Core Scientific: AMD Approval Could Be the Spark — Buy the Dip While Financing and Colocation Momentum Line Up

Hook and thesis

Core Scientific ($15.93) is presenting a classic “meltdown becomes opportunity” setup. Shares have retreated from the June highs near $30.46 to the current area, but several structural positives are lining up: a material shift from self-mining to higher-margin colocation, a fresh financing facility to fund expansion, and market reports that the company has secured AMD approval for hosting high-density AI workloads - a win that would materially improve demand visibility for its data-center capacity. Taken together, these developments argue for a core long position sized to your risk tolerance.

This is not a low-volatility, slam-dunk trade. Core Scientific still posts negative adjusted EBITDA and large negative free cash flow. But if you buy a proven infrastructure operator at a relatively compressed multiple amid operational reorientation, you get leverage to the AI/power-onfrastructure thematic without paying the froth that some peers command.

Business snapshot - why the market should care

Core Scientific operates purpose-built facilities for digital asset mining and a growing hosting/colocation business. The company is transitioning away from capital-intensive self-mining toward recurring, higher-margin hosting revenue. That shift is meaningful: colocation revenue jumped to $65 million in 2025 from $24 million in 2024 - a nearly threefold increase and proof that demand for leasable, high-density racks is real. With a goal to scale toward 1.5 gigawatts of leasable capacity, Core Scientific is positioning itself as a hybrid play - bitcoin mining tailwind optional, but hosting and AI infrastructure increasingly central.

Key fundamentals and valuation frame

At today's price ($15.93), market capitalization is roughly $5.12 billion and enterprise value about $7.65 billion. The company reports negative EPS (-$4.29) and sizeable negative free cash flow (-$957.7M), which explains the elevated EV/sales and EV/EBITDA optics: price_to_sales ~11.63 and ev_to_sales ~17.37 with ev_to_ebitda negative. These multiples read expensive on headline ratios, but the business is in transition: colocation revenue growth (2024 to 2025) shows tangible top-line progress that isn't yet fully reflected in operating leverage.

Put simply: headline multiples look rich because historical losses and negative cash flow remain. The bullish case is conditional - future valuation depends on (a) translating colocation growth into positive adjusted EBITDA, (b) stabilizing capital access, and (c) booking long-term contracts for AI/high-density hosting that meaningfully increase revenue visibility.

Why AMD approval matters (and the caveat)

Reports that Core Scientific has secured AMD's approval to host AMD-based GPUs for AI workloads - if confirmed in vendor documentation or mutually announced agreements - is a practical de-risker for the colocation story. Vendor certifications matter: customers seeking validated hardware stacks prefer facilities that pass vendor requirements. That approval would shorten sales cycles and allow Core Scientific to bid on a larger share of AI and high-performance computing workloads. I note the AMD approval point as a conditional but important catalyst - confirmation and subsequent customer bookings would materially increase revenue visibility.

Catalysts

• Contract wins and long-term colocation leases: Each announced multi-year lease for AI hosting would directly lift revenue visibility and improve adjusted EBITDA prospects.
• Vendor certifications and partnerships: Confirmation of AMD approval and additional OEM certifications would unlock new customer cohorts and accelerate bookings.
• Utilization ramp at new capacity and conversion of pipeline into signed deals, which would close the gap between revenue growth and unit economics.
• Additional finance availability or conversion of the current Morgan Stanley 364-day facility into longer-term, lower-cost capital to support the 1.5 GW buildout.
• Macro tailwinds - higher AI infrastructure demand and an energy market that supports predictable power pricing for high-density customers.

Trade plan - actionable rules

My base trade: establish a core long position at entry $15.90, place a stop loss at $13.25, and target $36.00 as the primary upside objective. Position size should reflect the stock's volatility and your portfolio risk tolerance; I recommend sizing this as a core conviction but not a full-tilt allocation given lingering cash-flow and execution risks.

Horizon guidance:

• Short term (10 trading days) - Expect continued chop. This window is useful for opportunistic layering into a position on intra-day weakness or high-volume pullbacks. Do not expect the AMD approval or major contract wins to show immediate revenue impact in this window.
• Mid term (45 trading days) - Look for newsflow: deal announcements, vendor confirmations, or utilization updates. If AMD approval is confirmed and a multi-year lease is announced within this window, consider adding to bring average cost down aggressively; conversely, if financing costs move sharply higher or a material operational setback emerges, reduce exposure.
• Long term (180 trading days) - Full thesis realization plays out here. The target $36.00 is ambitious but plausible if colocation growth continues, new host contracts are booked, and the company begins to show improving adjusted EBITDA and lower cash burn. Over this horizon, focus is on converting pipeline into signed, revenue-generating contracts and demonstrating path to positive free cash flow.

Trade mechanics: use limit orders to enter at $15.90 or below. If the position gaps lower overnight past $13.25, adhere to the stop and re-evaluate - large gap moves change risk math. Scale into the position on confirmed news (e.g., certified AMD approval update or signed multi-year hosting contract) rather than averaging into clear negative fundamental changes.

Risks and counterarguments

At least four risks warrant careful attention:

• Execution risk on capacity buildout - Scaling toward 1.5 GW requires smooth construction, interconnection, and power contracts. Delays or higher-than-expected capital intensity would push out breakeven timelines and keep cash burn elevated.
• Persistent negative cash flow - The company reported free cash flow near -$957.7M. Continued negative FCF without cheaper long-term financing would dilute equity or require asset sales, undermining upside.
• Commodity and cyclical demand for compute - AI hosting demand is strong today, but customer budgets or vendor roadmaps could shift. A slowdown in data-center demand or customers repatriating workloads would reduce utilization.
• Financing and interest-rate risk - The convertible or short-dated facility profile (364-day facility noted previously) needs conversion into lower-cost, longer-term capital to avoid liquidity squeezes. If credit markets tighten, refinancing risk could dent the thesis.
• High headline multiples - EV/sales and price-based multiples are high because earnings/cash flow remain negative. The stock could re-rate lower if growth doesn’t convert into profits quickly.
• Short-seller pressure and volatility - Short interest remains meaningful, and historical trading patterns show choppy price action. This can amplify downside in market downturns.

Counterargument: A valid bearish view is that the company is structurally unprofitable and reliant on continued equity or expensive debt to fund expansion. If AMD approval is not formalized, or if colocation growth stalls below expectations, the current valuation may not be sustainable. The bear case is straightforward and explains why position sizing and a tight stop are essential.

Conclusion and what would change my mind

I view Core Scientific as a tactical, core long at $15.90 with an upside target of $36.00 over a long-term horizon (180 trading days), provided the company continues to convert colocation pipeline into contracted revenue, secures vendor certifications like AMD approval, and stabilizes financing into lower-cost, longer-dated instruments. The setup offers asymmetric upside if the colocation pivot succeeds; downside is real and sizeable, which is why the stop at $13.25 protects against a structural deterioration.

What would change my view? Significantly worse-than-expected cash burn, an inability to convert short-term financing into longer-term capital, or clear evidence that colocation demand is evaporating would force me to abandon the long thesis. On the flip side, confirmed AMD certification, multi-year AI hosting contracts, or visible adjusted EBITDA improvement would accelerate conviction and justify adding to the position.

Execution checklist for new positions

• Enter at or below $15.90 using a limit order.
• Set stop loss at $13.25 and adhere to it.
• Size the position to limit single-stock exposure to a tolerable percentage of portfolio risk.
• Monitor financing updates, AMD/vendor confirmations, and monthly/quarterly utilization disclosures closely.

Bottom line: CORZ is a constructive, data-center-first trade with binary upside tied to execution and vendor validations. For long-horizon, conviction-oriented investors willing to accept near-term volatility, the current price provides a reasonable entry to participate in the colocation and AI hosting theme.