Independent student research — not an investment firm or financial advice
Competitive Comparison
Broadcom vs. Marvell vs. Qualcomm
Three custom/specialty silicon names serving hyperscalers and mobile — one held, one watched, one not owned — compared on growth-adjusted valuation.
Metric
Broadcom (AVGO)
Marvell (MRVL)
Qualcomm (QCOM)
PEG ratio
2.37
2.63
1.22
5yr EBITDA growth
19.3%
29.6%
9.2%
Capex (TTM)
$860.00M
$391.00M
n/a
PEG ratio, compared
AVGO
2.37
MRVL
2.63
QCOM
1.22
Same PEG ratio figures as the table above, plotted for a direct read. Bold = held in this book.
Market cap, P/E, PEG, EV/EBITDA, and capex sourced via public filings and financial-data aggregators (GuruFocus, StockAnalysis, company earnings releases), as of July 2026. PEG ratio sourced primarily from GuruFocus where available; different providers use different growth-rate assumptions, so figures elsewhere for the same stock can vary by several multiples. Gold-highlighted column(s) indicate the name(s) actually held in this book.
Why Broadcom over Marvell and Qualcomm
Marvell actually has the fastest 5-year EBITDA growth of the three (29.6%) and a lower PEG than Broadcom, which makes it a genuinely reasonable name to watch rather than a clear loser in this comparison. I hold Broadcom specifically because its custom-silicon business pairs with a large, sticky infrastructure-software franchise built through acquisitions — a more diversified earnings base than a pure custom-chip designer, which matters given how lumpy hyperscaler custom-silicon orders can be. Qualcomm, cheapest of the three on PEG at 1.22, is fundamentally a different bet on mobile and connectivity chips rather than AI-accelerator custom silicon, which is why it isn't part of this comparison's core thesis despite the attractive multiple.