BTC $83,935.18 +1.23%
ETH $2,560.05 +2.18%
BNB $753.04 +0.79%
XRP $1.42 +1.47%
SOL $112.76 +2.93%
TRX $0.3318 +0.51%
DOGE $0.0876 +2.56%
ADA $0.2533 +2.54%
BCH $285.89 +3.24%
LINK $13.46 +4.14%
HYPE $87.69 +3.15%
AAVE $170.43 +1.46%
SUI $1.12 +2.87%
XLM $0.2029 +3.82%
ZEC $1,333.77 +8.52%
AAPL $337.53 +0.32%
AMZN $264.31 +0.69%
GOOGL $353.41 +0.69%
MSFT $538.12 +0.59%
META $722.31 +0.43%
NVDA $230.45 -0.02%
TSLA $384.23 +0.47%
SNDK $1,584.46 -0.41%
INTC $105.30 +0.47%
SPCX $164.69 +0.64%
MU $1,029.94 -0.41%
AMD $613.79 +0.83%
BTC $83,935.18 +1.23%
ETH $2,560.05 +2.18%
BNB $753.04 +0.79%
XRP $1.42 +1.47%
SOL $112.76 +2.93%
TRX $0.3318 +0.51%
DOGE $0.0876 +2.56%
ADA $0.2533 +2.54%
BCH $285.89 +3.24%
LINK $13.46 +4.14%
HYPE $87.69 +3.15%
AAVE $170.43 +1.46%
SUI $1.12 +2.87%
XLM $0.2029 +3.82%
ZEC $1,333.77 +8.52%
AAPL $337.53 +0.32%
AMZN $264.31 +0.69%
GOOGL $353.41 +0.69%
MSFT $538.12 +0.59%
META $722.31 +0.43%
NVDA $230.45 -0.02%
TSLA $384.23 +0.47%
SNDK $1,584.46 -0.41%
INTC $105.30 +0.47%
SPCX $164.69 +0.64%
MU $1,029.94 -0.41%
AMD $613.79 +0.83%
first_img

Samsung Electronics plans three models of zNAND, aimed at customized AI storage

2026-10-12 11:55:50

According to a report by ZDNet Korea on October 12, Samsung Electronics is expanding the application range of the next-generation NAND flash memory zNAND and plans to subdivide the products into three types: zNAND-O, zNAND-T, and zNAND-I, in addition to the publicly announced zNAND-O. The report states that Samsung Electronics has not officially announced the latter two products, but the industry has learned that related plans have entered the early development stage to address the increasing demand for customized storage in AI infrastructure.

zNAND-O is aimed at edge AI. Unlike traditional wire bonding, this solution vertically stacks NAND in 4 or 8 layers through silicon vias, shortening the data transmission path and enhancing speed and energy efficiency. Samsung Electronics first publicly showcased the zNAND-O model at a memory-related event in the United States in August this year.

zNAND-T focuses on throughput and is aimed at high-capacity, high-performance AI data centers, also expanding bandwidth through vertical stacking and silicon vias, and is referred to by the industry as high-bandwidth flash memory. zNAND-I focuses on increasing input/output operations per second, used for frequent processing of small-capacity temporary data such as key-value caches in AI inference. The report also mentions that SK Hynix is developing AI NAND product series such as AIN P, AIN D, and AIN B.

app_icon
ChainCatcher Building the Web3 world with innovations.