Wealth Engine

Tag: DRAM Prices

  • Micron vs SanDisk: 7x vs 9.6x After MU’s $54B Quarter

    Micron vs SanDisk comes down to one number: price per dollar of forward earnings. After Micron’s record $54.23 billion fiscal Q4 print on September 30, 2026, MU trades at roughly 7.0x annualized guided earnings versus about 9.6x for SanDisk, using September 30 closes. Micron is the cheaper memory stock on that math — but it is also the one committing $25 billion of capex in six months.

    Key takeaways

    • Micron’s fiscal Q4 revenue hit $54.23 billion, up 379% year over year.
    • A $10,000 stake buys $1,433 of forward earnings in MU, $1,041 in SNDK.
    • Micron guided fiscal 2027 first-half capital expenditures to about $25 billion.

    What happened on September 30, 2026?

    Micron reported fiscal fourth-quarter 2026 results after the close on September 30, 2026. Revenue was $54.23 billion, up 379% year over year and 31% sequentially. Non-GAAP diluted earnings per share came in at $33.42. Gross margin reached 87%, above the roughly 86% the company had guided.

    Both lines beat the Street. LSEG consensus, as cited in coverage of the print, sat at $51.07 billion of revenue and $31.61 of adjusted EPS. That is a 6.2% revenue beat and a 5.7% earnings beat.

    The full year was the bigger number. Fiscal 2026 revenue reached $133.19 billion, up 256% from $37.38 billion in fiscal 2025, with non-GAAP diluted EPS of $75.52 — an 811% increase. DRAM revenue alone crossed $100 billion, at $100.68 billion.

    The after-hours reaction was muted rather than euphoric. Post-print coverage attributed the flat-to-lower tape not to the results but to the capital spending plan attached to them — the detail examined below.

    Which memory stock is cheaper, Micron or SanDisk?

    Micron is cheaper on forward earnings. Annualizing each company’s own next-quarter guidance and dividing by the September 30, 2026 close puts MU near 7.0x and SNDK near 9.6x. SanDisk carries roughly a 38% valuation premium on that measure, despite operating in the narrower of the two markets.

    The arithmetic is deliberately simple, because it uses only company guidance rather than analyst models. Micron guided fiscal Q1 2027 non-GAAP EPS to $38.15, plus or minus $1.00. SanDisk guided its fiscal Q1 2027 non-GAAP EPS to $44 to $46.

    The side-by-side numbers

    MetricMicron (MU)SanDisk (SNDK)
    Close, Sept 30, 2026$1,065.08$1,729.76
    Market capitalization~$1.20 trillion~$254.8 billion
    Latest quarter reportedSept 30, 2026Aug 5, 2026
    Latest quarterly revenue$54.23B (+379% YoY)$8.965B (+372% YoY)
    Latest non-GAAP EPS$33.42$39.25
    Fiscal 2026 revenue$133.19B$20.25B
    Fiscal 2026 non-GAAP EPS$75.52$70.88
    Next-quarter EPS guidance$38.15 ± $1.00$44–$46
    Forward P/E (annualized guide)~7.0x~9.6x
    Trailing P/E (FY26 non-GAAP)~14.1x~24.4x
    Latest gross margin87% (actual)83%–85% (guided)
    Product mixDRAM 75.6% of FY26 revenueNAND only
    Next reportDec 16, 2026 (forecast)Oct 29, 2026 (confirmed)
    Prices and market caps as of the September 30, 2026 close. Multiples are Wealth Engine calculations from company-guided figures.

    One caveat matters and is easy to miss. SanDisk’s trailing figures run through a fiscal year that ended in July 2026, while Micron’s run through August 2026. In a market where NAND pricing doubled inside a year, a two-month stale denominator flatters nobody consistently — it makes SNDK’s trailing 24.4x look worse than its current run-rate deserves.

    How much does the valuation gap cost you per $10,000?

    At a 7.0x forward multiple, every $10,000 committed to Micron buys $1,433 of annualized forward earnings. The same $10,000 in SanDisk at 9.6x buys $1,041. The difference is $392 per year per $10,000 — a 37.7% gap in earnings purchased per dollar invested.

    Here is the arithmetic, step by step, using the September 30, 2026 closes:

    1. Micron: guided EPS $38.15 × 4 = $152.60 annualized. $1,065.08 ÷ $152.60 = 6.98x. $10,000 ÷ 6.98 = $1,433 of forward earnings.
    2. SanDisk: guidance midpoint $45.00 × 4 = $180.00 annualized. $1,729.76 ÷ $180.00 = 9.61x. $10,000 ÷ 9.61 = $1,041 of forward earnings.
    3. Earnings yield: 14.3% for MU against 10.4% for SNDK, a 390-basis-point spread.
    4. Share count: $10,000 buys 9.389 MU shares or 5.781 SNDK shares, so fractional-share support is not optional at this position size.

    Annualizing one quarter is crude. It is also the only method that uses no forecast beyond what each company put in writing.

    Is Micron’s capex plan a reason the discount exists?

    Largely, yes. Micron told investors it plans to increase fiscal 2027 capital expenditures versus prior plans. Fiscal 2026 capex was $27 billion. First-half fiscal 2027 capex is guided to about $25 billion, with the second half expected to run higher than the first. Operating expenses are set to rise roughly $2.5 billion in fiscal 2027.

    That is close to a year of prior spending compressed into six months. Management framed the majority of the increase as construction capex to accelerate clean-room availability — capacity that arrives after the spending, not with it.

    This is the skeptical read on a 7.0x multiple: it is not obviously a bargain, it is a cyclical business being priced for the possibility that 87% gross margins do not survive the capacity it is currently funding. Micron itself guided fiscal Q1 2027 gross margin down to roughly 86.25%, calling it the low point for the year.

    SanDisk spends far less. Gross capex was 6.3% of revenue in its fiscal Q4 2026. A lighter capital model is a genuine argument for a premium multiple — it is simply not the argument most often made for the stock.

    Does DRAM or NAND exposure matter more here?

    It is the single biggest structural difference between the two. Micron sells both and is DRAM-weighted: DRAM was $100.68 billion of fiscal 2026 revenue, or 75.6% of the total, with NAND at $31.79 billion. SanDisk is pure NAND. Buying both is not diversification.

    In Micron’s fiscal Q4, DRAM contributed $39.77 billion, roughly 73% of quarterly revenue, with average selling prices up in the high-teens percentage range sequentially. Its Core Data Center unit generated $18.0 billion, up 56% sequentially at a 90% gross margin.

    NAND is not the weaker end of the trade. Micron’s data center SSD revenue alone approached $10 billion in the quarter, more than ten times the year-ago figure and over two-thirds of its total NAND revenue. SanDisk has forecast the NAND market at $300 billion in calendar 2026, rising to $500 billion in calendar 2027.

    Treat that $500 billion figure as a vendor estimate from a company with 11% NAND share as of the second quarter of 2026, not as an independent forecast.

    Which stock fits which investor profile?

    Neither is a low-volatility holding. The honest split is by what each investor is willing to underwrite: Micron asks you to accept a very large capital program, while SanDisk asks you to accept single-product concentration and a staler set of trailing numbers.

    Investor priorityBetter fit on the numbersWhy
    Lowest price per dollar of forward earningsMicron~7.0x vs ~9.6x on guided EPS
    Lowest capital intensitySanDiskGross capex 6.3% of FQ4 revenue
    Broadest memory exposureMicronDRAM and NAND; DRAM 75.6% of FY26 revenue
    Highest reported gross marginMicron87% actual vs 83%–85% guided
    Freshest disclosed financialsMicronReported Sept 30, 2026 vs Aug 5, 2026
    Largest, most liquid balance sheetMicron~$1.20T market cap vs ~$254.8B
    Nearest catalystSanDiskReports Oct 29, 2026
    Rankings reflect measurable attributes only, as of September 30, 2026. Not a recommendation.

    Sell-side targets exist on the Micron side and should be read as attributed opinion, not as fact. UBS maintained a Buy with a $1,625 target in a September 23, 2026 note. Baird’s Tristan Gerra raised his objective to $1,520 from $1,280 with a Buy rating, and JPMorgan held an Overweight with a $1,540 target ahead of the print.

    For related valuation work on non-semiconductor names, see our breakdowns of Costco vs BJ’s on earnings multiples, XLF vs KRE after the Fed’s hike, and dividend ETFs against Treasury yields.

    What to watch next

    • October 27–28, 2026: FOMC meeting, with the rate decision announced October 28 at 2:00 p.m. ET. No Summary of Economic Projections at this meeting.
    • October 29, 2026: SanDisk reports fiscal Q1 2027 after the close, with the call at 1:30 p.m. Pacific. This refreshes the stalest half of the comparison and tests the $44–$46 EPS guide.
    • December 8–9, 2026: FOMC meeting, decision December 9, including an updated Summary of Economic Projections and dot plot.
    • December 16, 2026 (forecast): Micron’s fiscal Q1 2027 report, covering the quarter ending December 3, 2026. Watch whether second-half capex lands above the $25 billion first-half figure.

    Frequently asked questions

    Is Micron cheaper than SanDisk right now?

    On forward earnings, yes. Using each company’s own next-quarter EPS guidance annualized against September 30, 2026 closes, Micron trades near 7.0x and SanDisk near 9.6x — a roughly 38% gap in Micron’s favor.

    Why did Micron stock not jump on a beat?

    Coverage of the September 30, 2026 print attributed the muted after-hours reaction to the capital spending outlook. Micron guided first-half fiscal 2027 capex to about $25 billion against $27 billion for all of fiscal 2026, and said the second half would be higher still.

    What is the difference between DRAM and NAND exposure?

    DRAM is working memory; NAND is persistent storage. Micron sells both, with DRAM at 75.6% of fiscal 2026 revenue. SanDisk sells NAND only, so its results track a single pricing cycle.

    How much of Micron’s business is data center?

    Its Core Data Center Business Unit was the largest contributor in fiscal Q4 2026 at $18.0 billion, up 56% sequentially, at a 90% gross margin. Data center SSD revenue was separately reported at nearly $10 billion for the quarter.

    Are these multiples comparable across the two companies?

    Only approximately. The fiscal calendars differ — SanDisk’s fiscal 2026 ended in July 2026, Micron’s in August 2026 — so trailing multiples are measured over different windows. Forward multiples built from current guidance are the closer comparison.

    Do share prices above $1,000 create a practical problem?

    They constrain position sizing. A $10,000 allocation is 9.389 shares of MU or 5.781 shares of SNDK at September 30, 2026 prices, so a broker without fractional-share support forces meaningful rounding.

    When do these two companies report next?

    SanDisk reports fiscal Q1 2027 on October 29, 2026, a confirmed date. Micron’s fiscal Q1 2027 report is forecast for December 16, 2026 and had not been formally confirmed as of September 30, 2026.

    The bottom line

    On the measurable question — price per dollar of forward earnings — Micron wins, and not narrowly. It is about 38% cheaper on annualized guided EPS, it reports the higher gross margin, it carries both DRAM and NAND exposure, and its numbers are two months fresher.

    What the discount depends on is specific, not vague: whether roughly $25 billion of first-half fiscal 2027 capex, plus a higher second half, arrives without collapsing the 87% gross margin that justifies the earnings base. Micron already guided that margin down to about 86.25% for fiscal Q1 2027.

    SanDisk’s premium is not indefensible — 6.3% of revenue in gross capex is a genuinely lighter model. But investors are paying 9.6x for one product line in a market whose $500 billion 2027 sizing comes from the vendor. The cheaper stock is also, on these disclosures, the better-diversified one.

    This article is journalism, not investment advice. Do your own research before investing.

    Sources

    Wealth Engine researches and drafts with AI tools and checks every figure against the sources above. How we report.

  • Nvidia AI Server Price Hike: 15% More, and Memory Is Why

    Nvidia is raising AI server prices by more than 15% on Grace Blackwell and Vera Rubin systems shipping in early 2027, Bloomberg reported on August 24, 2026. Memory is the reason. UBS puts memory at 62% of a Vera Rubin superchip’s $38,902 bill of materials, up from 53% on Grace Blackwell. TrendForce estimates the hike adds at least $5 billion to a 1-gigawatt data center.

    For three years the AI trade had one simple rule: Nvidia sets the price, and everyone pays it. That rule still holds. What changed is who Nvidia is paying.

    The Nvidia AI server price hike is not a margin grab. It is a pass-through. And the numbers underneath it say the memory makers, not the GPU designer, now control the cost curve of the AI build-out.

    How much is Nvidia raising AI server prices?

    More than 15% in many cases, effective on systems shipped early next year. Bloomberg reported the increases on August 24, citing people familiar with the matter. TrendForce, summarizing the same reporting, said some configurations could reach 17%. Nvidia did not respond to requests for comment.

    The warnings did not go to the cloud giants directly. According to Bloomberg, Nvidia notified the contract server manufacturers that assemble systems for Microsoft, Alphabet’s Google, and Oracle.

    That routing matters. The ODMs absorb the notice first, then reprice their own quotes. The cloud buyers find out when the invoice changes.

    Which systems are affected

    • Grace Blackwell systems — the current generation, still shipping in volume.
    • Vera Rubin systems — the next generation, with first shipments in early 2027.
    • Increases vary by chip generation and by memory configuration, per Bloomberg. Denser memory builds take the larger hit.

    Why is Nvidia raising prices now?

    Because memory has gone from a line item to the line item. Morgan Stanley estimates GPU silicon has fallen from more than 80% of AI server cost to roughly half that level in next-generation systems. The gap did not close because GPUs got cheaper. It closed because DRAM got expensive.

    A Vera Rubin NVL72 rack carries 74.7 TB of DRAM — 20.7 TB of HBM4 plus 54 TB of LPDDR5X, according to UBS’s teardown. That is the DRAM content of roughly 4,500 smartphones in a single rack.

    Every one of those bits is bought in the tightest memory market in a decade.

    What UBS found inside a Vera Rubin superchip

    UBS’s bill-of-materials analysis is the clearest picture available of where the money actually goes.

    Component Cost Share of superchip
    Total Vera Rubin superchip $38,902 100%
    All memory $24,297 62%
    SOCAMM2 LPDDR5X $19,355 49.8%
    HBM4 $4,943 12.7%
    Everything else $14,605 38%

    On Grace Blackwell, UBS put memory at 53% of cost. On Vera Rubin it is 62%, and the absolute memory bill rose about 2.5x between generations.

    One caveat worth holding onto: that 2.5x blends two different things. Vera Rubin carries more memory and pays more per gigabyte. It is not a pure price signal.

    How much has DRAM actually gone up?

    Steeply, and for longer than most forecasts allowed. TrendForce data cited by Tom’s Hardware shows conventional DRAM contract prices rising 90–95% quarter-over-quarter in Q1 2026 and a projected 58–63% in Q2 2026. Server DRAM is expected to climb every quarter through the second half of 2027.

    The consumer market tells the same story in plainer numbers. A mainstream 32GB DDR5-6000 kit runs about $392 today against $110–$140 a year ago, per Tom’s Hardware.

    Supply was committed early. SK hynix had sold out its entire 2026 production capacity by October 2025. Samsung and SK hynix raised 2026 HBM3E prices by roughly 20%.

    And HBM makes the squeeze worse mechanically: it consumes roughly four times the wafer area of conventional DRAM per bit shipped. Every HBM4 order crowds out ordinary server memory on the same fab.

    Who profits from the Nvidia AI server price hike?

    Not Nvidia, on the arithmetic. The memory suppliers capture the increase, the ODMs pass it through, and the hyperscalers eat it. Nvidia’s role here is closer to toll collector than beneficiary — and its own gross margin may be the quiet casualty.

    Work the math. Nvidia runs roughly a 75% gross margin, so the bill of materials is about 25% of the sale price. If memory is 62% of that BOM, memory is about 15.5% of the price. A 2.5x memory cost increase adds roughly 23 points of price to cost.

    A 15% price hike does not cover 23 points. Something has to give.

    Three readings, and the market has not settled on one:

    1. The 15% is an opening installment. More increases follow as 2027 contracts reprice.
    2. Nvidia is absorbing the difference. Gross margin drifts from ~75% toward the high 60s.
    3. The 2.5x is generational, not inflationary. Higher memory content is sold at a higher system ASP, so the comparison overstates the pass-through problem.

    Reading three is the most likely and the least discussed. It is also the one that would let Nvidia keep its margin story intact — which is precisely why it deserves scrutiny rather than acceptance.

    Why this matters for AI capex

    Because it reprices the entire build-out. TrendForce estimates the increase adds at least $5 billion to the cost of a 1-gigawatt AI data center. At the scale hyperscalers are now committing to, that is not a rounding error — it is a line in the capital plan that did not exist last quarter.

    The second-order effects are where this gets interesting.

    The uncomfortable version: AI compute has been getting cheaper per unit of intelligence for three straight years. This is the first credible input cost that pushes the other way.

    This post is reporting and analysis, not financial advice.

    Frequently asked questions

    How much is Nvidia raising AI server prices?

    More than 15% in many cases, with some configurations reaching 17% per TrendForce. Increases vary by chip generation and memory configuration.

    When do the new prices take effect?

    On systems shipped in early 2027, according to Bloomberg’s August 24, 2026 report.

    Which Nvidia systems are affected?

    Grace Blackwell and Vera Rubin server systems. Both are rack-scale platforms sold to cloud and enterprise data center operators.

    Why are AI server prices going up?

    Memory costs. UBS puts memory at 62% of a Vera Rubin superchip’s cost, and DRAM contract prices have risen every quarter through 2026 amid an HBM-driven supply squeeze.

    Who was notified about the price increases?

    Contract server manufacturers that build systems for Microsoft, Google, and Oracle, per Bloomberg. Nvidia did not comment publicly.

    How much does this add to a data center?

    TrendForce estimates at least $5 billion in additional cost for a 1-gigawatt AI data center.

    Does this hurt Nvidia’s margins?

    Possibly. Nvidia runs roughly a 75% gross margin. If memory costs rose 2.5x generationally, a 15% price increase may not fully offset it — though part of that increase reflects more memory content per system, not pure inflation.

    The bottom line

    The Nvidia AI server price hike is the clearest sign yet that the AI supply chain’s power center is shifting. For three years the scarce input was GPU wafer allocation. In 2027 it is memory, and the companies that own it — SK hynix, Samsung, Micron — are the ones setting terms.

    Watch two things next. First, whether Nvidia’s gross margin guidance holds through the fiscal year, because that is where the pass-through gap shows up. Second, whether any hyperscaler publicly revises a gigawatt commitment. The first cost-driven downgrade of an announced buildout would tell you the memory squeeze has stopped being an engineering problem and started being a financial one.

    Sources