Invest5 publishers2 min readPublished Updated
Memory inflation reaches the rack: Nvidia's 17% hike lands on Microsoft, Google and AWS
Reported increases of 15% to 17% on Grace Blackwell 300 and Vera Rubin 200 racks shipping next year turn DRAM and HBM contract pricing into the governing input for 2027 capacity budgets.
The Investor · Invest desk
What happened
- Seoul Economic Daily reports Nvidia will lift rack supply prices for the Grace Blackwell 300 and Vera Rubin 200 by up to 17% early next year.
- Bloomberg News, relayed by CNBC, puts the increase at more than 15% in many cases on servers containing Vera Rubin and Grace Blackwell chips.
- Microsoft, Google and Amazon Web Services have already been told, and The Information reported some customers heard it from server manufacturers.
- A Grace Blackwell 300 rack has been priced at about $3.7 million to $4 million.
Compiled by The InvestorSomething wrong?How this is made
Why it matters
- cost At the top of the range, a GB300 rack costs about $680,000 more, and a fixed budget buys roughly 14.5% fewer racks. The buyer of compute carries that, not the memory vendor's counterparty.
- decision Because the increase is sized by memory configuration, memory per rack becomes a procurement lever rather than a spec detail, and buyers who cannot trim it pay a different price from those who can.
- exposure Memory suppliers' windfall is now itemised inside hyperscaler capex, which a South Korean government official says gives customers reason to look for alternatives over the medium to long term.
- contradiction Seoul Economic Daily's own Vera Rubin 200 numbers imply a floor of just 2.6%, against a headline of up to 17% and Bloomberg's more than 15%.
The pass-through is easier to read against the capital expenditure line than the price list. Bloomberg's projection, as carried by Seoul Economic Daily, has major US Big Tech capex going from $781.9 billion this year to $975.4 billion next year [11], a rise of 24.7%, or $193.5 billion of new money [18]. If every one of those dollars went into racks, and racks cost 17% more [1], the installed fleet would grow about 6.6% [19]. It will not be every dollar, because that budget also buys land and power distribution: LS Cable & System is planning its first busduct plant outside its Gumi base, in Poland or Northern Europe [13], on the back of long-term supply contracts worth roughly 5 trillion won at its US arm [14]. The shape of the problem survives the caveat. Most of the visible capex increase is being consumed by unit price rather than unit count.
The mechanism is unglamorous. Memory sits inside the rack bill of materials, and Nvidia has been absorbing rising memory chip costs on parts essential to its GPUs and systems [10]. What is being repriced is not a new product tier but announced generations already in customers' plans, with the increase applied to systems shipped next year [9]. Both reports date from August 2026 [22][23], which puts the affected deliveries in 2027 [21] and leaves roughly five months between notice and the early-year effective date [24]. Any 2027 capacity plan built on a 2026 quote is already carrying a stale number.
The squeeze is also wider than HBM allocation. Alongside the continuing high-bandwidth memory shortage, general-purpose chip scarcity pushed Amazon to raise the Echo Dot from $49.99 to $79.99 [6], a 60% increase [25] on a device with no accelerator in it. That is the same cost curve arriving at two very different price points, which argues against treating the rack increase as an artefact of AI demand alone.
For anyone modelling compute cost, the practical consequence is that rack price has stopped being an input and become an output. It is a function of memory contract pricing, sized by configuration [8]. A fixed dollar-per-rack assumption in a 2027 model is a DRAM forecast that nobody has written down.
What to watch
- Whether Nvidia confirms the increase publicly, and whether the top of the range is tied to specific memory configurations rather than to whole generations.
- First-quarter 2027 HBM and DRAM contract settlements, and whether announced supplier capacity additions arrive in time to cool them.
- Whether hyperscalers start guiding capex in delivered units or rack counts instead of dollars, which would expose how much of the increase is price.