Sandisk ($1,877) Breaks a 3-Month Base at 8.8x Forward Earnings — $2,354 High Back in View

Sandisk (NASDAQ: SNDK) jumped 6.29% by mid-morning Tuesday, September 22, trading near $1,877 after closing Monday at $1,766.64. Monday's session ran from $1,737.01 to $1,834.49, ending down 1.4% from Friday's $1,791.82 close. Tuesday's move added roughly $16 billion of market value and lifted the company toward $275 billion, based on 146.42 million shares outstanding.

The catalyst was a new buy-rated initiation with a $2,400 price target, implying 36% upside from Monday's close. The argument behind it is that new AI compute platforms are repositioning NAND flash from a commodity storage medium into a system-critical component of AI infrastructure. As model sizes expand and inference workloads consume more data, buyers are prioritizing density, performance, endurance and supply certainty over the lowest price. That is a structural change in how the NAND market clears.

The thesis for this forecast follows from a single number. At Monday's close, Sandisk traded at 8.24 times next-twelve-month earnings. At Tuesday's $1,877, that multiple is 8.76. For a company that just posted an 84.6% non-GAAP gross margin, guided to $10.3 billion to $10.8 billion of quarterly revenue and signed ten business agreements worth a minimum of $93.9 billion with more than four years of visibility, single-digit forward earnings is the cheapest expression of the AI trade in the market. The stock trades at 23.97 times trailing earnings on TTM EPS of $73.70 and revenue of $20.248 billion, which shows how fast estimates have moved.

The market does not fully believe those estimates. That is the other half of the story. After the fiscal fourth-quarter report, Sandisk beat on both lines and the stock fell 12.1% after hours. It remains 20.3% below its 52-week high of $2,354.39, set on June 22, 2026. Insiders have sold $121.37 million of stock over the past three months.

The scale of the run explains the caution. The 52-week low is $93.54, set on September 25, 2025. Sandisk has gained more than 600% in 2026. It was added to the S&P 100 on September 21.

The forecast bias is bullish with defined risk. Support sits at $1,737 and $1,614. Resistance runs $1,834.49, then $2,000, then the $2,354.39 high.

Tuesday's Tape: A 6.29% Gap From a Pullback Base

The session's structure matters because it shows buyers stepping in after a two-week consolidation.

Sandisk had been pulling back before Tuesday. On September 20, the stock traded between $1,614.00 and $1,798.45, closing near the high at $1,797.98 with volume of 17.79 million shares against a 10.56 million daily average. That $184 intraday range is a 11.4% swing, which is normal volatility for this name. On Friday, September 18, the stock closed at $1,791.82. On Monday, it fell 1.4% to $1,766.64 after trading as high as $1,834.49 and as low as $1,737.01.

That Monday reversal from $1,834.49 to $1,766.64 is the key reference. It shows sellers active near $1,830, which makes Tuesday's break above that level the test to watch.

Tuesday's move began before the bell. Premarket trading showed the stock at $1,828.75 on 607,140 shares after the initiation crossed. By mid-morning, the gain had extended to 6.29%, putting the stock near $1,877 and above Monday's intraday high.

The broader tape supported the move. The Nasdaq Composite traded up 0.37% after closing at a record 27,122.09 on Monday, and a benchmark of semiconductor stocks rose for a sixth straight session, its longest winning streak since April. AMD crossed $1 trillion in market value on Monday after rising 9.95%, and Meta Platforms gained 11.34% as its Muse AI agent topped App Store download charts. That adoption evidence feeds directly into the storage argument: every autonomous agent session consumes far more tokens, and far more stored data, than a single chatbot query.

Storage peers moved with Sandisk. Western Digital rose 3% and Micron ticked higher in a recent session when the storage bid concentrated in Sandisk. That concentration matters. When capital wants NAND exposure, Sandisk is the purest listed vehicle, because it produces substantially all of its flash chips through a joint-venture framework with Kioxia across manufacturing sites in Japan.

The S&P 100 addition on September 21 added a technical bid. Index inclusion forces benchmark-tracking funds to buy, and it broadens the institutional ownership base.

The first test for traders is whether SNDK holds above $1,834.49, Monday's high. The second is $1,900, the next round number. A close back below $1,766.64 would mean the initiation move was a one-day spike.

The Q4 Numbers: $8.97 Billion in Revenue and an 84.6% Gross Margin

The fiscal fourth-quarter report is the fundamental base for the entire rerating, and the numbers are extraordinary.

Sandisk reported non-GAAP earnings per share of $39.25, beating estimates by 13.47%. Revenue came in at $8.97 billion, topping forecasts by 6.53%. Non-GAAP gross margin reached 84.6%, against 26.4% a year earlier. That is a 58.2 percentage point expansion in twelve months, which is close to unprecedented for a hardware manufacturer.

Data center revenue was the engine. It surged 103% sequentially to $2.98 billion, driven by AI demand. Sequential doubling in a single quarter shows how fast hyperscaler storage orders are arriving. In an earlier quarter, the data center segment grew 645% year over year while total revenue more than tripled to $5.95 billion, with adjusted gross margins at 78.4%.

The consumer business went the other way, falling 32% to $556 million. That divergence defines the company now. Sandisk is no longer a consumer flash drive and memory card business with a data center side. It is an AI storage supplier with a shrinking consumer legacy.

The guidance was stronger than the quarter. For the first quarter of fiscal 2027, Sandisk guided revenue of $10.3 billion to $10.8 billion and EPS of $44 to $46. At the midpoints, that is $10.55 billion and $45, sequential growth of 17.6% in revenue and 14.6% in earnings.

The contract disclosure was the most important item. Sandisk signed ten new business agreements worth a minimum of $93.9 billion with more than four years of visibility. For a company with trailing revenue of $20.248 billion, that is nearly five years of current revenue locked into contracts. Management expects these agreements to account for more than half of its bits in fiscal 2027.

Capital returns are running alongside. Sandisk repurchased $4.5 billion of stock in the fourth quarter and has $15.5 billion remaining under its authorization. At the current market value, that remaining authorization equals 5.6% of the company.

The stock fell 12.1% after hours despite the beat. That reaction is the single most important behavioral data point in this forecast. It tells you the market is not paying for peak-cycle earnings at face value.

The $93.9 Billion Contract Book: Why This Cycle Looks Different

The contracted revenue is the core of the bull case, and it is what separates this cycle from previous NAND cycles.

Historically, NAND has been the most brutally cyclical segment in semiconductors. Suppliers competed on price, capacity expanded during booms, oversupply arrived, prices collapsed and margins went negative. Sandisk's own history shows it: non-GAAP gross margin was 26.4% a year before hitting 84.6%.

The ten agreements worth a minimum of $93.9 billion, with more than four years of visibility, change that structure. Instead of selling bits into a spot market at whatever price clears, Sandisk has committed volumes at predetermined terms with named customers. Management expects these agreements to cover more than 50% of its bits in fiscal 2027. That leaves less than half of output exposed to spot pricing, which dampens the downside if the cycle turns.

The pricing backdrop supports the margins. One industry forecast has NAND average selling prices rising 186% in 2026 and enterprise solid-state drive prices rising 265%. Sandisk estimates NAND industry revenue will exceed $300 billion in calendar 2026 and approach $500 billion in 2027. That is an industry growing at roughly 67% in a single year.

The technology is the differentiator. Sandisk's BiCS8 and BiCS10 platforms deliver density gains using fewer 3D layers than competitors. Fewer layers means lower manufacturing cost per bit and better yields. The company operates under a 25-year manufacturing partnership with Kioxia, producing substantially all of its flash at joint-venture sites in Japan. That vertical integration gives it control over supply that pure-play designers lack.

The August 13 investor day set long-term targets: double-digit revenue growth through 2030 and an 80% non-GAAP gross margin. Holding 80% gross margin for four years would be remarkable for any hardware business.

Capacity expansion backs the demand view. Sandisk committed to a $31 billion NAND expansion, and it is raising capital expenditure for BiCS8 and BiCS10 production.

The risk is embedded in the same numbers. A $31 billion expansion, alongside similar moves from competitors, is how oversupply gets built. The industry is adding capacity at the top of a pricing cycle. The contracts protect Sandisk for four years. They do not protect the industry.

Valuation: 8.8 Times Forward Earnings Against 24 Times Trailing

The valuation gap is the most striking feature of this stock, and it is where the bull and bear cases collide.

At Monday's close of $1,766.64, Sandisk traded at 8.24 times next-twelve-month earnings. That implies NTM EPS near $214. At Tuesday's $1,877, the multiple is 8.76. The trailing P/E is 23.97 on TTM EPS of $73.70. Trailing revenue is $20.248 billion, TTM EBITDA is $12.617 billion, gross margin is 71.47%, net margin is 56.42% and return on equity is 91.56%.

Work through the forward math. First-quarter fiscal 2027 guidance calls for EPS of $44 to $46. Annualizing the $45 midpoint gives $180. At $1,877, that is 10.4 times. If the sequential growth continues at even a fraction of the recent pace, the implied multiple falls further. That is how the NTM figure reaches $214.

A single-digit forward multiple on a company with 84.6% gross margins, $93.9 billion of contracted business and a $15.5 billion buyback authorization is either a mispricing or a warning. The market is telling you it expects those earnings to prove temporary.

The historical pattern justifies the skepticism. NAND cycles peak and collapse. When they collapse, peak earnings become losses, and a stock that looked like 8 times earnings turns out to have been 30 times mid-cycle earnings. The 12.1% after-hours drop on a beat-and-raise quarter is the market pricing that risk.

The counterargument is the contract structure. If more than half of fiscal 2027 bits are sold under multi-year agreements, the earnings floor is higher than in previous cycles. That would justify a higher multiple, not a lower one.

Street targets show how wide the disagreement runs. The consensus 12-month target sits near $2,125, with 20 buy recommendations against one sell. Another consensus reading places it at $2,195 from 15 buys and two holds, and a third at $1,912. The high estimate is $3,600 and the low is $1,000. One target was raised to $2,500 from $2,025. Tuesday's initiation came in at $2,400.

A target range from $1,000 to $3,600 on the same stock is a 260% spread. That is not a disagreement about execution. It is a disagreement about whether this is a cycle or a structural change.

The AI Storage Thesis: Density, Endurance and Supply Certainty

The argument for a structural change rests on how AI workloads consume storage differently from traditional computing.

The case laid out in Tuesday's initiation is that expanding model sizes and data-intensive inference operations prioritize density, performance, endurance and supply certainty over lowest price. That is a shift in the buying criteria. When the buyer is a PC manufacturer sourcing a laptop SSD, price per gigabyte is the deciding factor, and suppliers compete it away. When the buyer is a hyperscaler building an AI cluster, the storage tier has to feed accelerators at high throughput without becoming the bottleneck, and supply has to be guaranteed for years to match the data center build schedule.

The demand mechanics support it. Training a frontier model requires enormous datasets held in fast storage. Inference at scale requires retaining model weights, context and retrieved data close to the compute. Agentic AI, the category Meta's Muse launch just validated with consumer-scale adoption, multiplies the token count per user session by orders of magnitude. Each of those tokens touches memory and storage.

The competitive framing has shifted from price wars toward density, performance, durability and supply stability. That favors suppliers with technology leadership and controlled capacity. Sandisk's BiCS8 and BiCS10 achieve density gains with fewer 3D layers, and its Kioxia joint venture gives it dedicated fabs.

The financial evidence is in the segment split. Data center revenue rose 103% sequentially to $2.98 billion while consumer fell 32% to $556 million. Data center is now more than five times the consumer business. A year earlier, that ratio was far lower.

The industry forecast is aggressive: NAND revenue above $300 billion in calendar 2026 and approaching $500 billion in 2027. For context, the entire memory industry, including DRAM, was a fraction of that a few years ago.

The skeptical view is equally clear and has been stated plainly: Sandisk sells commodity-like NAND flash with little differentiation or pricing power, which exposes it to volatile cyclicality. Under that view, tight supply is temporary and current margins will normalize.

The resolution comes down to whether AI demand growth stays rapid enough to keep NAND supply tight through 2030. That is the explicit bull scenario. It is also unknowable today, which is why the target range spans $1,000 to $3,600.

Risks: Insider Selling, CXMT and a 12.1% Post-Earnings Drop

The bearish case has specific, current evidence behind it, and it deserves the same scrutiny as the bull case.

The first risk is the market's own reaction to good news. Sandisk beat on revenue and earnings in its fiscal fourth quarter, guided first-quarter revenue to $10.3 billion to $10.8 billion and disclosed $93.9 billion of contracts. The stock fell 12.1% after hours. When a stock cannot rally on a beat of that magnitude, positioning is stretched and expectations have run ahead of results.

The second is insider selling. Insiders sold $121.37 million of stock over the past three months. Executives selling into strength after a 600% run is not unusual, and much of it may be scheduled. But the scale is meaningful relative to the float.

The third is new competition. China's CXMT is preparing to enter the NAND flash market with a dedicated research and development effort focused on AI customers. Chinese memory entrants have historically driven prices down aggressively as they scale, backed by state support and less concern for returns. A credible Chinese NAND supplier targeting the same AI customers is the single largest structural threat to the pricing that supports an 84.6% gross margin.

The fourth is industry capacity. Sandisk committed $31 billion to NAND expansion, and it is raising capital expenditure for BiCS8 and BiCS10. Competitors are doing the same. The lesson of every memory cycle is that capacity added at the top arrives just as demand growth slows.

The fifth is concentration. Data center revenue of $2.98 billion comes from a small number of hyperscale buyers whose capital expenditure plans drive the entire AI trade. If one large customer reduces orders, the revenue impact is immediate.

The sixth is macro. The Federal Reserve raised rates to 3.75%–4.00% on September 16 and markets price a 90% chance of another hike in December. The 10-year Treasury yield sits at 4.96%. High-multiple technology stocks are sensitive to rates, though at 8.8 times forward earnings Sandisk is less exposed on that measure than most AI names.

The seventh is volatility itself. The stock has swung from $93.54 to $2,354.39 within 52 weeks and fell more than 15% over two sessions during an earlier AI selloff. It dropped more than 7% in one session on reports that accelerator prices would rise, which the market read as a squeeze on customer memory budgets.

Technical Picture: A Consolidation Below the $2,354.39 High

The chart shows a stock in a strong uptrend that has been consolidating since June, and Tuesday's move is an attempt to break that consolidation.

The 52-week high of $2,354.39 was set on June 22, 2026. From that peak, Sandisk fell to the $1,600s, a drawdown of roughly 31%. It has since been building a base between $1,600 and $1,850. On September 20, it traded from $1,614.00 to $1,798.45. On Monday, September 21, it ranged from $1,737.01 to $1,834.49 before closing at $1,766.64.

That base is three months old and its boundaries are clear. The floor is $1,614, the recent low. The ceiling is $1,834.49, Monday's high, with stronger resistance near $1,900.

Tuesday's 6.29% move to approximately $1,877 breaks above the top of that range intraday. A daily close above $1,834.49 would confirm the breakout. A close above $1,900 would open the path toward $2,000 and then the June high at $2,354.39.

Volume is the confirmation to watch. The 10-day average volume is 10.56 million shares, and recent sessions have run above 17 million. High-volume breakouts hold more often than low-volume ones.

The momentum picture is mixed on longer readings. RSI has been oscillating in the 39 to 47 range during the consolidation, which is neutral-to-weak rather than overbought. That is constructive for a breakout attempt: it means momentum is not extended, so there is room to run before the stock becomes technically stretched.

The longer-term trend is unambiguous. From $93.54 on September 25, 2025, to $1,877 on September 22, 2026, the stock has risen 1,907%. Every pullback in that sequence has been bought.

Short interest is a factor. Recent readings have shown short interest above 7% of the float and in some periods above 11%. A stock with elevated short interest breaking out of a three-month base on a bullish initiation can squeeze hard.

The technical rule: above $1,834.49 on a daily close, the base breaks and $2,000 is the next target. Below $1,737, the breakout fails. Below $1,614, the three-month base breaks down and the uptrend is in question.

Key Resistance: $1,834.49, $1,900, $2,000 and the $2,354.39 Record

The upside map is defined by the consolidation's boundaries and the June peak.

The first resistance is $1,834.49, Monday's intraday high. Sellers reversed the stock from that level, taking it down to a $1,766.64 close. Clearing it on a daily close is the first confirmation.

The second is $1,900, a round number just above Tuesday's intraday range. It also sits near the $1,912 low-end consensus target, which means some holders will take profit there.

The third is $2,000, both a psychological level and the threshold referenced in the bullish argument that Sandisk could punch well above $2,000 per share as its memory becomes critical to AI models. From $1,877, reaching $2,000 is a 6.6% gain.

The fourth is the $2,125–$2,195 consensus target zone. Multiple consensus readings cluster there. That is 13.2% to 17.0% above the current price and represents where the average Street view sits.

The fifth is $2,354.39, the 52-week high from June 22, 2026. Reaching it requires a 25.4% gain. A break above that level would put the stock at an all-time high with no overhead supply.

Above that, the $2,400 initiation target and the $2,500 raised target define the next zone. The high estimate on the Street is $3,600, which would require a 92% gain and a full acceptance of the structural AI storage thesis.

The conditions for each level are specific. Clearing $1,834.49 and $1,900 requires only that Tuesday's momentum holds through the close on strong volume. Reaching $2,000 and the consensus zone would likely require confirmation from the November 4 fiscal first-quarter report, where revenue needs to land at or above the $10.8 billion high end of guidance with gross margins holding near 84%.

Reaching $2,354.39 and beyond would require evidence that the contracted revenue model is durable: additional multi-year agreements, stable NAND pricing into 2027 and no visible impact from Chinese capacity.

The resistance rule: below $1,834.49, this is still a consolidation. Above $1,900 on a close, the base has broken and $2,000 is in play.

Key Support: $1,766.64, $1,737.01, $1,614 and the $1,500 Line

The downside map defines risk for anyone buying the breakout.

The first support is $1,834.49 flipped from resistance, if the stock closes above it. Until then, the first real support is $1,766.64, Monday's close. A return below that level would erase Tuesday's gain entirely.

The second is $1,737.01, Monday's low. It marks where buyers stepped in during the most recent session. A break below it would signal that the initiation-driven bid has been fully absorbed.

The third is $1,614.00, the September 20 low and the floor of the three-month consolidation. That is 14.0% below Tuesday's price. Holding $1,614 keeps the base intact and the uptrend alive. Breaking it would be the first meaningful technical damage since the June peak.

The fourth is $1,500, a round number below the base. Reaching it would represent a 20.1% decline from current levels and a 36.3% drawdown from the June high.

The buyback provides support at lower levels. Sandisk has $15.5 billion remaining under its authorization after repurchasing $4.5 billion in the fiscal fourth quarter alone. At $1,877, $15.5 billion would retire 8.3 million shares, 5.6% of the 146.42 million outstanding. A company buying back stock at that pace creates a persistent bid, particularly during pullbacks.

The balance sheet supports the buyback. Total cash stands at $3.74 billion, the debt-to-equity ratio is 0.01 and the company repaid $2.00 billion of long-term debt over the trailing twelve months. Operating cash flow is $4.64 billion on a trailing basis with capital expenditure of $179 million, though the $31 billion expansion commitment will change that profile substantially.

The catalysts for a decline are identifiable. Weak guidance on November 4 would be the most direct. Evidence that NAND spot prices are rolling over would be second. News that CXMT is scaling faster than expected would be third. A broad AI selloff, like the one that took the stock down more than 15% over two sessions earlier this year, would be fourth.

The support rule: above $1,737, the breakout attempt is alive. Between $1,614 and $1,737, the base is under review. Below $1,614, the three-month base breaks and $1,500 becomes the target.

The November 4 Report: The Number That Settles the Debate

The next scheduled catalyst is the fiscal first-quarter report, estimated for November 4, and it will determine whether this stock trades at 8 times forward earnings or 20.

The bar is set by guidance: revenue of $10.3 billion to $10.8 billion and EPS of $44 to $46. Delivering at the high end would mean sequential revenue growth of 20.4% from the fourth quarter's $8.97 billion. Anything at or above the midpoint of $10.55 billion with gross margin holding near 84% would validate the current trajectory.

Four specific line items matter more than the headline.

The first is data center revenue. It rose 103% sequentially to $2.98 billion in the fourth quarter. Another sequential gain above 50% would show that hyperscaler orders are accelerating rather than front-loading. A flat or declining figure would be the clearest warning that the boom is peaking.

The second is gross margin. At 84.6% non-GAAP, margins are at a level that invites competition. The investor day target is 80% long term, which implies management expects some compression. A print above 84% would beat that trajectory. A print in the 70s would raise questions.

The third is the contract book. Sandisk disclosed ten agreements worth a minimum of $93.9 billion with over four years of visibility. New agreements announced alongside the quarter would extend the visibility and strengthen the structural argument. No new agreements would suggest the initial batch captured most available demand.

The fourth is the bit mix under contract. Management expects the agreements to cover more than 50% of bits in fiscal 2027. Confirmation of that figure, with pricing terms, would let investors model the earnings floor.

Capital returns are the fifth item. With $15.5 billion remaining on the authorization after $4.5 billion of repurchases in one quarter, the pace of buybacks signals management's own view of value.

Guidance for the fiscal second quarter will move the stock more than the reported quarter. The market is pricing in continued growth. Guidance that implies a plateau would compress the multiple quickly, as the after-hours reaction to the last beat demonstrated.

Between now and then, NAND spot pricing data and any news on Chinese capacity are the interim signals.

Price Targets: $2,000 Near Term, $2,354 on Confirmation, $1,614 as the Line

The forecast breaks into three scenarios with specific conditions.

The near-term base case, over the next two to four weeks, targets $2,000. It requires a daily close above $1,834.49 on volume above the 10.56 million average, followed by a push through $1,900. From $1,877, reaching $2,000 is a 6.6% gain. The conditions are the semiconductor complex holding its winning streak, the Nasdaq staying near records and no negative NAND pricing news. The elevated short interest and the S&P 100 index bid support this path.

The medium-term bull case targets the $2,125–$2,195 consensus zone and then the $2,354.39 52-week high, a 25.4% gain. It requires the November 4 report to deliver revenue at or above $10.8 billion, gross margin at or above 84% and data center revenue growing sequentially. New multi-year agreements added to the $93.9 billion book would accelerate it. In that scenario, the $2,400 initiation target becomes the next reference, and the stock would trade at roughly 11 times the NTM earnings implied by the current forward multiple.

The bear case targets $1,614 and then $1,500. It would be triggered by a guidance disappointment on November 4, evidence of NAND spot price weakness, faster-than-expected Chinese capacity additions or a broad AI selloff. The after-hours 12.1% drop on the last beat shows how quickly this stock can reprice. A break below $1,614 would end the three-month base.

The risk-reward favors the long side with discipline. From $1,877, the upside to $2,000 is $123 and to $2,354.39 is $477. The downside to $1,737 is $140, and to $1,614 is $263. A position with a stop below $1,737 targeting $2,354.39 offers a ratio of 3.4 to 1. Entering on a pullback toward $1,800 with the same stop would improve it further.

Position sizing should account for volatility. The stock moved 11.4% intraday on September 20 and has fallen more than 15% over two sessions in prior selloffs. A position sized for a normal large-cap will be too large here.

Level summary: support at $1,834.49 once flipped, $1,766.64, $1,737.01, $1,614 and $1,500. Resistance at $1,834.49, $1,900, $2,000, $2,125–$2,195 and $2,354.39.

Verdict: Bullish, Buy the Breakout Above $1,834, Stop Below $1,737

The verdict on SNDK near $1,877 is bullish, with a valuation argument that is difficult to dismiss and risks that are equally concrete.

The bullish case is the combination of growth and price. Sandisk delivered fiscal fourth-quarter revenue of $8.97 billion and non-GAAP EPS of $39.25, both above estimates, with a non-GAAP gross margin of 84.6% against 26.4% a year earlier. Data center revenue rose 103% sequentially to $2.98 billion. It guided the current quarter to $10.3 billion to $10.8 billion in revenue and $44 to $46 in EPS. It signed ten agreements worth at least $93.9 billion with more than four years of visibility, expected to cover more than half of fiscal 2027 bits. It has $15.5 billion left on its buyback after repurchasing $4.5 billion in one quarter, cash of $3.74 billion and a debt-to-equity ratio of 0.01. It targets double-digit revenue growth through 2030 and an 80% gross margin. And it trades at 8.76 times next-twelve-month earnings.

The bearish case is the cycle. NAND has destroyed capital in every prior cycle, and the market clearly remembers: the stock fell 12.1% after hours on a beat-and-raise quarter. Insiders sold $121.37 million in three months. China's CXMT is building dedicated NAND capacity aimed at AI customers. Sandisk itself is adding $31 billion of capacity into a pricing peak. Street targets range from $1,000 to $3,600, which shows how little consensus exists on whether these earnings are durable.

The trading plan follows. Buy a daily close above $1,834.49 with a stop below $1,737.01. Target $1,900 first, then $2,000, then the $2,125–$2,195 consensus zone. A close above $2,000 makes $2,354.39, the June 22 high, the objective. A daily close below $1,737 invalidates the breakout, and a close below $1,614 breaks the three-month base and targets $1,500.

The November 4 fiscal first-quarter report is the event that resolves the debate. Data center revenue growth, gross margin near 84% and new contract announcements would support a move toward $2,354.39. Any sign of a plateau would compress the multiple fast.

Verdict: bullish. Near-term target $2,000. Confirmation target $2,354.39. Stop below $1,737. Invalidation below $1,614.

That's TradingNEWS