectory through the dependency graph of subproblems. The physical model is enhanced using new recursions that operate over the complex ensemble including coaxial and dangle stacking subensembles. The recursions are coded generically and then compiled with a quantity-specific evaluation algebra and operation order to generate an executable for each physical quantity: partition function, equilibrium base-pairing probabilities, MFE energy and proxy structure, suboptimal proxy structures, and Boltzmann sampled structures. For large complexes (e.g., 30 000 nt), scalability is achieved for partition function calculations using an overflow-safe evaluation algebra, and for equilibrium base-pairing probabilities using a backtrack-free operation order. A new blockwise operation order that treats subcomplex blocks for the complex species in a test tube ensemble enables dramatic speedups (e.g., 20–120× ) using vectorization and caching. With these performance enhancements, equilibrium analysis of substantial test tube ensembles can be performed in ≤ 1 min on a single computational core (e.g., partition function and equilibrium concentration for all complex species of up to six strands formed from two strand species of 300 nt each, or for all complex species of up to two strands formed from 80 strand species of 100 nt each). A new sampling algorithm simultaneously samples multiple structures from the complex ensemble to yield speedups of an order of magnitude or more as the number of structures increases above ≈103. These advances are available within the NUPACK 4.0 code base () which can be flexibly scripted using the all-new NUPACK Python module.

show abstract

“…SIMD libraries. A flurry of C++ libraries provide a unified programming model for various SIMD architectures, such as Boot.SIMD [11], MIPP [8], UME::SIMD [14], Sierra [18] or VC [16]. These works are complementary to ours: our intent is to provide a language that (always) compiles to efficient bitsliced code while being amenable to formal verification.…”

Section: Related Workmentioning

confidence: 99%

“…Historically, bitslicing is a manual process. For instance, here is a snippet of a bitsliced implementation of the Data Encryption Standard (DES) by Kwan [17]: [11], r1^k[26], r2^k [3], r3^k [13] r11^k [17], r12^k [5], &l23, &l15, &l29, &l5);…”

Section: Introductionmentioning

confidence: 99%

Usuba

Mercadier¹,

Dagand²,

Lacassagne³

et al. 2018

Proceedings of the 2018 4th Workshop on Programming Models for SIMD/Vector Processing

View full text Add to dashboard Cite

Bitslicing is a programming technique commonly used in cryptography that consists in implementing a combinational circuit in software. It results in a massively parallel program immune to cache-timing attacks by design. However, writing a program in bitsliced form requires extreme minutia. This paper introduces Usuba, a synchronous dataflow language producing bitsliced C code. Usuba is both a domain-specific language-providing syntactic support for the implementation of cryptographic algorithms-as well as a domain-specific compiler-taking advantage of welldefined semantics invariants to perform various optimizations before handing the generated code to an (optimizing) C compiler. On the Data Encryption Standard (DES) algorithm, we show that Usuba outperforms a reference, hand-tuned implementation by 15% (using Intel's 64 bits general-purpose registers and depending on the underlying C compiler) whilst our implementation also transparently supports modern SIMD extensions (SSE, AVX, AVX-512), other architectures (ARM Neon, IBM Altivec) as well as multicore processors through an OpenMP backend.

show abstract

Boost.SIMD

Cited by 18 publications

References 14 publications

Faster Base64 Encoding and Decoding Using AVX2 Instructions

Faster Base64 Encoding and Decoding Using AVX2 Instructions

A Unified Dynamic Programming Framework for the Analysis of Interacting Nucleic Acid Strands: Enhanced Models, Scalability, and Speed

Usuba

Contact Info

Product

Resources

About