$SWKS blew through the top of the cut this scan — RSI 67.7 to an extreme 83.4, up 19.4% for the week and now the only name in extreme-overbought territory — while $AVGO fell the other way, sliding from RSI 38.1 to 42.2 but still down 13.0% over 30 days, extending its run as the cut's most persistent laggard. $NVDA itself cooled from RSI 60.4 to 49.8 on a 5.2% weekly pullback but remains 10.5% above its 200-day with a 23.2% year-over-year return — a normal cooldown inside an intact trend, not a breakdown. $TLN, the cut's only clean breakdown last scan, stayed below its 200-day and its year-over-year return worsened further to -22.0%.

Data as-of the 2026-09-11 close (Friday's settled session) — every name in this cut carries a fresh bar this scan.

Source: semis.json + ai-infra.json + energy.json, deduped to the core NVDA-adjacent chip/AI-infra/power-demand names — same 20-ticker cut and exclusion methodology as the prior scan (AMZN/PLTR/TAC/P/AIQ/SMH/SFTBY excluded from ai-infra.json; EOSE/NEE/FSLR/ENPH/TAC/XLE/XOP/CMBT excluded from energy.json; SMH/SOXX excluded from semis.json as broad-basket duplicates). Prior scan: 2026-08-28-nvda-ecosystem and 2026-09-04-nvda-ecosystem.

$SWKS is this scan's standout: RSI jumped from 67.7 to 83.4, the cut's highest print by a wide margin and the only name in extreme-overbought territory, on a 19.4% weekly gain and a 24.7% premium to its 20-day average. The news receipt on file for $SWKS was fetched 2026-08-05 — well before this week's move — and shows only routine coverage plus its already-known Qorvo merger story from October 2025; no matching stock-event was found for this window. The cause of this week's surge is not established in this scan's sources.

$AVGO continues as the cut's most persistent laggard even after last scan's broad bounce. RSI ticked up slightly (38.1 to 42.2) but the stock is still down 13.0% over 30 days and 26.9% off its 52-week high, despite reporting an EPS beat on 2026-09-02 ($3.32 actual versus $3.24 consensus, +2.5% surprise) — a beaten estimate that hasn't translated into a recovering chart. $AMAT joined it as a fresh laggard this scan, cooling to RSI 41.7 on a 16.7% 30-day decline, though it remains up 168.3% over the year and 10.9% above its 200-day — a deep pullback, not a breakdown.

$TLN remains the cut's only name that fails the long-trend test outright: still below its 200-day (-12.4%) with a year-over-year return that worsened further to -22.0% (from -16.7% last scan), even though its RSI (44.7) sits mid-pack. $CEG, flagged as a "recovering price, still-lagging trend" name for three straight scans, slipped back below its 200-day this scan (-3.6%, from +0.9%) while its year-over-year return stayed negative (-10.5%) — the recovery from two scans ago has partly reversed.

The bulk of the cohort is still intact. $MRVL, $MXL, $TSM, $INTC, $AMD, $BE, $QCOM all sit above their 200-day with strongly positive year-over-year returns, several extreme: $MXL +364.9%, $INTC +318.3%, $BE +309.8%, $MRVL +254.6%, $AMD +231.6%. $NVDA itself cooled from last scan's post-bounce RSI of 60.4 to 49.8 on a 5.2% weekly pullback, but remains 10.5% above its 200-day with a 23.2% year-over-year return and is still just 7.7% off its 52-week high — the cleanest, least-volatile chart in the cut continues to be the cleanest chart.

• $SWKS surged into extreme-overbought territory. RSI 67.7 → 83.4 (+15.7), price up 19.4% for the week and 26.8% for the month — no cause established in checked news/events.
• $AVGO stayed the cut's laggard despite an earnings beat. RSI 38.1 → 42.2 (+4.1), still -13.0% over 30 days even after a 2026-09-02 EPS beat ($3.32 vs $3.24 est).
• $AMAT cooled into laggard territory. RSI 39.8 → 41.7 (roughly flat), 30-day return -16.7% — still +168.3% 1Y and above its 200-day.
• $CEG slipped back below its 200-day. vs-200-day +0.9% → -3.6%, RSI 66.7 → 53.7 (-13.0) — the recovery flagged two scans ago has partly reversed; year-over-year return stayed negative (-10.5%).
• $TLN's year-over-year return worsened further. -16.7% → -22.0%, still the cut's only clean breakdown; RSI cooled slightly (46.2 → 44.7).
• $NVDA cooled off last scan's bounce. RSI 60.4 → 49.8 (-10.6), price down 5.2% for the week — still +23.2% 1Y and 10.5% above its 200-day.
• $MXL and $BE both extended strong weekly gains. MXL: RSI 44.6 → 56.4, +18.9% for the week. BE: RSI 63.1 → 65.0, +9.1% for the week.
• The oversold band emptied. No name in the cut sits below RSI 40 this scan (last scan $AVGO alone occupied 25-40).
• No stale bars this scan — every name carries a fresh 2026-09-11 close.
• Universe unchanged. Same 20 tickers and exclusion methodology as the prior scan.

RSI measures how stretched a stock is over the past 14 sessions — under 30 signals oversold, over 70 signals overbought, over 80 signals extreme. vs SMA20 is the price gap to the 20-day average. 1Y% and vs SMA200 are the numbers to check before calling anything a breakdown or a trend change: most of this cut sits above its 200-day with a strongly positive year despite this week's mixed moves. $TLN, trading below its 200-day with a worsening negative year, is the one name actually under structural test. $CEG is the trickiest read this scan: its RSI cooled and it slipped back below its 200-day, reversing two scans of "recovering" — watch whether that becomes a genuine re-test or a one-scan blip. $SWKS's extreme RSI print, with no news or event found to explain the week's 19.4% gain, is the name to verify before treating the move as fundamental rather than mechanical (its 20-day average sits 24.7% below the current price).

None.

None — the band emptied this scan; $AVGO (42.2) is now just above the line.

• Genuine breakdown (below 200-day, negative 1-year): $TLN — the cut's only remaining name that fails the long-trend check outright, and it worsened this scan.
• Slipped back, still-lagging trend: $CEG — dropped back below its 200-day this scan after two scans of recovery, year-over-year return still negative.
• Persistent laggard despite the tape: $AVGO — up on RSI slightly but still down double digits over 30 days despite a recent earnings beat.
• Deep pullback, long trend intact: $AMAT, $ASML, $GEV, $ON, $AMKR, $LSCC, $NVDA, $TXN, $ARM, $MRVL, $TSM — eleven names, the bulk of the cut, all above (or essentially at) their 200-day with strongly positive year-over-year returns.
• Strong weekly gain, intact: $MXL, $INTC, $AMD, $BE, $QCOM — clear uptrends posting the cut's best weekly moves.
• Extreme overbought, cause not established: $SWKS — the cut's highest RSI by a wide margin on a 19.4% weekly gain with no corroborating news or event found.

• $SWKS's 19.4% weekly surge into extreme-overbought territory has no established cause — does a driver surface (its pending Qorvo merger is the only known catalyst on file, dated October 2025), or does this look like a mechanical/technical move that fades?
• $AVGO stayed the cut's laggard even after a 2026-09-02 earnings beat — does the earnings result eventually pull the chart up, or does the market keep discounting it?
• $CEG slipped back below its 200-day after two scans of recovery — does it re-test and hold, or is the earlier recovery fully unwound next scan?
• $TLN's year-over-year return worsened for a second straight scan (-16.7% to -22.0%) — does the cut's one clean breakdown deepen further, or find a floor?

• Tape data (price/RSI/SMA/returns): full-scan pre-computed summaries as of the 2026-09-11 close — semis, ai-infra, energy.
• Earnings event for $AVGO: 2026-09-02-avgo-earnings (reported 2026-09-02, EPS beat).
• News check for $SWKS: latest (fetched 2026-08-05, predates this week's move) and research/stock-events/ (no matching event found).
• Prior scans for comparison: 2026-09-04-nvda-ecosystem, 2026-08-28-nvda-ecosystem.