惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

The GitHub Blog
The GitHub Blog
A
About on SuperTechFans
The Cloudflare Blog
G
Google Developers Blog
博客园_首页
Martin Fowler
Martin Fowler
Apple Machine Learning Research
Apple Machine Learning Research
L
LangChain Blog
D
Docker
C
Check Point Blog
T
Tailwind CSS Blog
博客园 - 司徒正美
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Hugging Face - Blog
Hugging Face - Blog
Microsoft Security Blog
Microsoft Security Blog
V
V2EX
博客园 - 叶小钗
T
The Blog of Author Tim Ferriss
酷 壳 – CoolShell
酷 壳 – CoolShell
IT之家
IT之家
M
MIT News - Artificial intelligence
Microsoft Azure Blog
Microsoft Azure Blog
博客园 - 【当耐特】
GbyAI
GbyAI

Hacker News - Newest: "AI"

AI can't read an investor deck AI as an attorney? Student uses ChatGPT, Gemini to sue UW over alleged racial discrimination Hacking MCP Servers in AI Systems – The Rug Pull: Tool Changes After Approval GitHub - MeepCastana/KubeezCut: Free Web based video editor Can AI judge journalism? A Thiel-backed startup says yes, even if it risks chilling whistleblowers Coming soon: 10 Things That Matter in AI Right Now DARPA built an AI to fact-check enemy weapons claims What explains heterogeneity in AI adoption? When AI Meets Muscle: Context-Aware Electrical Stimulation Promises a New Way to Guide Human Movements - Department of Computer Science AI Changed How We Build. It Did Not Change What Matters. Linux rules on using AI-generated code - Copilot is OK, but humans must take 'full responsibility for the… Meta spins up AI version of Mark Zuckerberg to engage with employees Code Mode: Let Your AI Write Programs, Not Just Call Tools | TanStack Blog GitHub - Delavalom/graft: Go framework for building AI agents. Type-safe tools, multi-provider (OpenAI, Anthropic, Gemini, Bedrock), zero vendor SDKs. India's TCS tops estimates, says new AI models did not dent services demand Gen Z's fading AI hype Strong feeling: we are in a folded AI reality GitHub - machinarii/total-recall-catalog: A reference catalog of latest knowledge retrieval, memory & RAG systems GitHub - mensfeld/code-on-incus: Give each AI agent its own isolated machine with root, Docker, and systemd. Active defense detects and stops threats automatically.. Quantization, LoRA, and the 8% Problem: Benchmarking Local LLMs for Production AI Iran war: We spoke to the man making Lego-style AI videos that experts say are powerful propaganda Powell, Bessent discussed Anthropic's Mythos AI cyber threat with major U.S. banks GitHub - immartian/bellamem: Persistent belief-graph memory for AI agents. Retrieves decisive context by importance — not recency, not RAG, not /compact. recursive-mode: The Repo-Native Operating System for AI Engineering After the attack on Sam Altman's home, will AI CEO's go on the offensive? The biggest advance in AI since the LLM Opus 4.6 vs GPT 5.4 One Prompt Unity World Generation Test “AI polls” are fake polls Client Challenge Can AI be a 'child of God'? Inside Anthropic's meeting with Christian leaders
This Tech Startup Wants To Put A Mini Nvidia Data Center ...
Joshua Hawkins · 2026-06-04 · via Hacker News - Newest: "AI"

Massive data centers are popping up everywhere across the U.S. these days, including Meta's newest data center in Louisiana that's as big as 70 football fields. As AI expands its reach, the need for these data centers is only going to increase, and one startup has come up with an idea to supplement the controversial, energy-consuming data centers. In addition to dedicated hyperscale AI data centers that make your utility bills go up, San Francisco-based Span wants to put a "distributed data center" in your backyard with a device the size of an air conditioner, called the XFRA.

While the specifics of how this arrangement would work have not been defined, Span is already preparing to test the system's potential with a 100-home trial run, which is expected to kick off later this year. The plan will see the company's XFRA units set up alongside new-construction homes. These units will act as distributed data centers, connecting to and feeding off the home's electrical connection to complement hyperscale data centers by supporting "workloads that can run on smaller hardware," such as AI inference, cloud gaming, and high-performance simulation.

How Span's XFRA system will work

Data servers in a line in a data center

shihabsarkar/Shutterstock

According to a white paper document, Span would pay the homeowner to subsidize their electricity and internet bills, while an article from Realtor.com suggests that the company could directly cover these utility costs and then potentially charge a flat-rate fee of up to $150 per month. These specifics don't appear to have been completely outlined just yet, but there's a chance the programs could vary by neighborhood. Additionally, Span says it could install 8,000 XFRA units roughly six times quicker than a new data center could be built, and these installs could be completed at a cost that's around five times lower than what it costs to build a hyperscale data center.

Span will use Nvidia technology to help bring the small servers to life, sourcing Nvidia RTX Pro 6000 GPUs, which have been liquid cooled to provide solid temperature ranges. Span says that by bringing these miniature data centers to people's homes, it can tap into the unused energy that residential homes offer instead of having to completely expand on the power grid. The company looked at how much electrical capacity the average home requires, and it set the maximum consumption for a single XFRA node to 80 amps, as they found that most homes with a 200-amp utility service almost always have 80 amps available.

Not everyone is convinced

Man looking down at papers with a skeptical look on his face

fizkes/Shutterstock

As with most big things involving AI these days, though, not everyone is convinced that this will be as great of an idea as Span makes it out to be. That's because many have posted online about how dangerous this feels as a security or theft issue, with one person on Reddit noting that there is a very big reason data centers tend to have as much security as they do. Others pointed out convenience-related issues, such as noise pollution, service intervals, or young kids using it as baseball target practice.

Others are also concerned about how easy it might make it for companies to harvest your online data. If your internet does run through the company as the implications suggest, then Span — and whatever other company has access to the database — could potentially use it to pull information about what websites you might be visiting and more. This is one of the issues with using public Wi-Fi networks, which is why many people recommend never using them in the first place. For now, the exact nature of how (or when) everything will work is still in the air. However, with the first iteration of its plan set to happen later this year, it shouldn't take long to see how Span's big — or small, rather — data center idea works out.