October 2nd FPTalks Community Meeting Sehyeok Park on RLibm
56 views
Skip to first unread message
Ian Briggs
Sep 25, 2025, 11:34:23 AM (13 days ago) Sep 25
to fpb...@fpbench.org
Hi Everyone!
Next Thursday, October 2nd at 9:00-10:00 AM pacific time, we’ll have the next FPTalks Community Meeting on this Zoom:
https://washington.zoom.us/j/99708186928?pwd=HbhpebAtCWvoP4VQahYb8G1QpQnTgm.1
Google Calender: https://calendar.google.com/calendar/u/0?cid=Y181N3Ayc3ZqdjFuN2JxNDkzNGNlbmYzNWxkMEBncm91cC5jYWxlbmRhci5nb29nbGUuY29t
We’re super excited to welcome Sehyeok Park from Rutgers University to present on RLibm-MultiRound: Correctly Rounded Math Libraries without Worrying about the Application's Rounding Mode
Our RLibm project generates a single implementation for an elementary
function that produces correctly rounded results for multiple rounding
modes and representations with up to 32-bits. The key insight is to
build polynomials that produce the correctly rounded result for a
representation with two additional bits when compared to the largest
target representation and with the "non-standard" round-to-odd rounding
mode, which makes double rounding the RLibm math library result to any
smaller target representation innocuous. The resulting approximations
are implemented with machine supported floating-point operations with
the round-to-nearest rounding mode. When an application uses any other
rounding mode, RLibm saves the application's rounding mode, changes the
system's rounding mode to round-to-nearest, computes the correctly
rounded result, and restores the application's rounding mode. This
frequent change of rounding modes has a performance cost.
In this talk I will cover two new methods to avoid the frequent changes
to the rounding mode and the dependence on round-to-nearest. First, our
new rounding-invariant outputs method emulates round-to-zero under all
rounding modes to implement RLibm's polynomial approximations. Second,
our rounding-invariant input bounds method factors any rounding error
due to different rounding modes using interval bounds in the RLibm
pipeline. Both methods make a different set of trade-offs and improve
function that produces correctly rounded results for multiple rounding
modes and representations with up to 32-bits. The key insight is to
build polynomials that produce the correctly rounded result for a
representation with two additional bits when compared to the largest
target representation and with the "non-standard" round-to-odd rounding
mode, which makes double rounding the RLibm math library result to any
smaller target representation innocuous. The resulting approximations
are implemented with machine supported floating-point operations with
the round-to-nearest rounding mode. When an application uses any other
rounding mode, RLibm saves the application's rounding mode, changes the
system's rounding mode to round-to-nearest, computes the correctly
rounded result, and restores the application's rounding mode. This
frequent change of rounding modes has a performance cost.
In this talk I will cover two new methods to avoid the frequent changes
to the rounding mode and the dependence on round-to-nearest. First, our
new rounding-invariant outputs method emulates round-to-zero under all
rounding modes to implement RLibm's polynomial approximations. Second,
our rounding-invariant input bounds method factors any rounding error
due to different rounding modes using interval bounds in the RLibm
pipeline. Both methods make a different set of trade-offs and improve
the performance of resulting libraries by more than 2X.
Looking forward to seeing everyone!
-Ian
As a reminder, if you would like to give a talk or know of someone that would be great for an FPBench Community meeting, please have them fill out the speaker suggestion form!
FPTalks Discussion: https://fpbench.org/subscribe
Ian Briggs
Sep 29, 2025, 11:01:12 AM (9 days ago) Sep 29
to fpb...@fpbench.org
Hi Everyone!
This Thursday, October 2nd at 9:00-10:00 AM pacific time, we’ll have the next FPTalks Community Meeting on this Zoom:
 |
Reply all
Reply to author
Forward
0 new messages