All three margin figures are real, but the comparison is not a cost test — and CXMT's own prospectus corroborates the >30% cost gap rather than contradicting it
The question
Verbatim: "Why did CXMT's Q2 2026 gross margin (87.6%) exceed SK hynix (83.2%) and Micron (84.9%)? Product mix, quarter timing, depreciation/subsidy accounting, or a real cost-gap narrowing? Does the CXMT prospectus disclose per-unit cost, utilization, or inventory days that would allow a direct cost-per-bit estimate?"
Context: this closes the first and fourth open follow-ups of [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]], which flagged ChangXin Memory Technologies' (CXMT's) Q2 2026 margin as "the most interesting number in the brief" and could not explain it. This brief sizes a framework parameter and audits a disclosure. It contains no position recommendation, price target, or buy/sell call.
What we already know (from the vault)
- The premise as it entered the queue, and its provenance. CXMT H1 2026 revenue of RMB 150.3bn with gross margin of 84.84% for the half, roughly 79% in Q1 and 87.59% in Q2, against Micron at 84.9% and SK hynix at 83.2%; revenue mix LPDDR 65.9% / DDR 31.6%. All three figures reached the vault through one secondary source, TrendForce, not from the three companies' own filings [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]].
- The parent brief named the four candidate explanations and declined to pick, saying "public data cannot distinguish between them" while noting that a real cost-gap closure "cannot be ruled out" but "is also not the most likely reading." Its Indicator 1 set the crossing condition as Q2's gap persisting through Q3 [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]].
- The >30% cost-per-bit gap is single-sourced to one SemiAnalysis 1Q26 estimate, never refreshed and never independently reproduced; the ~10% parity threshold was proposed without derivation [[2026-07-03-cxmt-dram-capacity-memory-oligopoly-risk]], [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]].
- Density, not yield, is the physical gap. CXMT's G4 16Gb DDR5 die measures 66.99 mm², or 0.239 gigabits per square millimetre (Gb/mm²), against roughly 0.40 Gb/mm² for the Samsung equivalent, and CXMT's bit-share-to-capacity-share ratio is forecast flat at about 0.7 through 2027 [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]], [[2026-09-08-cxmt-ymtc-bit-share-phase3-threshold]].
- The method this brief reuses, and the sibling that built it. Tonight's [[2026-11-04-dram-2023-trough-cash-cost-floor-cxmt-position]] derives Micron's dynamic random-access memory (DRAM) cash cost of goods sold from Securities and Exchange Commission (SEC) eXtensible Business Reporting Language (XBRL) tags using the identity
cash cost per bit = realised average selling price x (cash COGS / revenue). That brief's central mechanical warning is the one this question needs: a reported cost-of-goods-sold ratio is contaminated by inventory write-downs and by the depreciation share inside cost of goods sold, and selling previously written-down inventory flatters later quarters at an artificially low carrying value. This brief does not redo that derivation; it applies the same caution to CXMT. - The 2026 price spike was concentrated in conventional DRAM, not high-bandwidth memory (HBM). Conventional DRAM contract prices rose about 90-95% quarter on quarter in Q1 2026 and 58-63% in Q2 2026, while 2026 HBM capacity was sold out in advance under long-term agreements and HBM repricing was expected only with the 2027 HBM4 agreements [[2026-05-18-memory-cycle-v1.1]], [[2026-06-28-chip-memory-cycle-phase2-phase3-indicators]], [[2026-07-07-dram-hbm-phase2-phase3-early-signals]]. Pricing in those agreements is dominantly a floor-and-ceiling band rather than a flat fixed price, which caps but does not eliminate the effect [[2026-11-01-hbm-fixed-price-offtake-primary-filings]].
What the web says
Premise check, figure by figure.
- Micron's 84.9% is real, hard disclosure, and on the wrong period. Micron's fiscal Q3 2026 Form 8-K exhibit reports revenue of $41,456M, GAAP gross margin of $35,056M or 84.6%, and non-GAAP gross margin of $35,199M or 84.9%, for the quarter ended May 28, 2026. Prior quarter: 74.4% GAAP / 74.9% non-GAAP on $23,860M. Year-ago quarter: 37.7% / 39.0% on $9,301M. Fiscal Q4 2026 guidance is revenue of $50.0bn ± $1.0bn at gross margin of approximately 86% on both bases (Micron FQ3 FY26 8-K Ex. 99.1). Micron's fiscal Q3 spans roughly March to May, so it is not calendar Q2, and it sits inside a sequence rising 10.0 percentage points (pp) and then about 1.1pp per quarter.
- Micron's business-unit margins are the single most useful disclosure in this brief, and they invert the naive reading. Same 8-K, fiscal Q3 2026 revenue and operating margin by unit: Cloud Memory $13,769M at 78%, Core Data Center $11,524M at 83%, Mobile and Client $11,521M at 86%, Automotive and Embedded $4,634M at 75%. The HBM-heavy unit has the lowest margin of the four and the mobile/client unit, which is the closest analogue to CXMT's 65.9%-LPDDR mix, has the highest. Micron gives no DRAM-versus-NAND revenue or margin split, and reports no inventory write-down or underutilisation charge in the period (both confirmed absent).
- SK hynix's 83.2% is consistent with its own hard numbers but was not confirmed from its gross-profit line in this run. SK hynix announced Q2 2026 (calendar April to June, Korean International Financial Reporting Standards) revenue of about KRW 79.3tn and operating profit of about KRW 60.5tn, an operating margin reported at 76% and an all-time high, against roughly 72% in Q1 (Pulse2, Khan, search-result extracts). A 76% operating margin plus operating expense of roughly 7pp of revenue reconciles to an 83.2% gross margin, so the figure is arithmetically plausible. Flagged: two SK hynix newsroom uniform resource locators returned HTTP 404 and were not retried, so the gross-profit line itself is unverified here. One outlet reports a net margin of 118%, which is a signal to treat secondary SK hynix coverage carefully.
- CXMT's 87.59% is corroborated, and the gap widened in Q3 2026. CXMT's Q3 2026 report shows revenue of about RMB 118.71bn, up 613.18% year on year, attributable net profit of about RMB 54.43bn against total net profit of about RMB 88.98bn, and gross margin of 89.53%, up 1.94pp from the previous quarter (Gizmochina, Unite.AI, Techtroduce, all secondary; the Shanghai Stock Exchange filing was not reached). 89.53 minus 1.94 equals 87.59, so CXMT's own sequential comparison independently reproduces the Q2 figure that entered this queue through TrendForce. The premise's CXMT number survives.
- The prospectus is reachable, and it is the 2025-12-29 filing draft, not the final prospectus. A full English translation of CXMT's Shanghai Stock Exchange Science and Technology Innovation Board (STAR Market) initial-public-offering prospectus filing draft is published at saranormous.github.io/filing-websites/cxmt (730,712 characters of text; authorship of the translation is not stated, so it is an unofficial translation and a lower evidence tier than a filing). Latest period covered is January to June 2025, balance sheet dated 2025-06-30. It therefore cannot speak to Q2 2026 at all. Disclosed, under Chinese Accounting Standards for Business Enterprises:
- Main-business gross margin -3.67% (2022), -2.19% (2023), 5.00% (2024), 12.72% (H1 2025); overall gross margin -3.13%, -1.93%, 5.58%, 13.0%. Revenue $1.1bn, $1.3bn, $3.3bn, $2.1bn.
- Year-on-year change in the selling price per unit of major DRAM products: -43.54% (2023), +55.08% (2024). A percentage change only; no price level is given.
- Depreciation: RMB 5,243M, 10,555M, 14,875M, 11,349M. Fixed-asset book value RMB 68,016M, 84,452M, 153,132M, 171,593M, being 46.05%, 43.81%, 56.38%, 59.19% of total assets.
- Inventory balance RMB 9,434M, 14,180M, 21,215M, 28,073M; impairment provision RMB 2,517M, 3,444M, 2,665M, 1,986M; inventory write-down loss RMB -4,078M, -11,500M, -1,827M, -628M (cross-checked against the same figures given in 万元: 1,150,021.06 万元 confirms RMB 11.5bn for 2023).
- Inventory turnover ratio, excluding the effect of write-off of write-down provisions: 1.47, 1.64, 1.42, 1.19, with the stated cause of the 2022 and 2024 lows being work-in-progress build-up as new production lines started.
- Not disclosed anywhere in that document, on a targeted term sweep of the full text: any absolute per-unit or per-bit cost; any capacity utilisation rate; any wafer-start, wafer-output or bit-shipment volume; any cost-of-revenue composition split into materials, depreciation, labour and manufacturing overhead; any useful-lives table for machinery and buildings (the policy section refers to useful lives being reviewed annually but states no years); and any government-grant accounting policy or annual grant amount. The phrase "unit cost" appears only in use-of-proceeds narrative ("reduce unit costs"), and "utilisation" only inside a quoted government policy document.
Convergences and contradictions
- Convergence: the premise is real. All three figures are genuine gross-margin numbers of the stated magnitude, and CXMT's is now double-sourced through its own Q3 sequential comparison. Nothing here is a transcription error or a mislabelled metric. The parent brief's arithmetic was sound.
- Contradiction: the three figures are not the same measurement, and the comparison cannot carry the weight placed on it. Three separate mismatches stack in the same direction. Period: Micron's is March to May under United States Generally Accepted Accounting Principles with an August fiscal year end, SK hynix's is April to June under Korean International Financial Reporting Standards, CXMT's is April to June under Chinese standards. Scope: Micron and SK hynix are blended DRAM plus NAND flash with HBM inside the DRAM, CXMT is effectively pure conventional DRAM with no HBM and no NAND. Standard: Chinese Accounting Standards for Business Enterprises place inventory write-downs in asset-impairment losses below gross profit, whereas Micron books them inside cost of goods sold, which the sibling brief documents at $1,430M in one 2023 quarter [[2026-11-04-dram-2023-trough-cash-cost-floor-cxmt-position]].
- The write-down placement is confirmed by arithmetic, not inferred. CXMT reported a main-business gross margin of -2.19% in 2023 while taking an inventory write-down loss of RMB 11,500M against revenue of about RMB 9.2bn. A write-down of roughly 125% of revenue sitting inside cost of goods sold is incompatible with a cost-of-goods-sold ratio of 102.19%. The write-downs are therefore definitively below the gross-profit line.
- Resolved contradiction with the parent brief. The parent brief read the margin ordering as evidence straining against the >30% cost gap. Mix-corrected and run on pre-boom periods, the same method points the other way: it corroborates the gap. The parent brief's framing of the number as anomalous should be retired.
Synthesis for RDCO
Verdict on the four candidates: product mix is supported and is sufficient on its own; quarter timing is supported and small; two specific Chinese-accounting channels are supported and sized; a real cost-gap narrowing is not supported and the new evidence argues against it. Confidence: medium-high on mix and timing, medium on the accounting channels, medium-high on the rejection of a cost-gap closure.
Why mix alone closes the gap, by two independent routes. Route one uses Micron's business units. The four unit operating margins weighted by unit revenue reconcile to 81.3%, against a consolidated gross margin of 84.6%, so Micron allocates essentially all operating expense to units and the unit-level gross margin is roughly the unit operating margin plus about 3.3pp. That puts Micron's Mobile and Client unit at a gross margin near 89% and its HBM-heavy Cloud Memory unit near 81% (Ray derivation; the uniform operating-expense allocation is an assumption, since research-and-development intensity differs by unit). Route two solves the blend. Writing Micron's cost-of-goods-sold ratio as a revenue-weighted average of its DRAM and NAND ratios, with NAND at 15-20% of revenue and a NAND gross margin of 60-70%, the DRAM-only gross margin solves to about 89% (Ray derivation; the NAND weight and margin are assumptions this brief could not verify, because Micron publishes no DRAM-versus-NAND split). Two methods converge on Micron DRAM-only at roughly 88-90% for a quarter whose consolidated print was 84.9%, which is above CXMT's 87.59%. The ordering in the premise reverses once scope is matched. The mechanism is the one the vault already documents: 2026 was a conventional-DRAM price spike, HBM was pre-sold under multi-year agreements inside floor-and-ceiling bands, and NAND margins trailed, so a producer's margin rank in 2026 is mostly a ranking of exposure to spot-repriced conventional DRAM and only faintly a ranking of cost. CXMT, at 65.9% LPDDR with no HBM and no NAND, had the purest exposure available. SK hynix, carrying the largest HBM book and a NAND business, had the least, which is exactly the observed order.
Quarter timing is real but second-order, and the accounting channels are larger than expected. Micron's calendar-Q2 equivalent, month-weighting fiscal Q3's 84.9% over April and May with fiscal Q4's guided 86% over June, is roughly 85.3-86.5%, so timing accounts for about 0.4-1.6pp of the 2.7pp headline gap. On accounting, the subsidy question could not be answered: the filing draft discloses no grant policy and no grant amounts, so whether CXMT uses the net method, which reduces an asset's carrying value and therefore its depreciation and its cost of goods sold, or the gross method, which routes grants to other income below gross profit, remains open and is the largest unresolved confounder in this brief. Two adjacent channels are sizeable and quantifiable. First, written-down inventory: CXMT carried an impairment provision of RMB 1,986M at 2025-06-30, the residue of RMB 15.6bn of 2022-2023 write-downs, and it is selling that inventory into 2026 boom prices at a written-down carrying value. If the full balance was consumed in H1 2026, it is worth roughly 1.3pp of H1 2026 gross margin and up to about 2.3pp of Q2 2026 (Ray derivation; full-consumption-in-period is an assumption, and the Q2 figure uses a Q2 revenue of roughly RMB 85-90bn backed out of the disclosed H1 total, since CXMT's Q1-versus-Q2 revenue split was not sourced). That single channel is the same order of magnitude as the entire anomaly. Second, overhead absorption: CXMT's H1 2025 depreciation of RMB 11,349M is roughly 76% of its H1 2025 revenue, which cannot all sit in cost of revenue alongside a positive 12.72% gross margin, so a large share is being capitalised into an inventory balance that tripled from RMB 9.4bn to RMB 28.1bn. At a turnover ratio of 1.2-1.6 against Micron's typical two-to-three turns, those low-cost layers reach the income statement slowly and land in high-price quarters. The implied depreciation rate on net fixed assets of 12.5-14.0% (Ray derivation from the disclosed series) hints at useful lives longer than Micron's five-to-seven-year equipment lives, but with no useful-lives table disclosed this stays a hint, not a finding.
The load-bearing result: the margin-residual method, run where it has discriminating power, corroborates the >30% cost gap. Boom-quarter margins are a terrible cost instrument because the metric is violently leveraged there. At an 87.6% margin, one percentage point of gross margin corresponds to roughly an 8% move in unit cost or a 9% move in price, so the whole 2.7pp headline gap is equally well explained by CXMT's unit cost being about 18% lower than Micron's or by its realised revenue per unit of output being about 22% higher, and nothing in public data separates the two. Pre-boom periods are the opposite. Taking CXMT's own H1 2025 main-business gross margin of 12.72% against Micron's hard-disclosed overlapping fiscal Q3 2025 GAAP gross margin of 37.7%, the cost-of-goods-sold ratios are 0.873 and 0.623. Their ratio is 1.40, and applying the lineage's 5-10% CXMT price discount gives a cost-per-unit-of-output gap of 26-33% (Ray derivation). That brackets SemiAnalysis's >30% almost exactly, from two sources neither of which is SemiAnalysis. Three corrections push it wider, not narrower: Micron's 37.7% is blended and its DRAM-only figure was higher, so a scope-matched comparison implies a gap above 40%; CXMT's 12.72% excludes RMB 628M of write-downs that Micron-style accounting would have put in cost of goods sold, worth roughly 4pp of margin; and CXMT's periods carry ramp underloading. The honest read of the direction is that the >30% figure stopped being single-sourced tonight, and the appropriate posture is that the gap was intact as of mid-2025 with no public evidence of closure since.
What to change in the instrumentation. Three reversible edits, no posture change to [[2026-05-18-memory-cycle-v1.1]]. First, retire Indicator 1 of the parent brief as written. Its crossing condition, Q2's margin gap persisting through Q3, has now technically fired at 89.53%, and this brief shows that firing carries no information about cost, because the condition tests exposure rather than cost. Replace it with a mix-and-period-matched comparison, CXMT total gross margin against Micron's Mobile and Client unit margin plus about 3.3pp, on the nearest fiscal quarter, with a note that the residual is a price-versus-cost composite and not a cost reading. Second, re-key the cost-gap refresh to down-quarters. The parent brief's Indicator 2, margin beta in the first quarter of falling contract prices, is the right instrument and this brief explains why: margin differences regain discriminating power as margins fall, and the pre-boom comparison above is the proof of concept. Add the mechanical correction that CXMT's Chinese-standards margin excludes write-downs that Micron's includes, so any down-cycle comparison must add CXMT's asset-impairment loss back into its cost of goods sold before comparing. Third, mark Indicator 4 partly satisfied. The >30% estimate now has a second, independent, filing-based corroboration for mid-2025, so the open item narrows to whether the gap moved during 2026, which no public source addresses.
Answering the second half separately: no, the prospectus does not permit a direct cost-per-bit estimate, and the gap is specific. The binding absence is volume. The filing draft discloses no bit shipments, no wafer output, no wafer starts, and no capacity utilisation rate, and it gives no absolute per-unit cost or per-unit price, only a year-on-year percentage change in the unit selling price of major DRAM products. Without a bit or wafer denominator, no cost-per-bit level is recoverable by any route. Three weaker things are recoverable. A relative unit-cost index follows from combining gross margin with the disclosed price change: unit cost fell about 44% in 2023 and rose about 44% in 2024 (Ray derivation), but "per unit" means per piece and CXMT was migrating from DDR4 to denser DDR5 and LPDDR5 over exactly that window, so mix migration contaminates it and it is emphatically not per-bit. Inventory days are recoverable from the disclosed turnover ratio: roughly 248, 223 and 257 days for 2022, 2023 and 2024, with H1 2025's 1.19 ratio reading as about 153 days if it is a half-year figure and about 307 if annualised, an ambiguity the document does not resolve (Ray derivation; compare the Micron days-inventory-outstanding series in [[2026-09-07-memory-supplier-inventory-days-dio-anchor]]). And the margin-residual cost gap above, which is the most useful thing the document supports. One caveat on the whole answer: this is the 2025-12-29 filing draft read through an unofficial translation, covering only through H1 2025. A final registration prospectus or a listed-company annual report would carry updated periods and may add a cost-composition table; neither was reached in this run.
Why this is in the vault
It closes two of the four open follow-ups in [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]] and changes what the CXMT cost leg is instrumented with. Concretely: Indicator 1 of that brief's monitoring table should be retired rather than marked as fired, the ">30% cost gap is single-sourced" caveat carried through four briefs in the CXMT lineage can be downgraded, and any future CXMT-versus-peer margin comparison in the memory-cycle thesis needs the write-down add-back correction documented here before it is used.
Open follow-ups
- How does CXMT account for government grants, and how large are they? The filing draft discloses no grant policy and no amounts. Whether grants are netted against asset carrying values, which lowers depreciation inside cost of goods sold, or routed to other income below gross profit, is the single largest unresolved confounder on CXMT's reported gross margin. A listed-company annual report or the final registration prospectus should state the policy and the annual figure.
- What useful lives does CXMT assign to machinery and buildings? The implied 12.5-14.0% depreciation rate on net fixed assets suggests lives longer than Micron's five-to-seven-year equipment lives, which would flatter CXMT's cost of goods sold structurally and permanently. The policy section of the filing draft refers to useful lives without stating years.
- Does Micron publish a DRAM-versus-NAND margin or revenue split anywhere? The two independent routes to Micron DRAM-only gross margin in this brief both rest on an assumed NAND weight and margin. A segment footnote, 10-K disclosure, or earnings-call statement giving the split would convert the central mix finding from derivation to disclosure.
- Is the final CXMT registration prospectus or its first annual report reachable, with periods through 2025 or H1 2026? That document would extend the margin-residual cost-gap series past mid-2025, which is the one period where this brief can say nothing and where a real cost-gap closure would have to show up.
- Did CXMT's Q2 or Q3 2026 reports disclose capacity utilisation, output volume, or an operating-expense ratio? The sibling brief argues utilisation is the observable for the "does CXMT cut production" term that should carry more weight in the Phase-3 severity modifier than the cost gap. The filing draft has none of it; the listed-company reports may.
- Does the margin-residual method reproduce the cost gap for SK hynix and Samsung in 2024 and H1 2025? The method was run only against Micron here. Running it against the two lower-cost producers would test whether CXMT's gap to the industry's cost leader is wider than 30%, which is the number that matters for trough depth.
Related
- [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]] - parent brief; this closes its first and fourth follow-ups and retires its Indicator 1
- [[2026-11-04-dram-2023-trough-cash-cost-floor-cxmt-position]] - sibling, same night; source of the cash-cost method and the write-down-contamination warning reused here
- [[2026-07-03-cxmt-dram-capacity-memory-oligopoly-risk]] - origin of the >30% gap and the 5-10% price discount
- [[2026-09-08-cxmt-ymtc-bit-share-phase3-threshold]] - share leg; flat bit-share-to-capacity ratio of about 0.7 through 2027
- [[2026-07-07-dram-hbm-phase2-phase3-early-signals]] - tripwire table carrying row 5; conventional-DRAM contract-price path
- [[2026-11-01-hbm-fixed-price-offtake-primary-filings]] - HBM long-term-agreement pricing is a floor-and-ceiling band, which caps the mix mechanism
- [[2026-09-07-memory-supplier-inventory-days-dio-anchor]] - Micron days-inventory-outstanding series for the inventory-days comparison
- [[2026-05-18-memory-cycle-v1.1]] - executable thesis; posture unchanged by this brief
- [[2026-06-28-chip-memory-cycle-phase2-phase3-indicators]] - Q1-Q2 2026 conventional DRAM contract-price moves
Sources
Vault
- [[2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire]] -
~/rdco-vault/06-reference/research/2026-09-12-cxmt-ddr5-cost-per-bit-yield-parity-tripwire.md - [[2026-11-04-dram-2023-trough-cash-cost-floor-cxmt-position]] -
~/rdco-vault/06-reference/research/2026-11-04-dram-2023-trough-cash-cost-floor-cxmt-position.md - [[2026-07-03-cxmt-dram-capacity-memory-oligopoly-risk]] -
~/rdco-vault/06-reference/research/2026-07-03-cxmt-dram-capacity-memory-oligopoly-risk.md - [[2026-09-08-cxmt-ymtc-bit-share-phase3-threshold]] -
~/rdco-vault/06-reference/research/2026-09-08-cxmt-ymtc-bit-share-phase3-threshold.md - [[2026-07-07-dram-hbm-phase2-phase3-early-signals]] -
~/rdco-vault/06-reference/research/2026-07-07-dram-hbm-phase2-phase3-early-signals.md - [[2026-11-01-hbm-fixed-price-offtake-primary-filings]] -
~/rdco-vault/06-reference/research/2026-11-01-hbm-fixed-price-offtake-primary-filings.md - [[2026-09-07-memory-supplier-inventory-days-dio-anchor]] -
~/rdco-vault/06-reference/research/2026-09-07-memory-supplier-inventory-days-dio-anchor.md - [[2026-05-18-memory-cycle-v1.1]] -
~/rdco-vault/01-projects/investing/theses/2026-05-18-memory-cycle-v1.1.md - [[2026-06-28-chip-memory-cycle-phase2-phase3-indicators]] -
~/rdco-vault/06-reference/research/2026-06-28-chip-memory-cycle-phase2-phase3-indicators.md
Web — hard disclosure
- Micron Technology FQ3 FY2026 Form 8-K, Exhibit 99.1 press release (read; revenue, GAAP and non-GAAP gross margin, business-unit revenue and operating margin, FQ4 guidance, nine-month depreciation)
- CXMT STAR Market IPO prospectus filing draft, 2025-12-29, full English translation (read in full via direct download and term-swept; unofficial translation, translator not stated — treat as a lower evidence tier than the filing itself; periods 2022 through H1 2025)
Web — secondary
- CXMT reports $8.16 billion Q3 2026 profit - Gizmochina, Nov 2 2026 (search extract; Q3 revenue, profit, capacity plan)
- CXMT Q3 Revenue Rises 613% as DRAM Shortage Swings It to Profit - Unite.AI (search extract)
- CXMT Hits 89.53% Gross Margin - Techtroduce (search extract; the +1.94pp sequential figure)
- SK hynix Reports Record Q2 2026 Revenue Of 79.3 Trillion Won - Pulse2 (search extract)
- SK hynix Q2 2026 results coverage - Khan (search extract; 76% operating margin)
- CXMT revenue set for sharp acceleration as memory prices surge - S&P Global Market Intelligence, Oct 2026 (search extract; forecast, 2026 gross margin 85.7% — an analyst projection, not a reported figure)
Flagged, not retried
- SK hynix newsroom Q2 2026 release: two candidate URLs under
news.skhynix.comreturned HTTP 404. The 83.2% gross-margin figure is therefore corroborated only arithmetically from hard-disclosed revenue and operating profit, not from SK hynix's own gross-profit line. - CXMT Q2 and Q3 2026 filings on the Shanghai Stock Exchange were not reached in primary form; all 2026 CXMT figures in this brief are secondary coverage, with the Q2 figure double-sourced via the Q3 report's own sequential comparison.