Analog Devices is paying $1.35bn in cash for a neural front end already in production

The contingent slice is where the negotiation shows. It runs to 14.8 percent of the upfront cash ^[1], lifting the ceiling to $1.55 billion ^[2], and the release ties it only to the terms of the definitive agreement, without disclosing what triggers payment ^[13]. A tail that size is usually the residue of a disagreement about a forecast, which means the two sides settled what Alif is and deferred what it becomes.

There is a number on the buy side and none on the sell side. No Alif revenue, headcount or prior round appears in the announcement ^[12], so the $1.35 billion that gets quoted in every banker deck for the next year is a price. It is not yet a multiple, because no one has supplied the denominator. Against ADI's own scale it is a manageable price: no more than about 12 percent of the more than $11 billion of revenue ADI reported for FY25 ^[6][3], with the ceiling at no more than about 14 percent ^[4].

Read the two quotes against each other and the division of labour is explicit. Vincent Roche puts the electro-physical interface, where real-world signals become actionable intelligence, in ADI's column and says it is the domain ADI has mastered for decades ^[9]; Reza Kazerounian describes what Alif built, a heterogeneous architecture with dedicated low-power neural processing engineered in from the start ^[10]; and the release frames the combination as Alif's digital platform alongside ADI's sensing, signal processing, power, connectivity and software ^[11]. The digital compute came by acquisition.

The category reading, that analog incumbents will keep buying their neural compute because internal design cycles are slower than what customers are asking for, is the one that will get repeated. A narrower reading fits the terms better: ADI bought calendar time, since a programme started in September 2026 would be years from the customer designs Alif already sits inside ^[5], and ADI wants this closed within the 113 days between announcement and year end ^[5], with the release itself flagging any extension of the Hart-Scott-Rodino waiting period as the thing that moves the date ^[4]. A third reading deserves a hearing too: at about an eighth of one year's revenue ^[3], this is small enough that treating it as the price of a category may be reading strategy into a line item.

The second reading carries the most weight here. Disclosure will settle it. If ADI books an Alif revenue line at close and the implied multiple lands in ordinary analog M&A territory, the price-without-a-denominator framing is finished and the comp becomes arithmetic anyone can run. If the contingent tail lapses unpaid, that will confirm $1.35 billion was the real number and the earnout added no substance.