Kernel optimization speedup
METR's line chart drew 51.1K views on X: 0.3x its typical post and #26 of 26 METR charts we track by views. Read Oct 10, 2026.
Can frontier models cost-effectively accelerate ML workloads via optimizing GPU kernels? Our take at METR: yes, and they’re improving pretty steeply, but it’s easy to miss these capabilities without good elicitation and “fair” compute spend
Why it spread
It spread because it explained itself and rode a hot topic that week, reaching 51,067 views, 203 likes, 30 reposts, 10 quotes, and 83 saves.
What worked
- The chart is a line chart that shows KernelAgent achieving increasing speedup over baseline models, so the trend is visible at a glance.
- The post text frames the finding as a clear takeaway about frontier models cost-effectively accelerating ML workloads via optimizing GPU kernels, which makes the chart easy to understand without extra context.
- The topic was timely and the account posted it on 2025-02-14, landing in the top quarter for both source reach and quote rate.
What to borrow
We could pair a simple line chart with a plain-language takeaway in the post text, label the chart clearly, and time it to a hot topic so viewers immediately see the trend and the point.
Careful
Do not over-read the absence of a source line or branding as a weakness, and do not assume the same framing will work for every topic just because it worked for this hot topic.
Written by our analysis model from the measured numbers above, not by METR.
Make it yours
Your numbers, in this line chart.
Starts from the real numbers of "Average speedup on kernel optimization tasks, by model and elicitation method" by METR. 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 22e766f7e6f5.
| Path | Returns | Plan | |
|---|---|---|---|
| GET | /api/v1/charts/22e766f7e6f5 | Title, publisher, date, form, point count; with a Max key, the data link | Open |
| GET | /api/v1/charts/22e766f7e6f5/data | Every point with its provenance and confidence, plus the redraw spec. Add ?format=csv for CSV | Max |
| GET | /api/v1/charts/22e766f7e6f5/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/22e766f7e6f5/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 METR
All 26
Other giants using a line chart
All 295
Questions people ask
How far did METR's "Kernel optimization speedup" chart travel?
51,067 views, 203 likes, 30 reposts and 83 saves on X, read Oct 10, 2026. A normal post from @METR_Evals gets about 153K views.
What kind of chart is it?
A line chart. KernelAgent shows increasing speedup over baseline models.
Can I use METR's chart?
Use the format, not the picture. The chart, logo and design belong to METR. ViralChart redraws the chart type with your numbers and words, and credits the original publisher.