-
As per the kernel ABI for `/sys/bus/pci/devices/.../irq` which is used by `lspci` to report INTx:
```
If a driver has enabled MSI (not MSI-X), "irq" contains the
I…
-
### Objective:
Implement sub byte unsigned integer quantization baselines from 1-7 to enable users to experiment with low bit quantization in pytorch.
### Tracker:
- [x] Create a UIntx Tensor Subcla…
-
Was chatting with @Chillee about our plans in AO today and he mentioned we should be focusing on a few concrete problems like
1. Demonstrate compelling perf for fp8 gemm at a variety of batch sizes.
…
-
I have this setup and it can work with Frigate.
But 1-2mins of detection will force a reboot on the machine. On the screen there is a flash of "hardware error"
If I do not use Frigate with pcie, th…
-
```
In file included from /home/yansendao/git/riscv-port-jdk8u-ysd/hotspot/src/share/vm/memory/allocation.hpp:28,
from /home/yansendao/git/riscv-port-jdk8u-ysd/hotspot/src/share/vm/…
-
```
Based on -ve int32 discussion (and how this impacts int16/byte if
implemented), prefer to use sint32 if no preference is given.
Will need to rename SignedAttribute ;-p
```
---
Original issue r…
-
I've just added a [unit test](https://github.com/getkyo/kyo/blob/main/kyo-core/jvm/src/test/scala/kyoTest/coreBytecodeSizeTest.scala) to validate the bytecode size of Kyo's core methods. Ideally, all …
-
## Motivation
`Comparable#clamp` is fantastic, and it works very well. However, its signature allows any types for the clamp boundaries (implicitly requiring that `self` defines `# max
return mi…
-
Can't build new version of qtum *mainnet-fastlane-v0.20.2*:
```
/usr/bin/ccache g++ -m64 -std=c++14 -DHAVE_CONFIG_H -I. -I../src/config -U_FORTIFY_SOURCE -D_FORTIFY_SOURCE=2 -DQTUM_BUILD -I. …
-
- [x] (U)IntX.add
- [x] (U)IntX.sub
- [x] (U)IntX.mul
- [x] (U)IntX.v
- [x] UIntX.div
- [x] UIntX.mod
- [ ] UIntX.add_mod
- [ ] UIntX.sub_mod
- [ ] UIntX.mul_mod
- [ ] (U)IntX.lt
- others?
…