Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
thinxer
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
thinxer
3y ago
For “cloud-native” apps, JuiceFS is not needed. S3 is not designed for intensive metadata operations, like listing, renaming etc. For these operations, you will need a somewhat POSIX-complaint system. For example, if you want to train on
2.
▲
MTIA v1: Meta’s first-generation AI inference accelerator
(ai.facebook.com)
110 points
by
thinxer
3y ago
|
44 comments
3.
▲
by
thinxer
3y ago
For context: 1. https://twitter.com/rus/status/1641908582814830592 2. https://twitter.com/bighuman/status/1641910129384648706
4.
▲
Text to Shareable Web Apps
(literallyanything.io)
1 points
by
thinxer
3y ago
|
1 comments
5.
▲
Site Generator for Apple Notes
(devlog.notespub.com)
1 points
by
thinxer
3y ago
|
0 comments
6.
▲
Sparks of Artificial General Intelligence: Early Experiments with GPT-4
(arxiv.org)
180 points
by
thinxer
3y ago
|
236 comments
7.
▲
Apple AMX Instructions
(github.com)
6 points
by
thinxer
4y ago
|
1 comments
8.
▲
MLIR CodeGen Dialects for Machine Learning Compilers
(lei.chat)
2 points
by
thinxer
5y ago
|
0 comments
9.
▲
Programming Language Memory Models
(research.swtch.com)
180 points
by
thinxer
5y ago
|
96 comments
10.
▲
Where Are Pixels? – A Deep Learning Perspective
(ppwwyyxx.com)
76 points
by
thinxer
5y ago
|
5 comments
11.
▲
Accelerate
(huggingface.co)
2 points
by
thinxer
5y ago
|
0 comments
12.
▲
by
thinxer
5y ago
On the first sight I fired up Developer Tools to check the banner is *not* <marquee>.
13.
▲
OctoML raises $28M Series B for its machine learning acceleration platform
(techcrunch.com)
1 points
by
thinxer
6y ago
|
0 comments
14.
▲
by
thinxer
6y ago
I expect Photoshop to contain many highly optimized function routines for photo processing. These routines are optimized by hand and usually use architecture-specific features such as SSE/AVX which is not portable. Unoptimized (more po
15.
▲
by
thinxer
6y ago
This looks like Entity-Component-System[1] in game development. Entities are composed of components, and components are often stored vectorized. [1]: https://en.wikipedia.org/wiki/Entity_component_system
16.
▲
Vcpkg: C++ Library Manager for Windows, Linux, and macOS
(github.com)
1 points
by
thinxer
6y ago
|
0 comments
17.
▲
by
thinxer
6y ago
It is because Google wants a monorepo, then Google choose to use Perforce (and later Piper). It is not that Google uses Perforce and thus are limited to a monorepo. The core value behind monorepo (and monorepo-like approaches) explained in
18.
▲
by
thinxer
6y ago
For anyone interested in why monorepo works, I'd recommend the book Software Engineering at Google: Lessons Learned from Programming Over Time. It has detailed the reasons for the One Version Rule and Version Control over Dependency Ma
19.
▲
Hackers and Painters (2003)
(paulgraham.com)
3 points
by
thinxer
7y ago
|
0 comments
20.
▲
ONiO.zero: ultra-low-power wireless MCU using energy harvesting technology
(onio.com)
100 points
by
thinxer
7y ago
|
26 comments
21.
▲
DevOps tech: Trunk-based development
(cloud.google.com)
2 points
by
thinxer
7y ago
|
0 comments
22.
▲
DevOps Practices by Google Cloud
(cloud.google.com)
4 points
by
thinxer
7y ago
|
0 comments
23.
▲
by
thinxer
7y ago
I'm using this: https://github.com/uxbh/ztdns .
24.
▲
Libra: The Path Forward
(developers.libra.org)
3 points
by
thinxer
7y ago
|
0 comments
25.
▲
CRFS: Mount container image directly from registry
(github.com)
1 points
by
thinxer
8y ago
|
0 comments
26.
▲
ScoutFS: the first GPL archiving file system
(scoutfs.org)
3 points
by
thinxer
8y ago
|
0 comments
27.
▲
Glow: Compiler for neural network hardware accelerators
(github.com)
2 points
by
thinxer
8y ago
|
0 comments
28.
▲
WireGuarding the mainline
(lwn.net)
206 points
by
thinxer
8y ago
|
95 comments
29.
▲
Supervisely v2.0: supercharge your training data pipeline with Deep Learning
(hackernoon.com)
2 points
by
thinxer
8y ago
|
0 comments
30.
▲
Caffe2 merging into PyTorch codebase
(github.com)
1 points
by
thinxer
8y ago
|
0 comments
More ›