bazel # reports are there for making 32-bit compile, see <a href="https://github.com/bazelbuild/bazel/issues/3643">upstream</a>
ceph # fails fatally in check() - see <a href="https://bugs.archlinux32.org/index.php?do=details&task_id=21">FS32#21</a>
cuda # cuda 9.0 is 64-bit only
delve # currently no 32-bit support, see <a href="https://github.com/derekparker/delve/issues/20">upstream</a>
dmd # fails to build, <a href="https://issues.dlang.org/show_bug.cgi?id=17833">upstream</a> is unresponsive, see <a href="https://bbs.archlinux32.org/viewtopic.php?pid=492#p492">forum</a>
dotnet-core # not available for 32-bit on linux, see <a href="https://github.com/dotnet/core/blob/master/release-notes/download-archives/2.0.5-download.md">here</a>
linux-hardened # not <a href="https://mirror.archlinux32.org/irc-logs/%23archlinux-ports/2017-06-28.html#04:04:01">really useful</a> on i686 + hard to maintain
libretro-citra # do not build, see FS32#23
libretro-mupen64plus # do not build, see FS32#23
libretro-parallel-n64 # do not build, see FS32#23
libretro-ppsspp # do not build, see FS32#23
luxrays # does not compile on i686 - see FS#54960
mongodb # dropped support for 32-bit in version 3.4
nginx-mod-pagespeed # bundles a 64-bit only version of psol, doesn't build that way
nsjail # kafel has no 32-bit Intel support, see also <a href="https://bbs.archlinux32.org/viewtopic.php?pid=795#p795">forum</a>
nvidia-lts # dropped support after release 390: https://www.phoronix.com/scan.php?page=news_item&px=32-bit-NVIDIA-Drop-Dropping
nvidia-utils # see nvidia-lts
opera # dropped 32-bit version in version 46.0.2597.26
opera-ffmpeg-codecs # as opera itself is blacklisted
ospray # "Your Embree build needs to support at least one ISA >= SSE4.1!"
ovmf # "ERROR: This package must be built under the x86_64 architecture.", UEFI for VMS, of little use anyway with 32-bit machines
parallel-netcdf-openmpi # requires MPI_Offset >= 8, which is not (yet?) possible on 32 bits - see https://github.com/open-mpi/ompi/issues/3195
parity # "32-bit targets are not supported at the moment, You can use '32bit_opt_in' feature. See https://github.com/pepyakin/wasmi/issues/43
python-pytorch # Gloo can only be built on 64-bit systems
qcef # contains CEF binaries (not current ones) for 64-bit and MIPS only, no 32-bit Intel. Discuss with the authors if you need it.
reflector # supports Archlinux mirrors and not Archlinux32 mirrors (yet)
rubinius # Support for non-64bit platforms was deprecated 1 Jun 2016 and has now been removed. If non-64bit support is a critical feature for your application, please email contact@rubinius.com
skia-sharp # does not provide a bin/gn for 32-bit Linux, see FS32#8
skia-sharp58 # does not provide a bin/gn for 32-bit Linux, see FS32#8
tensorflow # bazel doesn't build, unclear whether we could build without cuda/cudnn if bazel would work?
vc # "Unsupported target architecture 'i686'. No support_???.cpp file exists for this architecture", complain upstream if you need it
wiredtiger # WiredTiger requires a 64-bit build
zcash # <a href="https://github.com/zcash/zcash/issues/914">We don't yet support 32bit systems (and we won't by launch) but we can look into this after launch.</a>