blob: c10f357dc770ab43f4cbe7edb10d9afc6b23e6e9 (
plain)
1
2
3
4
5
6
7
8
9
10
11
|
# Benchmarks
The main idea is to run the existing HLS benchmarks to see if they work - they can only test correctness for a single set of input by a customised test bench.
- jacob_2d: a benchmark from Polybench
- sobel: a benchmark from LegUp HLS
- getTanh: a benchmark from DASS
- fft: a benchmark from MachSuite
- KMeans: a benchmark from Felix's work
Note all the benchmark set above (links included in the source code) should be all synthesisable, so you may be able to show coverage instead of a single benchmark.
|
