Check of the Muse report "COTS usage trends in RF modules"¶
Checked 2026-10-12. The report was pasted into chat. Its Muse link (https://muse.ai/library/artifacts/cots-usage-trends-in-rf-modules) sent a signed-out browser to the Muse login page, so readers without a Muse account probably can't open it.
Figures spot-checked (6 of the most important; all 6 hold)¶
| Claim in the report | Result | Source opened |
|---|---|---|
| Yole: RF GaN $1.3B (2025) to $2.4B (2031), 11% a year, defence leads | Holds. Note: Yole's defence slice alone grows 10% a year ($592M to $1B), so 11% is the whole market, not defence | Compound Semiconductor, Yole report page |
| Skyworks–Qorvo deal, about $22B, completion "reported" around Oct 2026 | Holds, and can now be firmer: completion on 2026-10-05 is reported (still second-hand summaries, not the closing 8-K) | Globe and Mail, TipRanks |
| ERAI: 1,055 suspect parts in 2024, 748 in 2025 | Holds (ERAI's own annual reports) | ERAI 2025, ERAI 2024 |
| Mercury $8.5M IBAS award, X-band package, more than 80% smaller SWaP | Holds, but it is Mercury's own target, not a test result. The report says "targets", which is right | Mercury press release |
| DDTC proposed narrower USML controls in Sept 2026, not yet law | Holds (proposed 2026-09-30, comments due 2026-11-30) | Thompson Hine, Export Compliance Daily |
| MACOM ENGSD00088: 2–18 GHz, 2.5 W, 24% PAE, 3.5 dB NF | Holds per MACOM's launch note; Mouser lists 25% "average" PAE | MACOM IMS 2025 preview, Semiconductor Today |
Not checked: the other figures (ThinkRF, Per Vices, Keysight FieldFox, LEO payload market, IndexBox, DataInsightsMarket, GrowthMarketReports, GII/Mordor, Honeywell/CAES details).
Weak spots for a public post¶
- The growth chart mixes one strong source with weak ones. The report itself calls IndexBox a "content-marketing estimate" and says the bars are not comparable. A chart readers screenshot loses those caveats. Only the Yole bar is solid.
- None of the bars measures COTS. The report says so ("no tier-1 COTS-only market size"). A title promising "COTS usage trends" over a chart of other markets invites the comment "these numbers aren't about COTS".
- The MagicON banner sits on work MagicON didn't do. The report is a web-research summary, not a MagicON engine result. Fine on LinkedIn if presented as reading notes; on Reddit a branded report reads as an ad.
- Overlap with COTS Post 1 (planned Wed 2026-10-21, same topic, its own Visiongain chart).
Decision (2026-10-12)¶
Attach the report as a PDF to COTS Post 1 on LinkedIn. No Reddit version of the report; Post 1R stays a plain question.
Fixes to ask Muse for (paste this into Muse)¶
Please revise the report "COTS usage trends in RF modules": 1. In section 02, drop the bar chart, or keep only the Yole RF GaN bar. Keep the table, with its caveats, for the other figures. 2. In section 01, replace the 5.8% IndexBox stat tile with one from a stronger source, or drop it. 3. In section 01, make clear that the 11% Yole rate is for the whole RF GaN market; Yole puts the defence slice at 10% a year ($592M in 2025 to $1B in 2031). 4. Update the Skyworks–Qorvo items: completion on 5 October 2026 is now reported (Globe and Mail, TipRanks summaries of the company announcements). Keep saying it is secondary reporting. 5. Add one line under the title: "Reading notes from public sources, not MagicON engine results." 6. Export as a PDF.