29 Apr
2026
29 Apr
'26
9:58 p.m.
> Huge integers I can understand> (because sometimes you need> results that are not just precise> but exact), but what's the use> for float1024? float256
> Huge integers I can understand> (because sometimes you need results> that are not just precise but exact),> but what's the use for float1024?> float256
So far there have been severalinteresting reactions regarding:
* behavior of the (u)int128 typein relation to built-insand standards,
* documentation queries,
* concerns/ideas that even more "big"types could potentially be useful.
I'm very interested in "big number"types. Yet I feel that enlargingthe scope of this thread to othertypes may probably better left forfuture threads.
I can reaffirm need for (u)int128,as recently as we've used it inongoing specification work today.
Thanks for the enriching discussionso far.
- Christopher
On Wednesday, April 29, 2026 at 10:21:13 PM GMT+2, Peter Taraba via Boost <boost@lists.boost.org> wrote:
these two publications should explain it:
https://www.sciencedirect.com/science/article/pii/S2666827021000633?via%3Dihubhttps://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2023.1144886/full going wider with DNNs for higher accuracy. let me know if you have more questions. Best,
Peter Sent: Wednesday, April 29, 2026 at 1:14 PM
From: "Rainer Deyke via Boost" <boost@lists.boost.org>
To: boost@lists.boost.org
Cc: "Rainer Deyke" <rdeyke@gmail.com>
Subject: [boost] Re: Seeking Endorsement: Boost.int128On 4/29/26 21:24, Peter Taraba via Boost wrote:
> Me as a python user for large numbers, I would also love to have int128 in C++
> as well as float128, float256, float512 & float1024 ...
Huge integers I can understand (because sometimes you need results that
are not just precise but exact), but what's the use for float1024?
float256 would already have enough precision to represent measure the
size of the observable universe at Planck length precision.
I'm not saying there isn't a use for such a type, but I'm genuinely
curious as to what that use is.
--
Rainer Deyke - rainerd@eldwood.com
_______________________________________________
Boost mailing list -- boost@lists.boost.org
To unsubscribe send an email to boost-leave@lists.boost.org
https://lists.boost.org/mailman3/lists/boost.lists.boost.org/
Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/3KBA7AHE7ZOB6OQKTFHRDLXOSIGXC3ZJ/_______________________________________________
Boost mailing list -- boost@lists.boost.org
To unsubscribe send an email to boost-leave@lists.boost.org
https://lists.boost.org/mailman3/lists/boost.lists.boost.org/
Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/3PORE4XPNMOEMGPAD72ENF4PJSEYFR7B/