SIMD-accelerated compute kernel benchmarks for the Kazkade zero-copy columnar analytics engine. The dataset includes AVX2 vs SSE4.2 vs scalar GEMM performance, vector operation throughput across 1M elements, columnar scan predicate filtering, and multi-layer perceptron inference timing. All measurements were conducted on an Intel i7-1260P processor by Alpasan, Lois-Kleinner of The Anticloud by Lois-Kleinner and last updated on June 24, 2026.
Use Cases
- Benchmarking GEMM (General Matrix Multiply) performance across different SIMD instruction sets based on the AVX2 vs SSE4.2 vs scalar comparisons mentioned.
- Evaluating vector operation throughput for large-scale data processing based on the 1M element vector operation tests described.
- Profiling columnar database scan performance based on the predicate filtering benchmarks included.
- Timing neural network inference for multi-layer perceptrons based on the MLP inference timing measurements.
Strengths
- Benchmarks include specific hardware context: all measurements were taken on an Intel i7-1260P processor.
- Performance comparisons are provided across three distinct instruction set architectures: AVX2, SSE4.2, and scalar.
- Vector operation tests are conducted at a defined scale of 1 million elements.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- The dataset's focus is on a single hardware platform (Intel i7-1260P), limiting generalizability to other architectures.
Provenance
- Source
- The Anticloud by Lois-Kleinner
- Collection Method
- Benchmark measurements of compute kernels.
- Freshness
- Last updated 2026-06-24 15:29:44; freshness should be verified.