- Giants
- SemiAnalysis
- Token throughput vs interactivity
Token throughput vs interactivity
SemiAnalysis's line chart drew 71.9K views on X: 1.4x its typical post and #14 of 33 SemiAnalysis charts we track by views. Read Oct 10, 2026.
GB200 NVL72 is incredibly power efficiency compared to H200 & B200. GB200 NVL72 is 10x more power (all in provisioned MegaWatt) efficient per token across certain interactivity (tok/s/user) compared to H200 Single Node on document querying
Why it spread
This chart post spread because an independent author with a big source published a branded line chart showing that token throughput per GPU declines as interactivity increases, and it earned 314 likes, 27 reposts, 8 quotes, 110 saves and 71,874 views.
What worked
- The post pairs a clear line chart with a strong claim that GB200 NVL72 is 10x more power efficient per token across certain interactivity compared to H200 Single Node on document querying scenarios.
- The chart is branded and shows a source line, which supports trust while the account is an independent author drawing on a big source.
- The framing ties a technical optimization story to a simple visual trend, so the chart carries the argument without extra explanation.
What to borrow
We could use the same form by plotting one of our own metrics as a line that declines or rises with a clear axis, then attach a single bold comparison claim and a visible source line.
Careful
Do not over-read the 10x claim as universal, since it applies only across certain interactivity on document querying scenarios, and do not copy the below average quote rate or assume the same reach will follow.
Written by our analysis model from the measured numbers above, not by SemiAnalysis.
Make it yours
Your numbers, the bar chart way.
Starts from the real numbers of "Token Querying Scenario per All in Utility Token Throughput per GPU SemiAnalysis Inference" by SemiAnalysis. Replace every number with yours before you post. Make one like this, step by step
Free pictures carry a small "Made with ViralChart" line. That line is how other people find us.
API · Max
Use this chart via API
On the Max plan you can pull this chart's numbers with the provenance of each one, as JSON or CSV, with its redraw spec, or call it from an agent over MCP. Chart id ed0db06b52a4.
Approximate values. Our two readings of this chart agreed only in part (confidence 0.34 of 1), so check the original before you quote a number. The API marks it lowConfidence, and ?min_confidence=0.5 leaves it out.
| Path | Returns | Plan | |
|---|---|---|---|
| GET | /api/v1/charts/ed0db06b52a4 | Title, publisher, date, form, point count; with a Max key, the data link | Open |
| GET | /api/v1/charts/ed0db06b52a4/data | Every point with its provenance and confidence, plus the redraw spec. Add ?format=csv for CSV | Max |
| GET | /api/v1/charts/ed0db06b52a4/render?view=1 | The draw-only view this page uses | Open |
| POST | /mcp | Tool get_chart_data with this id, for Claude, Cursor and other agents | Max |
curl -s https://viral-chart.com/api/v1/charts/ed0db06b52a4/data \
-H "Authorization: Bearer $VIRALCHART_KEY"claude mcp add --transport http viralchart https://viral-chart.com/mcp \
--header "Authorization: Bearer $VIRALCHART_KEY"60 requests a minute per key; over that you get 429 with Retry-After. Up to 5 keys per account, made on your account page and shown once. Without Max, the data route answers 402 with a link to the plans. The numbers describe a chart its publisher owns, and every file carries a license note.
More from SemiAnalysis
All 33
Other giants using a line chart
All 295
Questions people ask
How far did SemiAnalysis's "Token throughput vs interactivity" chart travel?
71,874 views, 314 likes, 27 reposts and 110 saves on X, read Oct 10, 2026. A normal post from @SemiAnalysis_ gets about 52.2K views.
What kind of chart is it?
A line chart. Token throughput per GPU declines as interactivity increases.
Can I use SemiAnalysis's chart?
Use the format, not the picture. The chart, logo and design belong to SemiAnalysis. ViralChart redraws the chart type with your numbers and words, and credits the original publisher.