# Strassen Algorithm for Dense

**URL:** <https://discuss.tvm.apache.org/t/strassen-algorithm-for-dense/2661>\
**Category:** Development\
**Created:** [May 21, 2019, 7:53am UTC](https://discuss.tvm.apache.org/t/strassen-algorithm-for-dense/2661 "2019-05-21T07:53:55Z")\
**Posts on this page:** 1\
**Showing post:** 15

<div class="post-metadata">

**Author:** ![githubzj](https://discuss.tvm.apache.org/letter_avatar_proxy/v4/letter/g/e9c0ed/32.png) [@githubzj](https://discuss.tvm.apache.org/u/githubzj)\
**Post date:** [September 21, 2020, 1:35pm UTC](https://discuss.tvm.apache.org/t/strassen-algorithm-for-dense/2661/15 "2020-09-21T13:35:20Z")

</div>

Thank you for your reply.

Regarding time-consuming fluctuations, I didn’t make it clear. After autotvm tune is completed, I picked the best record for time-consuming testing, and its time-consuming fluctuates significantly.I calculate the time difference between the start and the end to get the time-consuming.

> struct timeval curTime1;
> 
> gettimeofday(&curTime1, NULL);
> 
> size\_t milli\_start = curTime1.tv\_sec\*1000000 + curTime1.tv\_usec;
> 
> tvm::runtime::TVMRetValue ret = f(x, y, z);
> 
> struct timeval curTime2;
> 
> gettimeofday(&curTime2, NULL);
> 
> size\_t milli\_end = curTime2.tv\_sec\*1000000 + curTime2.tv\_usec;
> 
> size\_t run\_time = milli\_end - milli\_start;

However, the time-consuming of strassen algorithm does not fluctuate significantly. So I am curious whether time-consuming fluctuation is related to tvm, or it is just caused by cpu load changes(After all, cpu is not dedicated).

---

_[View the full topic](https://discuss.tvm.apache.org/t/strassen-algorithm-for-dense/2661)._