Top all-time · #6585
dstack
Company blog · dstack.ai/blogaisystems
dstack is the company blog at dstack.ai, writing about ai and systems. Since 2020, 12 different Hacker News readers have linked it 40 times, most often “Show HN: Dstack – an open-source engine for running GPU workloads”. No feed was found, so read it on the site itself.
Too few sampled comments (7) to say how often HN recommends it.?
Why HN links it
HN readers link dstack.ai as the docs, tutorials and product site for an open-source tool for orchestrating AI compute, and cite its inference benchmark post. The comments do not judge the writing.
Summarised by Claude from a sample of 7 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
dstack is a unified control plane for GPU provisioning and orchestration that works with any GPU cloud, Kubernetes, or on-prem clusters.
Most-linked posts?
- Show HN: Dstack – an open-source engine for running GPU workloads4 people · 6 mentions · first linked 2020-08-12
- Show HN: dstack.ai – publish, track, and share data visualizations3 people · 4 mentions · first linked 2020-03-17
- Orchestrating GPUs in data centers and private clouds2 people · 2 mentions · first linked 2025-02-18
What Hacker News says?
Sure, 1. Here's the most simple tutorial how to make an interactive dashboard and share it: https://docs.dstack.ai/tutorials/dashboards-tutorial It includes screenshots. 2. Here's another tutorial with more realistic data: Output: https://dstack.ai/gallery/d/b56128a3-522e-42d7-8662-9b1a768d... The code for it is available at https://github.com/dstackai/dstack-tutorials-py/blob/master/... Actually we have very few ex…
Comment linking /gallery/d/b56128a3-522e-42d7-8662-9b1a768dbc8c
on “Show HN: dstack – an open-source tool to build data applications easily”Submitted Show HN: Dstack – an open-source engine for running GPU workloads as a story · 101 points
HN discussionThe github README is a bit confusing to understand (for someone with no experience). The [1] landing page shows a much clearer outline on what this actually is. Seems like it could be good for making data-driven dashboard graphs. Although [2] the react library looks like it needs a bit more work. Congrats on shipping something though! [1] https://dstack.ai/ [1] https://github.com/dstackai/dstack-react
Comment linking /
on “Show HN: dstack – an open-source tool to build data applications easily”Submitted Exploring inference memory saturation effect: H100 vs. MI300x as a story · 57 points
HN discussionIm building dstack [1], a tool for continuous training ML models in the cloud. And I’m looking for a co-founder! Happy to chat if you are interested. [1] https://dstack.ai
Comment linking /
on “Ask HN: What developer tools would you like to see?”Submitted dstack (K8s alternative) adds support for AMD accelerators on RunPod as a story · 31 points
HN discussionTo be frank, I’m not a fan of Kubernetes as long as GPU is concerned. Kubernetes has a very large legacy. Why we think dstack can do better: 1. Lightweight-ness - that it very easy to integrate dstack with modern cloud GPU providers. 2. AI-friendly interface built-in - no need to things like KubeFlow and alike. FTR, we integrate dstack with Kubernetes too - already [1] But our native cloud integrations can be a lot…
Comment linking /changelog/0.15.1/
on “Show HN: Dstack – an open-source engine for running GPU workloads”Submitted Orchestrating GPUs in data centers and private clouds as a story · 25 points
HN discussionVast.ai is supported. RunPod not yet but on our list! The full list of all supported targets can be found at https://dstack.ai/docs/installation/
Comment linking /docs/installation/
on “Show HN: Dstack – an open-source engine for running GPU workloads”Submitted Show HN: dstack.ai – publish, track, and share data visualizations as a story · 20 points
HN discussionFrom the screencast on the homepage and in the docs [1], I see Google and Azure clouds are supported as targets. Is there a list of other targets? Is Runpod and vast.ai supported for example (since they're more focused on GPUs)? 1: https://dstack.ai/docs/concepts/dev-environments/#define-a-c...
Comment linking /docs/concepts/dev-environments/
on “Show HN: Dstack – an open-source engine for running GPU workloads”Submitted Show HN: Dstack – Train models and manage data effortlessly as a story · 17 points
HN discussionThis is from https://dstack.ai/blog/2023/06/29/say-goodbye-to-managed-not... Yup, that's why we support dev environments as first-class citizens. Now, also support tasks (for fine-tuning, other batch jobs) and services (deployment, incl. OpenAI compatibility for LLMs).
Comment linking /blog/2023/06/29/say-goodbye-to-managed-notebooks/
on “Show HN: Dstack – an open-source engine for running GPU workloads”Submitted Benchmarking Llama 3.1 405B on 8x AMD MI300X GPUs as a story · 11 points
HN discussion> Any explanation for how it simplifies orchestration. There are many examples in the docs. Here are few: 1. ALl accelerators are supported natively (no operators are required). 2. Distributed task work out of the box [1] 3. For inference, OpenAI compatible gateway is provided automatically for any models deployed; along with authentication. [2] 4. Cluster management is a lot more convenient for AI compared to K8S […
Comment linking /docs/concepts/fleets/
on “Dstack: An alternative to k8s for AI/ML tasks”Submitted Dstack is introducing a Sky-computing platform for GPU as a story · 10 points
HN discussionIts weird, i looked up whether AMD has any benchmarks on the 405B for the MI300x, and came across this one -- https://dstack.ai/blog/amd-mi300x-inference-benchmark/#token... From my understanding, it can get up to around 2500 tokens/s? Both are 8x units (h200 and MI300x)
Comment linking /blog/amd-mi300x-inference-benchmark/
on “Llama 405B 506 tokens/second on an H200”Submitted The state of cloud GPUs in 2025: costs, performance, playbooks as a story · 8 points
HN discussionTL;DR -- ML platforms solve for a million different problems, and most people don't have all (or any) of them. Hamilton is a pretty simple way of organizing code, so it can plug in with a bunch of different approaches (and that's why we think its general purpose). > Data scientist here, stuck in the Dark Ages of "deploying" my models by writing bespoke Python apps that run on some kind of cloud container host like E…
Comment linking /
on “Launch HN: DAGWorks – ML platform for data science teams”Submitted Benchmarking Prefill–Decode ratios: fixed vs. dynamic as a story · 5 points
HN discussionCongrats to the MosaicML and wish all the luck! If anyone is interested in MosaicML alternatives, check out dstack, we are building its OSS and cloud-agnostic alternative: https://dstack.ai . Disclaimer: I'm the founder and CEO at dstack.
Comment linking /
on “Databricks Strikes $1.3B Deal for Generative AI Startup MosaicML”Submitted Benchmarking AMD GPUs: bare-metal, containers, partitions as a story · 4 points
HN discussionSounds similar to https://dstack.ai/docs/
Comment linking /docs/
on “A PyTorch Approach to ML Infrastructure”Submitted DeepSeek R1 inference performance: MI300X vs. H200 as a story · 4 points
HN discussionSubmitted Infrastructure orchestration is an agent skill as a story · 3 points
HN discussionSubmitted Beyond Kubernetes: 2024 recap and what's ahead for AI infra as a story · 3 points
HN discussionSubmitted Built simple CLI to spin up VMs in the cloud to run Python code on-demand as a story · 2 points
HN discussionSubmitted Deploying inference endpoints with PD disaggregation on AMD GPUs as a story · 1 points
HN discussionSubmitted Show HN: Orchestrating GPUs in data centers and private clouds as a story · 1 points
HN discussionSubmitted Auto-shutdown for inactive dev environments–no idle GPUs as a story · 1 points
HN discussion