Skip to content

Commit 736c212

Browse files
authored
Update README.md
1 parent ab6405a commit 736c212

File tree

1 file changed

+5
-5
lines changed

1 file changed

+5
-5
lines changed

README.md

Lines changed: 5 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -9,7 +9,7 @@
99
PhastFT is a high-performance, "quantum-inspired" Fast Fourier
1010
Transform (FFT) library written in safe Rust.
1111

12-
Designed for large FFTs, such as in a quantum computer simulator.
12+
Designed for large FFTs common in scientific workloads, e.g. in quantum computer simulators.
1313

1414
## Features
1515

@@ -29,12 +29,12 @@ Designed for large FFTs, such as in a quantum computer simulator.
2929
## Planned features
3030

3131
- Real-to-complex FFT
32-
- More work on cache-optimal FFT
32+
- Even more work on performance
3333

3434
## How is it so fast?
3535

36-
PhastFT is designed around the capabilities and limitations of modern hardware (that is, anything made in the last 15
37-
years or so).
36+
PhastFT is designed around the capabilities and limitations of modern hardware
37+
(that is, anything made in the last 10 years or so).
3838

3939
The two major bottlenecks in FFT are the **CPU cycles** and **memory accesses**.
4040

@@ -44,7 +44,7 @@ It doesn't matter how much or how little CPU instructions you need to execute
4444
if the CPU spends most of the time just waiting on memory anyway!
4545

4646
[Notes on FFTs for implementers](https://fgiesen.wordpress.com/2023/03/19/notes-on-ffts-for-implementers/) is a good read
47-
if you want to understand the trade-offs on modern hardware. The author is not affiliated with PhastFT.
47+
if you want to understand the trade-offs on modern hardware. Its author is not affiliated with PhastFT.
4848

4949
The trade-offs we chose are:
5050

0 commit comments

Comments
 (0)