--- title: "Marvell trades at 77.4 times earnings as SK hynix's first commercial memory runs on a startup's chip" description: "At the Open Compute Project summit, SK hynix put its first named commercial CXL memory product on a startup's compute-in-memory chip, due for mass production before Marvell's rival expander samples. Marvell still trades at 77.4 times trailing earnings on a fiscal 2028 outlook management just raised to $20 billion." publisher: "The Inference" section: "Silicon" published: 2026-10-20T13:35:58.474Z modified: 2026-10-20T13:35:58.474Z canonical: https://theinference.org/article/marvell-trades-at-77-4-times-earnings-as-sk-hynix-s-first-commercial-memory-runs-on-a-startup-s-chip language: en keywords: "Marvell, SK hynix, Semiconductors, Memory, AI Infrastructure" --- # Marvell trades at 77.4 times earnings as SK hynix's first commercial memory runs on a startup's chip > At the Open Compute Project summit, SK hynix put its first named commercial CXL memory product on a startup's compute-in-memory chip, due for mass production before Marvell's rival expander samples. Marvell still trades at 77.4 times trailing earnings on a fiscal 2028 outlook management just raised to $20 billion. [SK hynix](https://theinference.org/markets/companies/sk-hynix), the world's second-largest DRAM maker, spent the Open Compute Project summit in San Jose this month showing [CXL Hybrid Memory, its first named commercial product](https://www.techtimes.com/articles/329194/20261019/sk-hynix-debuts-cxl-computational-memory-xcena-mx1-samsung-commits-foundry.htm) on CXL, [the interconnect that pools server memory](https://theinference.org/article/everspin-s-cxl-mram-demo-is-a-sales-call-and-the-rig-on-the-crate-shows-it). The controller inside came from XCENA, a startup founded in 2022 by Samsung and SK hynix veterans, not from [Marvell](https://theinference.org/markets/companies/marvell) [1]. Marvell has priced that link for investors. It [sizes the market for silicon connecting AI chips at $37 billion](https://chipdispatch.com/articles/marvell-puts-a-37-billion-price-tag-on-ai-chip-interconnect-3b5a2803), its own estimate relayed through MarketBeat [2], and LightCounting founder Vladimir Kozlov puts AI scale-up networking above $50 billion by 2030 [3]. ## Marvell papered the rack in one week On October 12 it announced [second-generation Structera X memory expanders](https://www.techpowerup.com/353633/marvell-unveils-second-generation-structera-x-industry-first-six-channel-cxl-3-2-memory-expanders-for-ai-inference) carrying up to 9 terabytes per controller [4]. On October 13 came [the Alaska E 3.2T retimer](https://www.streetinsider.com/Business+Wire/Marvell+Bolsters+Connectivity+Portfolio+with+New+3.2T+Protocol-aware+Retimer+and+Second-generation+1.6T+AEC+DSP/27178567.html), 16 lanes of 224G signaling, and a 3nm Alaska A 1.6T cable DSP drawing 12.8 watts, both already sampling [3]. On October 19 it [co-founded an OCP security group](https://ava.fund/news/marvell-launches-six-cxl-memory-expanders-and-alaska-connectivity-chips-forms-oc-19102026) with [Microsoft](https://theinference.org/markets/companies/microsoft) and [Oracle](https://theinference.org/markets/companies/oracle) [5]. At its October 6 investor day, management [raised its fiscal 2028 revenue outlook to about $20 billion](https://finance.yahoo.com/markets/stocks/articles/just-happened-marvell-stock-180620226.html), against $9.45 billion of trailing revenue, and set a fiscal 2031 target of $70 billion to $90 billion [6]. *Chart: **Ten straight quarters of revenue growth, to $2.74 billion in August** Line chart of Marvell quarterly revenue rising for ten consecutive quarters to $2.74 billion in the quarter ended August 2026* *Marvell quarterly revenue, US$ billions, from SEC filings; labels mark the calendar quarter in which each fiscal quarter closed, latest being the quarter ended August 1, 2026. [7]* ## Compute moved inside the memory Structera X makes memory bigger; the host processor still does the thinking. XCENA's MX1 puts 3,072 RISC-V cores inside the controller, and at Hot Chips 2026 it measured up to 4.7 times a host CPU's throughput over CXL at 18.7 times the energy efficiency, on 40 watts [1]. Samsung starts MX1 mass production on its foundry lines by year-end, with revenue from 2027 [1]. Marvell's second-generation expander samples in the first quarter of 2027 [4]. The rival chip reaches production before Marvell's answer takes its first sample. The DRAM makers hedged: Samsung and SK hynix both supplied endorsement quotes for the Structera launch even as they backed the startup [4]. And on October 14 two accounts posted that Marvell's OCP booth showed "significant levels of cheating": waveform averaging used for an eye measurement, a TDECQ reading flagged with a question mark, an RLM of 0.9 against a 0.95 spec [8][9]. The claims are unverified, and no Marvell response appears in the coverage gathered. ## The bill arrives in 2028 [A KeyBanc note relayed on X on October 19](https://x.com/pequityresearch/status/2112198335197044770) has Marvell likely to win Microsoft's Maia, the in-house AI accelerator, over MediaTek and [Broadcom](https://theinference.org/markets/companies/broadcom) for the Maia 400 generation [10]. Maia 300 enters production in the first half of 2027 at roughly 500,000 units; Maia 400 ramps in the first half of 2028 [10]. [A sector table posted the same day](https://x.com/jamescal2026/status/2112307158390042974) puts Marvell at 40 times forward earnings, against 17 for [Nvidia](https://theinference.org/markets/companies/nvidia) and 5 for [Micron](https://theinference.org/markets/companies/micron) [11]. The shares fell 17.2% across the five sessions to October 16 and still closed at 77.4 times trailing earnings [6]. *Chart: **Marvell's 40x forward earnings trails only AMD and Intel** Bar chart of forward P/E ratios across nine semiconductor stocks, Marvell at 40x behind AMD 47x and Intel 52x, Micron lowest at 5x* *Forward price-to-earnings ratios from a semiconductor sector table posted on X, October 19, 2026; a crowd-posted table, not consensus estimates. [11]* Three dates settle it: XCENA on Samsung's production lines by year-end [1], Marvell's Structera sampling in the first quarter of 2027 [4], and the report against a $3.15 billion quarterly guide due around November 30 [6]. Nothing in the Maia thesis collects revenue before 2028 [10]. ## Takeaway The rival chip reaches Samsung's production lines by year-end while Marvell's answer is still months from sampling; a 77.4x trailing multiple rests on a Maia win that collects no revenue before 2028. ## Sources 1. [TechTimes, SK Hynix Debuts CXL Computational Memory on XCENA MX1; Samsung Commits as Foundry, 19 October 2026](https://www.techtimes.com/articles/329194/20261019/sk-hynix-debuts-cxl-computational-memory-xcena-mx1-samsung-commits-foundry.htm) 2. [Chip Dispatch, Marvell Values AI Chip Interconnect Market at $37 Billion, 13 October 2026](https://chipdispatch.com/articles/marvell-puts-a-37-billion-price-tag-on-ai-chip-interconnect-3b5a2803) 3. [Business Wire via StreetInsider, Marvell Bolsters Connectivity Portfolio with New 3.2T Protocol-aware Retimer and Second-generation 1.6T AEC DSP, 13 October 2026](https://www.streetinsider.com/Business+Wire/Marvell+Bolsters+Connectivity+Portfolio+with+New+3.2T+Protocol-aware+Retimer+and+Second-generation+1.6T+AEC+DSP/27178567.html) 4. [Marvell press release via TechPowerUp, Marvell Unveils Second-generation Structera X, Industry-first Six-channel CXL 3.2 Memory Expanders for AI Inference, 12 October 2026](https://www.techpowerup.com/353633/marvell-unveils-second-generation-structera-x-industry-first-six-channel-cxl-3-2-memory-expanders-for-ai-inference) 5. [Ava Advisor (Simply Wall St), Marvell Launches Six CXL Memory Expanders and Alaska Connectivity Chips, Forms OCP Security Group, 19 October 2026](https://ava.fund/news/marvell-launches-six-cxl-memory-expanders-and-alaska-connectivity-chips-forms-oc-19102026) 6. [Yahoo Finance (Trefis), What Just Happened To Marvell Stock?, 19 October 2026](https://finance.yahoo.com/markets/stocks/articles/just-happened-marvell-stock-180620226.html) 7. [Marvell quarterly figures from its SEC filings, retrieved 20 October 2026](https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001835632) 8. [@insane_analyst on X, post on Marvell OCP 2026 booth measurements, 14 October 2026](https://x.com/insane_analyst/status/2110437821970604296) 9. [@bubbleboi on X, post on Marvell OCP 2026 booth measurements, 14 October 2026](https://x.com/bubbleboi/status/2110446890399715819) 10. [@pequityresearch on X, relaying KeyBanc note on Microsoft Maia 300 and Maia 400, 19 October 2026](https://x.com/pequityresearch/status/2112198335197044770) 11. [@jamescal2026 on X, semiconductor forward P/E table, 19 October 2026](https://x.com/jamescal2026/status/2112307158390042974) --- Marvell trades at 77.4 times earnings as SK hynix's first commercial memory runs on a startup's chip — The Inference. Canonical: https://theinference.org/article/marvell-trades-at-77-4-times-earnings-as-sk-hynix-s-first-commercial-memory-runs-on-a-startup-s-chip