The Memory Market in 2025. A Forecast
We looked at Q4 performance, Q1 guidance and some of the market
forces. Plugged them into the memory model and came up with some numbers. There
are some big trade offs and as a result, the numbers were surprising. This is
for 2025 overall.
Caveat: I do not usually believe the “macro” normally affects memory
market, but if the US continues to implement trade restrictions with China, That
could single handedly put the entire tech market into a holding pattern.
I am a free trader because free trade helps the market grow.
The Guesstimates:
1)
DRAM Bits overall increase high teens percentage
2)
DDR Pricing decreases mid teens percentage
3)
HBM Bit volume increases roughly 90%
4)
HBM Bit pricing is flat
5)
DRAM ASP is roughly flat (HBM offsets DDR)
6)
DRAM Revenue increases mid teens percentage
7)
DRAM Operating Margin dollars increase mid teens
percentage
8)
NAND Bits increase mid teens percentage
9)
NAND ASP decreases Mid teens percentage
10 NAND Revenue is flat to slightly down
End result: Expect big 3 memory company margins (dollars) to
increase Mid teens percentage over the year… growth almost totally dominated by
HBM.
The “cycle disrupting” product is HBM. It sucks up capex,
sells for 6x bit price of DDR. Costs 3.3x the price of DDR bits. Potentially
sucks up excess DRAM capacity. Adds lots of margin for ever bit converted from
DDR. We expect supply to catch up by the end of 2025 and we will see what
happens next in 2026
We have all the gory details and reasons in excel spreadsheets to look
at the tradeoffs and some possible other scenarios
Mark Webb