Memory is 30% of hyperscaler capex, and the cheapest fix is the DDR4 you already own

Meta's case for salvage does not start with price. Its paper puts server life at three to five years and the DRAM inside at seven to 10 ^[16], so the modules outlive the chassis they shipped in by two to seven years ^[6]. Around 40% of the fleet is memory-capacity bound ^[16]. The shortage and the surplus are in the same building.

The part that closes the gap is deliberately unambitious: a CXL 2.0 Type-3 expander on a PCIe 5.0 x16 link, bridging two DDR4 channels, 128GB per chip assembled from 32GB modules pulled out of retired machines ^[14]. That works out to four salvaged modules per chip and two chips per MemServer to reach the 256GB of DDR4-2400 sitting beside 768GB of local DDR5-6400 ^[5]. A quarter of each machine's memory is second-hand ^[4], at a claimed 50ns idle round trip on the controller path ^[15], and server counts for disaggregated ML inference fall by up to 25% ^[13].

Supply is the part that caps the idea. Samsung, SK hynix and Micron have been winding DDR4 down since last year, and TrendForce measured the legacy rally at up to 50% for DDR4 in Q1 alone ^[6]. No new DDR4 is coming, so the pool of poolable memory is whatever you bought in 2021 and 2022 and are retiring now. This scales with your own decommissioning calendar, not with anyone's fab.

The ladder underneath is steep. Contract prices rose 90 to 95% in Q1 2026 and another 58 to 63% in Q2 ^[2][3], which compounds to roughly 3.0 to 3.2 times in two quarters ^[1], and 3.4 to 3.8 times by the end of Q3 if server DRAM lands inside the forecast 13 to 18% ^[4][2]. U.S. hyperscalers capped part of that with long-term agreements running through 2027 and 2028 ^[4]. Buyers without one were already being filled at about 70% of orders late last year, while wafers went to HBM ^[5].

Marvell answers this on a slower clock. The Structera X expansion controllers were announced in July 2024 and are in hyperscaler deployments, with inline LZ4 that VP Khurram Malik told EE Times returns 2 to 2.5 times effective capacity ^[9]; on Meta's 256GB that would read as 512 to 640GB ^[8]. The Structera S switch samples this quarter ^[10], and the one genuinely new part, the Bravera SC6, samples in Q4 and puts drives in 2027 at the earliest ^[8]. The switch's under-460ns round trip is roughly nine times Vistara's idle figure ^[7], for a different job: 48TB shared across 16 or 32 sockets at 4 TB/s ^[10] rather than two channels bolted to one host.

So the three-tier portfolio is an assembly. The optical tier came in with the $3.25 billion Celestial AI purchase that closed in February ^[11], and in March Marvell took $2 billion from Nvidia tied to NVLink Fusion, the proprietary scale-up fabric that CXL pooling has to sit alongside ^[12]. In 2024, DDR4 reuse was a sustainability line; the crunch turned it into a capex line ^[17]. The proof point for that argument is a socket Marvell does not hold ^[13].