Cerebras Jumps 8% on Microsoft Deal; Leveraged ETF Surges 16%

The Leverage Shares 2X Long CBRS Daily ETF (CBOE:CBRG) surged 16% to $0.81 in afternoon trading on November 13, after Cerebras Systems (NASDAQ:CBRS) reported a Q3 beat, raised its full-year guidance and disclosed Microsoft as a new hyperscaler customer. Cerebras stock rose 8% to $105.92. The move matters beyond one ticker: it is the first sharp test of investor appetite for a leveraged fund built on a newly public AI-chip maker, and a data point in the broader race to sell inference speed against NVIDIA and AMD.

By James Calloway

What did Cerebras actually report?

Cerebras posted Q3 2026 core revenue of $230.9 million, ahead of the $214.7 million LSEG consensus, CNBC reported. Adjusted earnings per share came in at a profit of 12 cents, against an LSEG consensus expectation of a 14-cent loss. On a GAAP basis, the company lost $0.67 per share, according to its earnings release filed with the SEC. The gap between the two figures reflects $155.1 million in stock-based compensation — a cost shareholders still bear, whatever the adjusted number says.

Guidance did more work than the results. Cerebras forecast Q4 2026 revenue of $310 million to $330 million, well above the $271 million LSEG consensus. It also raised its full-year 2026 core revenue outlook to $940 million to $960 million. Management repeated its plan to triple revenue in 2027, which CNBC calculated implies roughly $2.85 billion — slightly below the $2.92 billion LSEG consensus. The tripling pledge is a promise, not a booked contract; the consensus shortfall on that 2027 figure is the quiet caveat inside an otherwise strong print.

Why is the Microsoft deal significant?

Microsoft (NASDAQ:MSFT) is the new hyperscaler customer. CEO Andrew Feldman said the deal was signed in October, with revenue expected to begin in Q1 2027. Combined with two earlier hyperscaler agreements, the company now counts three hyperscalers using its platform for inference — the stage of AI computing where a trained model generates answers.

Feldman did not hold back on the competitive claim. "Cerebras' technology remains the undisputed champion of speed," he said. That is a marketing assertion from a CEO whose stock is still down 47% since its May IPO; the Microsoft revenue has not started flowing yet, and will not until next year.

Why did the leveraged ETF move twice as much?

The CBRG fund targets twice Cerebras' daily move. It resets each day, a structure that means its returns drift from that multiple over longer holding periods. That reset has been punishing lately: the ETF traded under a dollar even after Thursday's rally and remains down 55% over the past month. A 16% single-day gain does not repair a month of decay, and the same reset that amplifies rallies magnifies any drop in Cerebras stock.

The stakes are straightforward. Leveraged single-stock ETFs on recently IPO'd AI names concentrate volatility in a product aimed at short-term traders, and Cerebras' post-IPO slide shows how quickly the mathematics can turn against holders.

Why are NVIDIA and AMD falling while Cerebras climbs?

Chip stocks sold off broadly on the day. NVIDIA (NASDAQ:NVDA) traded at $161.09, down 3%, with the NASDAQ 100 down 2% and technology the weakest sector. AMD (NASDAQ:AMD) fell harder, down 11% to $522.93.

AMD's slide is notable because it has a direct commercial tie to Cerebras. Management said on the November 12 earnings call that Cerebras' production disaggregated inference platform pairs AMD GPUs for prefill — the step where a model reads the input — with Cerebras CS-3 systems for decode, the step where it generates output. The partnership did not shield AMD from sector-wide selling.

The divergence has a structural explanation. Cerebras is a narrow pure play on inference speed. NVIDIA sells a far broader accelerated computing platform — GPUs, networking, complete systems and the CUDA software ecosystem — to hyperscalers, enterprises and sovereign governments. AMD spreads across data center accelerators, server and PC processors, gaming, embedded products and Helios rack-scale systems. Both giants carry greater exposure to overall chip-sector sentiment, so they fall together while Cerebras trades on company-specific news.

What makes Cerebras' wafer-scale approach different?

Cerebras builds wafer-scale processors — chips the size of an entire silicon wafer rather than a diced die — designed for very fast AI inference. It sells them as systems and rents capacity through Cerebras Cloud, which now generates most of its revenue.

One design choice stands out: the chips skip high-bandwidth memory (HBM), the stacked memory modules that dominate AI accelerators. Management expects that will help Cerebras avoid the memory-supply squeeze it sees facing competitors through 2028. If that thesis holds, it is a genuine structural advantage; if HBM supply loosens earlier, the edge narrows.

What could still go wrong?

The next test is whether Microsoft revenue starts flowing on the schedule Feldman described. Optimists can point to $26.1 billion in remaining performance obligations — contracted future revenue — as of September 30, plus the repeated 2027 growth plan. Management also expects a return to breakeven core operating profit in Q4, which would add support to the story.

For the bears, OpenAI offers a clear warning. Cerebras acknowledged it is running behind on initial capacity under its OpenAI agreement, citing land, power and shell constraints, and expects to be back on schedule by the second half of 2027. Management said OpenAI had not cancelled any of the delayed capacity as of the call — but that is a fact about today, not a guarantee about tomorrow.

For CBRG holders, the risk is a larger version of the same problem. Any slip in the Microsoft ramp or any OpenAI cancellation would hit Cerebras stock, and the fund's daily reset would magnify the drop. Investors watching the name have a concrete signal to monitor: whether OpenAI's capacity timeline holds, which management itself flagged as the swing factor for the business — and by extension for the leveraged fund tied to it.