IEEE Solid-States Circuits Magazine - Summer 2023 - 60

(HPC) applications, and autono mous
driving. As a result, the classical
memory subsystem hierarchy of
SRAM, DRAM and NAND/HDD is
strained in supporting this growth.
SRAM and DRAM cost-per-bit scaling
has slowed down, and a large
gap (~10×) exists between SRAM and
DDR DRAM in terms of the cost per
bit, latency, and power. Hence, there
is an opportunity for new memory
options, which do not need to be as
fast as SRAM nor as dense as DRAM,
to fill this gap. Likewise, there is
also a large gap in density and performance
between DRAM and NAND
storage to justify a new technology
to be placed between them. On the
embedded NVM (eNVM) side, eFlash
scaling beyond 28 nm has proven
challenging, as eFlash requires a
significant number of additional
masks, plus extremely high voltages
for program and erase. These
challenges ignited the need for new
memories with low-voltage operations
to support eNVM applications.
Magnetoresistive RAM (MRAM),
resistive RAM (RRAM), and ferroelectric
RAM (FeRAM) have been the primary
focus of academia and industry
researchers in the last decade to address
both the memory hierarchy
and eNVM applications. The most
common usage of these memories
is to combine the storage element
with a transistor switch, as shown in
Figure 10(a). MRAM and RRAM store
data as resistance, which can be toggled
between high and low resistance
states by the write current direction.
FeRAM, on the other hand, stores
data as ferroelectric material polarization
in its ferroelectric capacitor,
which can be programmed through
electric field directions.
While all three of these memories
are proposed as eNVM [18], [19],
[20], MRAM and FeRAM have shown
the most promise for high-speed and
high-endurance cache and memory
applications [21], [22] to close the
gap between SRAM and DRAM, which
can further increase memory-critical
workload performances in graphics,
AI, and HPC. In addition, while MRAM
60 SUMMER 2023
and RRAM have already been offered
as eNVM for high-volume manufacturing,
RRAM has an opportunity to
fill the gap between DRAM and Flash.
Summary
Semiconductor memories have come
a long way since the birth of the transistor.
Incumbent memory technologies
introduced along the way have
leaned on relentless innovations and
execution to sustain decades of scaling,
playing indispensable roles in
the transistor revolution. In parallel,
new forms of memories continue to
be invented and researched, seeking
to meet the rising challenges
of the future. With the accelerating
demands for compute and storage
showing no signs of slowdown, further
advancements in semiconductor
memories will remain necessary and
present many exciting opportunities
for the industry.
References
[1] L. Su and S. Naffziger, " Innovation for
the next decade of compute efficiency, "
in Proc. IEEE Int. Solid-State Circuits
Conf. (ISSCC), San Francisco, CA, USA,
2023, pp. 8-12, doi: 10.1109/ISSCC42615.
2023.10067810.
[2] " Frontier supercomputer debuts as world's
fastest, breaking exascale barrier, " Oak
Ridge National Laboratory, Oak Ridge,
TN, USA, May 2022. [Online]. Available:
https://www.ornl.gov/news/frontiersupercomputer-debuts-worlds-fastestbreaking-exascale-barrier
[3]
D. Reinsel, J. Gantz, and J. Rydning, " The
digitization of the world: From edge to
core, " Seagate, Dublin, Ireland, IDC White
Paper - #US44413318, Nov. 2018. [Online].
Available: https://www.seagate.com/files/
www-content/our-story/trends/files/
idc-seagate-dataage-whitepaper.pdf
[4] J. Chang et al., " Embedded memories for mobile,
IoT, automotive and high performance
computing, " in Proc. VLSI Symp. Technol.
Circuits, Kyoto, Japan, 2017, pp. T26-T27,
doi: 10.23919/VLSIT.2017.7998187.
[5] G. Yeap, " 5nm CMOS production technology
platform featuring full-fledged EUV, and
high mobility channel FinFETs with densest
0.021µm2 SRAM cells for mobile SoC
and high performance computing applications, "
in Proc. IEEE Int. Electron Devices
Meeting, San Francisco, CA, USA, 2019,
pp. 36.7.1-36.7.4, doi: 10.1109/IEDM19573.
2019.8993577.
[6] Y. Mii, " Semiconductor innovations, from
device to system, " in Proc. IEEE Symp. VLSI
Technol. Circuits, Honolulu, HI, USA,
2022, pp. 276-281, doi: 10.1109/VLSITechnologyandCir46769.2022.9830423.
[7]
N. Loubet et al., " Stacked nanosheet gateall-around
transistor
to enable scaling
beyond FinFET, " in Proc. Symp. VLSI Technol.,
Kyoto, Japan, 2017, pp. T230-T231,
doi: 10.23919/VLSIT.2017.7998183.
IEEE SOLID-STATE CIRCUITS MAGAZINE
[8] D. Kwon, " A 1.1V 6.4Gb/s/pin 24-Gb DDR5
SDRAM with a highly-accurate duty corrector
and NBTI-tolerant DLL, " in Proc.
IEEE Int. Solid-State Circuits Conf. (ISSCC),
2023, pp. 27-29, doi: 10.1109/ISSCC42615.2023.10067651.
[9]
D. Ha and H.-S. Kim, " Prospective innovation
of DRAM, flash, and logic technologies
for digital transformation (DX) era, "
in Proc. IEEE Symp. VLSI Technol. Circuits,
2022, pp. 417-418, doi: 10.1109/VLSITechnologyandCir46769.2022.9830465.
[10]
M.-J. Park et al., " A 192-Gb 12-high 896GB/s
HBM3 DRAM with a TSV auto-calibration
scheme and machine-learning-based
layout optimization, " IEEE J. Solid-State
Circuits, vol. 58, no. 1, pp. 256-269, Jan.
2023, doi: 10.1109/JSSC.2022.3193354.
[11] D. Lee et al., " A 16Gb 27Gb/s/pin T-coil
based GDDR6 DRAM with merged-MUX TX,
optimized WCK operation, and alternative-data-bus, "
in Proc. IEEE Int. Solid-State
Circuits Conf. (ISSCC), 2022, pp. 446-448,
doi: 10.1109/ISSCC42614.2022.9731614.
[12] J.-H. Kang et al., " A 24-Gb/s/pin 8-Gb
GDDR6 with a half-rate daisy-chain-based
clocking architecture and I/O circuitry
for low-noise operation, " IEEE J. Solid-State
Circuits, vol. 57, no. 1, pp. 212-223, Jan.
2022, doi: 10.1109/JSSC.2021.3114205.
[13] T. M. Hollis et al., " An 8-Gb GDDR6X DRAM
achieving 22 Gb/s/pin with single-ended
PAM-4 signaling, " IEEE J. Solid-State Circuits,
vol. 57, no. 1, pp. 224-235, Jan. 2022, doi:
10.1109/JSSC.2021.3104093.
[14] D.-H. Kim, " A 16Gb 9.5Gb/S/pin LPDDR5X
SDRAM with low-power schemes exploiting
dynamic voltage-frequency scaling
and offset-calibrated readout sense amplifiers
in a fourth generation 10nm DRAM
process, " in Proc. IEEE Int. Solid-State Circuits
Conf. (ISSCC), 2022, pp. 448-450, doi:
10.1109/ISSCC42614.2022.9731537.
[15] Y.-C. Kwon, " A 20nm 6GB function-inmemory
DRAM, based on HBM2 with a
1.2TFLOPS programmable computing unit
using bank-level parallelism, for machine
learning applications, " in Proc. IEEE Int.
Solid-State Circuits Conf. (ISSCC), 2021, pp.
350-352, doi: 10.1109/ISSCC42613.2021.
9365862.
[16] S. Lee, " A 1ynm 1.25V 8Gb, 16Gb/s/pin
GDDR6-based accelerator-in-memory supporting
1TFLOPS MAC operation and various
activation functions for deep-learning
applications, " in Proc. IEEE Int. Solid-State
Circuits Conf. (ISSCC), 2022, pp. 1-3, doi:
10.1109/ISSCC42614.2022.9731711.
[17] A. Khakifirooz et al., " A 1.67Tb 5b/cell
flash memory fabricated in a 192-layer
floating-gate 3D-NAND technology and
featuring a 23.3Gb/mm2 bit density, " IEEE
Solid-State Circuits
Lett., early access,
2023, doi: 10.1109/LSSC.2023.3285508.
[18] Y.-C. Shih et al., " A reflow-capable, embedded
8Mb STT-MRAM macro with 9ns read
access time in 16nm FinFet logic CMOS
process, " in Proc. IEEE Int. Electron Devices
Meeting (IEDM), 2020, pp. 11.4.1-11.4.4,
doi: 10.1109/IEDM13553.2020.9372115.
[19] O. Golonzka et al., " Non-volatile RRAM embedded
into 22FFL FinFET technology, " in
Proc. Symp. VLSI Technol., 2019, pp. T230-
T231, doi: 10.23919/VLSIT.2019.8776570.
[20] M. Trentzsch et al., " A 28nm HKMG super
low power embedded NVM technology
based on ferroelectric FETs, " in Proc.
IEEE Int. Electron Devices Meeting (IEDM),
2016, pp. 11.5.1-11.5.4, doi: 10.1109/
IEDM.2016.7838397.
[21] J. G. Alzate et al., " 2 MB array-level demonstration
of STT-MRAM process and performance
towards L4 cache applications, "
https://www.ornl.gov/news/frontier-supercomputer-debuts-worlds-fastest-breaking-exascale-barrier https://www.ornl.gov/news/frontier-supercomputer-debuts-worlds-fastest-breaking-exascale-barrier https://www.ornl.gov/news/frontier-supercomputer-debuts-worlds-fastest-breaking-exascale-barrier https://www.seagate.com/files/www-content/our-story/trends/files/idc-seagate-dataage-whitepaper.pdf https://www.seagate.com/files/www-content/our-story/trends/files/idc-seagate-dataage-whitepaper.pdf https://www.seagate.com/files/www-content/our-story/trends/files/idc-seagate-dataage-whitepaper.pdf
IEEE Solid-States Circuits Magazine - Summer 2023

Table of Contents for the Digital Edition of IEEE Solid-States Circuits Magazine - Summer 2023
