newest
Open in hackernews
Implementing a Fast Tensor Core Matmul on the Ada Architecture
https://www.spatters.ca/mma-matmul
2
•
skidrow
•
3h ago
Comments
jhlee525
•
3h ago
This is incredibly useful. Thanks for making the kernels public.
I'm curious if anyone has tried generalizing this to batched matmuls or to sparse inputs on Ada?
Fair software licensing group brands Microsoft EU concessions 'stalling tactic'
https://www.theregister.com/2025/07/18/cispe_microsoft_concessions/
1
•
rntn
•
1m ago
•
0 comments
I got tired of explaining my project to AI assistant so I gave it memory
https://codecontextpro.com/
1
•
Jahboukie
•
1m ago
•
1 comments
Using Omni's AI Assistant on the Semantic Layer
https://mikkeldengsoe.substack.com/p/using-omnis-ai-assistant-on-the-semantic
1
•
mikkelenzo
•
2m ago
•
0 comments
Algorithm on Fleek: How TikTok Is Transforming the English Language
https://lithub.com/algorithm-on-fleek-how-tiktok-is-transforming-the-english-language/
1
•
Michelangelo11
•
3m ago
•
0 comments
Sameness entices, but novelty enchants in fanfiction online
https://www.nature.com/articles/s41599-025-05166-3
1
•
sohkamyung
•
4m ago
•
0 comments
Unlock Gemini's reasoning: A step-by-step guide to logprobs on Vertex AI
https://developers.googleblog.com/en/unlock-gemini-reasoning-with-logprobs-on-vertex-ai/
1
•
simonpure
•
6m ago
•
0 comments
Mr T: Be Somebody or Be Somebody's Fool (1984) [video]
https://www.youtube.com/watch?v=0Y1abMt1UGw
1
•
austinallegro
•
7m ago
•
0 comments
Is that QR code a scam? Here's what to know about 'quishing' before you scan
https://www.cbc.ca/news/canada/qr-code-scams-quishing-1.7587253
1
•
uladzislau
•
7m ago
•
0 comments
Show HN: From Wan video to Wan image
https://www.pruna.ai/blog/wan-image-juiced-image-generation
2
•
dberenstein1957
•
12m ago
•
0 comments
Tao: The Power of the Graph (2013)
https://engineering.fb.com/2013/06/25/core-infra/tao-the-power-of-the-graph/
1
•
tosh
•
12m ago
•
0 comments
Flexoki
https://stephango.com/flexoki
1
•
tosh
•
13m ago
•
0 comments
Theories of Rational Decision
https://github.com/kzollman/RationalDecision
1
•
synthetictask
•
13m ago
•
0 comments
OpenAI's ChatGPT Agent Outperforms the Model Alone: Impact for Developers
https://belitsoft.com/news/chatgpt-agent-openai-20250717
1
•
customllm
•
16m ago
•
1 comments
Google Took Everything Back
2
•
scorpiock
•
18m ago
•
1 comments
SyncMyGroup –> Sync Users, Groups Across Tools
https://www.syncmygroup.com/
1
•
falitjain
•
21m ago
•
1 comments
The Conspiracy of Reason
https://medium.com/sadasant/the-conspiracy-of-reason-f02d6d991a60
1
•
psychoslave
•
22m ago
•
0 comments
Tell HN: Thank you, HN – from shadowban to one of the happiest days of my life
5
•
FerkiHN
•
25m ago
•
1 comments
Global operation targets NoName057(16) pro-Russian cybercrime network
https://www.europol.europa.eu/media-press/newsroom/news/global-operation-targets-noname05716-pro-russian-cybercrime-network
1
•
doener
•
27m ago
•
0 comments
Hermes-3-Dataset
https://huggingface.co/datasets/NousResearch/Hermes-3-Dataset
1
•
tosh
•
28m ago
•
0 comments
UnplugTrump: Free Yourself Digitally from Trump and Big Tech
https://www.kuketz-blog.de/unplugtrump-free-yourself-digitally-from-trump-and-big-tech/
2
•
doener
•
28m ago
•
0 comments
Liqid unveils composable GPU servers with CXL 2.0 memory pooling
https://blocksandfiles.com/2025/07/17/liqid-pcie-gen-5-cxl-composability/
1
•
rbanffy
•
29m ago
•
0 comments
I just shipped the most requested feature for DeadliQ: Outlook Calendar Sync
https://www.deadliq.com/
1
•
raresAIQ
•
29m ago
•
1 comments
RISC-V China Summit: Nvidia: CUDA will soon be ported to the RISC-V architecture
https://en.eeworld.com.cn/news/manufacture/eic701276.html
2
•
fork-bomber
•
29m ago
•
0 comments
Show HN: Presentations at the Speed of Markdown
https://slidesdown.github.io
1
•
jceb81
•
29m ago
•
0 comments
Why not to use iframes for embedded dashboards
https://embeddable.com/blog/iframes-for-embedding
1
•
rogansage
•
29m ago
•
0 comments
OWASP's cure for a sick AI supply chain
https://www.scworld.com/feature/owasps-cure-for-a-sick-ai-supply-chain
1
•
Bender
•
31m ago
•
0 comments
Fortinet FortiWeb Flaw Exploited in the Wild After PoC Publication
https://www.securityweek.com/fortinet-fortiweb-flaw-exploited-in-the-wild-after-poc-publication/
1
•
Bender
•
32m ago
•
0 comments
Google Sues Operators of 10M-Device Badbox 2.0 Botnet
https://www.securityweek.com/google-sues-operators-of-10-million-device-badbox-2-0-botnet/
1
•
Bender
•
32m ago
•
0 comments
Unix, Plan 9, and the lurking Smalltalk [pdf]
https://www.humprog.org/~stephen/research/papers/kell19unix-personal.pdf
2
•
fanf2
•
36m ago
•
1 comments
3D-printed magnetoelastic smart pen may help diagnose Parkinson's
https://techxplore.com/news/2025-07-3d-magnetoelastic-smart-pen-parkinson.html
1
•
PaulHoule
•
40m ago
•
0 comments
jhlee525•3h ago
I'm curious if anyone has tried generalizing this to batched matmuls or to sparse inputs on Ada?