Nanya Technology

Research evidence excerpt

Nanya Technology

new suppliers for key customers to produce enough ARM-architecture CPU Adj. EPS - 26E (NT$) 39.82 58.43 46.7%

chip-grade memory. There appear to be various types of density and Adj. EPS - 27E (NT$) 43.10 89.27 107.1%

performance LPDDR for SOCAMM and we believe there is a reasonably high

Quarterly Forecasts (FYE Dec) possibility NYT will tap into this market starting next year, supporting mid to

long-term product mix improvement. We estimate NYT’s server end-market Adj. EPS (NT$)

2025A 2026E 2027E

exposure to be in the low-to-mid teens % and that every 100bps of SOCAMM Q1 (0.63) 8.41A 19.91

market share gain will help NYT expand server revenue by US$2+bn, Q2 (1.32) 14.33 21.50

substantially increasing server market exposure. In addition, we are witnessing Q3 0.50 17.11 22.59

Q4 3.58 18.58 25.28

a severe shortage in SSD controller grade cache memory (currently mainly FY 2.13 58.43 89.27

DDR4) and we see a high chance that NYT expands eSSD (categorized as

broad-base server compute market) grade memory sales in the mid-to-long Style Exposure

term. We think this new product diversification will help NYT gain better

bargaining power to secure LTAs at lucrative contract terms, supporting mid-

to-long term profitability and product mix improvement (adding new server

package module and eSSD grade memory, broadbase server exposure could be

near 30%). We pencil in a 37% blended ASP increase next year and forecast

further OPM advancement (73-77% in FY26-28E).

• Strong technology optionality against competition. We believe that NYT’s

competition with its Chinese DRAM peer will continue to make headline

noise, however, we would highlight that the companies have limited customer

overlap, with NYT having a more diversified customer base across