frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Emacs-tramp-RPC: high-performance TRAMP back end using MsgPack-RPC

https://github.com/ArthurHeymans/emacs-tramp-rpc
1•fanf2•27s ago•0 comments

Nintendo Wii Themed Portfolio

https://akiraux.vercel.app/
1•s4074433•4m ago•1 comments

"There must be something like the opposite of suicide "

https://post.substack.com/p/there-must-be-something-like-the
1•rbanffy•6m ago•0 comments

Ask HN: Why doesn't Netflix add a “Theater Mode” that recreates the worst parts?

2•amichail•7m ago•0 comments

Show HN: Engineering Perception with Combinatorial Memetics

1•alan_sass•13m ago•1 comments

Show HN: Steam Daily – A Wordle-like daily puzzle game for Steam fans

https://steamdaily.xyz
1•itshellboy•15m ago•0 comments

The Anthropic Hive Mind

https://steve-yegge.medium.com/the-anthropic-hive-mind-d01f768f3d7b
1•spenvo•15m ago•0 comments

Just Started Using AmpCode

https://intelligenttools.co/blog/ampcode-multi-agent-production
1•BojanTomic•17m ago•0 comments

LLM as an Engineer vs. a Founder?

1•dm03514•17m ago•0 comments

Crosstalk inside cells helps pathogens evade drugs, study finds

https://phys.org/news/2026-01-crosstalk-cells-pathogens-evade-drugs.html
2•PaulHoule•19m ago•0 comments

Show HN: Design system generator (mood to CSS in <1 second)

https://huesly.app
1•egeuysall•19m ago•1 comments

Show HN: 26/02/26 – 5 songs in a day

https://playingwith.variousbits.net/saturday
1•dmje•19m ago•0 comments

Toroidal Logit Bias – Reduce LLM hallucinations 40% with no fine-tuning

https://github.com/Paraxiom/topological-coherence
1•slye514•22m ago•1 comments

Top AI models fail at >96% of tasks

https://www.zdnet.com/article/ai-failed-test-on-remote-freelance-jobs/
5•codexon•22m ago•2 comments

The Science of the Perfect Second (2023)

https://harpers.org/archive/2023/04/the-science-of-the-perfect-second/
1•NaOH•23m ago•0 comments

Bob Beck (OpenBSD) on why vi should stay vi (2006)

https://marc.info/?l=openbsd-misc&m=115820462402673&w=2
2•birdculture•27m ago•0 comments

Show HN: a glimpse into the future of eye tracking for multi-agent use

https://github.com/dchrty/glimpsh
1•dochrty•27m ago•0 comments

The Optima-l Situation: A deep dive into the classic humanist sans-serif

https://micahblachman.beehiiv.com/p/the-optima-l-situation
2•subdomain•28m ago•1 comments

Barn Owls Know When to Wait

https://blog.typeobject.com/posts/2026-barn-owls-know-when-to-wait/
1•fintler•28m ago•0 comments

Implementing TCP Echo Server in Rust [video]

https://www.youtube.com/watch?v=qjOBZ_Xzuio
1•sheerluck•28m ago•0 comments

LicGen – Offline License Generator (CLI and Web UI)

1•tejavvo•31m ago•0 comments

Service Degradation in West US Region

https://azure.status.microsoft/en-gb/status?gsid=5616bb85-f380-4a04-85ed-95674eec3d87&utm_source=...
2•_____k•32m ago•0 comments

The Janitor on Mars

https://www.newyorker.com/magazine/1998/10/26/the-janitor-on-mars
1•evo_9•33m ago•0 comments

Bringing Polars to .NET

https://github.com/ErrorLSC/Polars.NET
3•CurtHagenlocher•35m ago•0 comments

Adventures in Guix Packaging

https://nemin.hu/guix-packaging.html
1•todsacerdoti•36m ago•0 comments

Show HN: We had 20 Claude terminals open, so we built Orcha

1•buildingwdavid•37m ago•0 comments

Your Best Thinking Is Wasted on the Wrong Decisions

https://www.iankduncan.com/engineering/2026-02-07-your-best-thinking-is-wasted-on-the-wrong-decis...
1•iand675•37m ago•0 comments

Warcraftcn/UI – UI component library inspired by classic Warcraft III aesthetics

https://www.warcraftcn.com/
2•vyrotek•38m ago•0 comments

Velocity of Money

https://en.wikipedia.org/wiki/Velocity_of_money
1•gurjeet•42m ago•0 comments

Stop building automations. Start running your business

https://www.fluxtopus.com/automate-your-business
1•valboa•46m ago•1 comments
Open in hackernews

Tokasaurus: An LLM inference engine for high-throughput workloads

https://scalingintelligence.stanford.edu/blogs/tokasaurus/
218•rsehrlich•8mo ago

Comments

behnamoh•8mo ago
While Tokasaurus’s Async-TP shows impressive throughput gains, it seems over-engineered for common use cases. The CPU overhead from async tensor parallelism only pays off at 6k+ token batches, and you need NVLink-connected GPUs to see real benefits. Most prod deployments don’t need this complexity — you’re better off with simpler approaches unless you’re specifically optimizing for massive batch throughput. The adaptive manager skipping “optional” tasks under load also feels concerning from a reliability perspective.
bjt12345•8mo ago
Buy surely next years production deployments will be very different to right now, with different use cases...etc
jdiff•8mo ago
Sure. Things change over time. Is there a reason to believe they'd be different in such a way that this would be more useful than in today's landscape? I haven't seen such a forecast myself.
YetAnotherNick•8mo ago
Depends on what production means for you. This is useful for batch production jobs.

Also, this seems very useful for generating synthetic data or labelling a bunch of data. 6k batch size is small for data labelling.

cpard•8mo ago
How big of a use case is synthetic data generation? I’m curious as I see a lot about it coming from academic projects but I haven’t seen much related to commercial use cases
electroglyph•8mo ago
tiny NNs distilled from LLMs can produce some amazing results, i'm surprised it's not more common tbh
cpard•8mo ago
I agree, there are impressive results. This just came out from Berkeley https://arxiv.org/abs/2506.04178

But still, I mainly see work on this direction in academia.

nabakin•8mo ago
> On throughput-focused benchmarks, Tokasaurus can outperform vLLM and SGLang by up to 3x+.

Looks like they don't compare to TensorRT-LLM throughput numbers which, last I checked, are SOTA in open source.

andersa•8mo ago
TensorRT-LLM being open source is a lie, all the important kernels are loaded from cubins.
nabakin•8mo ago
Yeah you're right (although, they started to open source some of that recently iirc). I meant SOTA for inference engines we can actually download and use ourselves.
qeternity•8mo ago
It also appears that this was a sampling benchmark...which is not representative.

Generation benchmark was 5% faster than SGLang.

symbolicAGI•8mo ago
Given chat and API needs for low-latency, llama.cpp is probably still the best choice for self hosted models with or without GPU support. And Ollama is the leader for wrapping llama.cpp.

Because Tokasaurus was mentioned as better than Ollama for conducting darwinian godel machine operations (self-improvement), I looked for the linked repo on GitHub and it was 404. So glad it is back https://github.com/ScalingIntelligence/tokasaurus.

radq•8mo ago
Cool project! The codebase is simple and well documented, a good starting point for anyone interested in how to implement a high-performance inference engine. The prefix sharing is very relevant for anyone running batch inference to generate RL rollouts.
refibrillator•8mo ago
The code has few comments but gotta love when you can tell someone was having fun!

https://github.com/ScalingIntelligence/tokasaurus/blob/65efb...

I’m honestly impressed that a pure python implementation can beat out vLLM and SGLang. Granted they lean on FlashInfer, and of course torch.compile has gotten incredibly powerful in the last few years. Though dynamic shapes have still been a huge thorn in my side, I’ll need to look closer at how they pulled it off…

bobrenjc93•8mo ago
Hi! I work on dynamic shapes in pytorch and would love to hear more about the challenges you’ve run into. We’re always looking to improve the experience, so if you’re open to chatting, feel free to DM me on Twitter (@bobrenjc93) or email me at bobren@meta.com.
gricardo99•8mo ago
since you work on pytorch, what would you say is the best place to ask questions about general usage, trouble shooting? I’ve been struggling with a, what I would consider, a simple torchrun elastic training example, and haven’t found any good resources online. I’ve been spelunking through pytorch but have a feeling a little back and forth with someone familiar with these features would immensely clear things up.
bobrenjc93•8mo ago
PyTorch Dev Discuss is a fantastic forum where many core devs actively participate and answer questions: https://dev-discuss.pytorch.org

In addition to Dev Discuss, a number of core contributors are also active on Twitter. Two particularly helpful and prolific voices are @ezyang and @cHHillee.

Finally, don’t overlook GitHub issues—they’re a surprisingly effective way to start conversations. If you’ve found a bug or have ideas on how to improve the APIs, opening an issue is always welcome.

almostgotcaught•8mo ago
There's also the slack but you gotta know someone to get on that ;)
chillee•8mo ago
I mean, vllm and sglang are both "pure python" essentially as well. But yeah, in ML you rarely require C++ to get good performance for most of the systems people are writing.
AStonesThrow•8mo ago
Stanford was edgy enough to reefer to “toking” in the moniker, but exercises restraint by depicting the titular thunder lizard smoking a putatively conventional tobacco cigarette.

I am hoping to use this “Tokasaurus” nickname with affection for my neighbors. If Stanford is ok with informal usage.

Success with Meta AI / Llama 4:

Hey Meta, I would like to see an image of a Tyrannosaurus Rex, who is clad in a leather jacket, sunglasses, and fedora. He is so cool looking, and smoking a joint of marijuana, and his image is superimposed against a skyline of Phoenix in the golden glow of sunset.

Can you light up the joint with a glowing tip?

Art9681•8mo ago
Proof that attention is not only highly desired by Stanford tech bros, but HN keyboard warriors equipped with LLM tech. Everyone is clever all of the time.
catlifeonmars•8mo ago
I appreciate the double entendre
DiabloD3•8mo ago
Shame this is written in Python, looks very interesting, but I'm no expert in this field.

If there is anything here worth using, it's entirely possible that the llama.cpp crew can save it from vanishing into obscurity.

Szpadel•8mo ago
I'm curious what how big is latency tradeoff. I know assumption here is that it does not matter in those use cases but what order of magnitude it is? 10x? 100x?

this is important for usage in "soft realtime" application, where you do not need instant response but someone is still waiting.

if latency is really big, then it can only be used for basically background processes.