Token Economics Calculator for AI inference hardware
Hi HN, I'm Paul from Tensordyne. We build AI inference systems and chips on logarithmic math. We've put together an interactive Token Economics Calculator to help make apples-to-apples comparisons of inference hardware across vendors: We're interested in how closely it lines up with the community's view of the market. Why we built this Investors and customers kept asking how our system compares to others (NVIDIA and a growing list of startups). Plenty of publicly available data exists, but it's scattered and inconsistent. News articles, provider sites, Artificial Analysis, MLCommons, and now…
In plain words
Token Economics Calculator for AI Inference Hardware is an interactive tool from Tensordyne that compares AI inference hardware across different vendors using consistent metrics. It addresses the fragmented landscape of performance data by consolidating information from multiple sources into standardized comparisons. The tool helps investors and customers evaluate systems from NVIDIA and other inference hardware providers on equal terms, making it easier to assess cost and performance tradeoffs across the market.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hi HN, I'm Paul from Tensordyne. We build AI inference systems and chips on logarithmic math. We've put together an interactive Token Economics Calculator to help make apples-to-apples comparisons of inference hardware across vendors: <https://www.tensordyne.ai/token-economics-calculator> We're interested in how closely it lines up with the community's view of the market. Why we built this Investors and customers kept asking how our system compares to others (NVIDIA and a growing list of startups). Plenty of publicly available data exists, but it's scattered and inconsistent. News articles, provider sites, Artificial Analysis, MLCommons, and now SemiAnalysis’s new InferenceMAX all publish useful numbers — but using different metrics. That makes “what’s actually better (and at what cost)?” surprisingly hard to answer, especially across the broad range of system providers. What the calculator does - Scenario-normalized comparisons: we’ve chosen a few model scenarios and normalized data from different sources to the same key metrics. - Capacity modeling: we estimate racks needed to support a target user load based on model size and KV-cache needs. - Cost & power economics: we estimate tokens/$ and tokens/kWh. You can input your own capex, amortization, colocation and energy costs and see how that impacts TCO. - Architecture: see how different memory architectures (e.g. SRAM-only vs. HBM) could impact profitability. - Like-for-like: see how model performance varies significantly depending on use case by comparing two configurations for the same model. Data sources & gaps We pull from publicly available materials. Where numbers are missing, we estimate (e.g. from chip size / process / HBM capacity) to take a first cut at pricing — and let you swap in your own values. What we’re hoping to learn from you - Which metrics matter most for your use case. - Where our defaults are off (power, users, utilization, etc.). - Systems we should add (including startups) and links to data. If you try it, please tell us what’s confusing, missing, or flat-out wrong. We’ll be in the thread answering questions.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, November 2025
the whole month →
- IB
Life & fun · Nov 2025 · bitsnpieces.dev
- BBoing▲782
Life & fun · Nov 2025 · boing.greg.technology