7
In e na ional Jou nal o Ad ance and Applied Resea ch
www.ijaa .co.in
ISSN – 2347-7075
Impac Fac o – 8.141
Pee Re iewed
Bi-Mon hly
Vol. 6 No. 38
Sep embe - Oc obe - 2025
The Role o VLSI Technologies in AI-Based Bioelec onics Tools and
Technologies o Diagnos ic Applica ions
Rashmi Singh
Depa men o Compu e Science
D . D. Y. Pa il A s, Comme ce and Science College, Aku di, Pune, MH, India
Co esponding Au ho – Rashmi Singh
DOI - 10.5281/zenodo.17309801
Abs ac :
The con e gence o e y-la ge-scale in eg a ion (VLSI), a i icial in elligence (AI), and
bioelec onics is e olu ionizing heal hca e diagnos ics. Wea ables, implan ables, and lab-on-chip
pla o ms gene a e complex biosignals equi ing secu e, low-powe , and eal- ime analysis. VLSI
add esses hese needs h ough minia u iza ion, e icien compu a ion, and AI-speci ic a chi ec u es.
This pape e iews VLSI’s ole in bioelec onic diagnos ics, emphasizing signal acquisi ion, on-chip
AI accele a o s, ene gy-e icien design, da a p o ec ion, and scalabili y. Applica ions span poin -o -
ca e es ing, neu ological moni o ing, cance bioma ke de ec ion, and con inuous heal h assessmen .
Case s udies o neu omo phic p ocesso s, IoBT sys ems, and lab-on-chip de ices highligh ad ances.
Challenges include ene gy limi s, co-design issues, cos s, and egula ion, wi h u u e p ospec s in
neu omo phic and quan um-inspi ed sys ems.
Keywo ds: VLSI, Bielec onics, A i icial In elligence, Diagnso ics, Signals
In oduc ion:
The con e gence o bioelec onics,
a i icial in elligence (AI), and e y-la ge-
scale in eg a ion (VLSI) echnologies has
opened new on ie s in heal hca e diagnos ics.
Bioelec onic de ices anging om wea able
heal h moni o s o implan able senso s
gene a e as amoun s o physiological and
biochemical da a.
Ha nessing hese signals o eal- ime
diagnos ic applica ions equi es no only
sophis ica ed AI algo i hms bu also e icien
ha dwa e o execu e hem. VLSI echnology,
wi h i s capaci y o minia u iza ion, high-
speed p ocessing, and low-powe ope a ion,
p o ides he essen ial ounda ion o
in eg a ing AI in o bioelec onic sys ems.
This pape explo es he ole o VLSI
in enabling AI-d i en diagnos ic ools,
highligh ing ecen ad ancemen s,
applica ions, challenges, and u u e di ec ions.
P oblem S a emen :
The apid g ow h o bioelec onic
de ices has c ea ed unp eceden ed
oppo uni ies o eal- ime heal hca e
diagnos ics, anging om wea able moni o s
o implan able senso s. These de ices gene a e
as s eams o complex physiological and
biochemical da a ha equi e as , accu a e,
and ene gy-e icien p ocessing o deli e
clinically ele an insigh s. A i icial
in elligence (AI) has shown g ea p omise in
ex ac ing pa e ns, p edic ing disease
p og ession, and enabling pe sonalized
diagnos ics om such da a. Howe e , he
e ec i e deploymen o AI in bioelec onics is
hinde ed by se e al challenges:
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
8
Compu a ional cons ain s o po able
and implan able de ices, which limi he
abili y o un ad anced AI algo i hms in eal
ime.
Powe consump ion and he mal
dissipa ion issues, especially in con inuous
moni o ing and implan able sys ems whe e
ba e y li e is c i ical.
La ency and p i acy conce ns
associa ed wi h cloud-based AI p ocessing,
necessi a ing secu e and e icien on-de ice
compu a ion.
Ha dwa e-so wa e misma ches,
whe e exis ing bioelec onic pla o ms a e no
ully op imized o AI wo kloads.
Ve y-la ge-scale in eg a ion (VLSI)
echnologies o e a po en ial solu ion by
enabling minia u ized, low-powe , and high-
pe o mance ha dwa e a chi ec u es
speci ically ailo ed o AI-d i en diagnos ic
applica ions. Ye , he sys ema ic ole o VLSI
in b idging AI and bioelec onics o scalable,
secu e, and clinically iable diagnos ic
sys ems emains unde explo ed.
Objec i es:
1. To highligh he con e gence o VLSI,
AI, and bioelec onics and i s
ans o ma i e po en ial in ad ancing
heal hca e diagnos ics.
2. To explain he ounda ions and ecen
ad ancemen s in VLSI echnologies ha
enable minia u iza ion, high-speed
p ocessing, low-powe design, and
neu omo phic compu ing o diagnos ic
applica ions.
3. To examine he in e ace be ween AI
and bioelec onics and illus a e how
VLSI acili a es e icien , secu e, and
eal- ime execu ion o AI algo i hms a
he edge.
4. To analyze he ole o VLSI in AI-based
diagnos ic bioelec onic sys ems,
ocusing on signal acquisi ion,
p ep ocessing, on-chip accele a o s,
low-powe a chi ec u es, scalabili y, and
da a secu i y.
5. To explo e di e se diagnos ic
applica ions— om poin -o -ca e es ing
o neu ological moni o ing, cance
de ec ion, genomics, p o eomics, and
wea able/implan able de ices—enabled
by AI-VLSI in eg a ion.
6. To p esen case s udies and eme ging
ends demons a ing he p ac ical
deploymen o VLSI-enabled AI
bioelec onics in eal-wo ld heal hca e
diagnos ics.
7. To c i ically assess challenges and
limi a ions such as powe e iciency,
ha dwa e-so wa e co-design, cos ,
egula o y hu dles, and da a p i acy
conce ns.
8. To p ojec u u e pe spec i es and
oppo uni ies, including quan um-
inspi ed designs, nex -gene a ion
neu omo phic a chi ec u es, edge AI
wi h 6G, pe sonalized diagnos ics, and
con e gence wi h syn he ic biology.
9. To es ablish VLSI as he ounda ional
enable ha b idges biological signal
acquisi ion and AI-d i en diagnos ic
insigh s, se ing he s age o
pe sonalized, eal- ime, and accessible
heal hca e.
VLSI Technologies and AI Bioelec onics:
Founda ions and Ad ancemen s VLSI
echnology e e s o he p ocess o in eg a ing
millions (and now billions) o ansis o s on o a
single silicon chip. Since i s incep ion in he
1970s, VLSI has unde gone exponen ial
g ow h, ollowing Moo e’s Law o decades.
Recen ad ancemen s ex end beyond ansis o
densi y, including: 3D In eg a ed Ci cui s (ICs):
Enabling e ical s acking o logic and memo y
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
9
uni s o compac ness and imp o ed
pe o mance. He e ogeneous In eg a ion:
Combining mul iple unc ionali ies (senso s,
p ocesso s, memo y, analog/digi al in e aces)
wi hin a single package. Low-Powe
A chi ec u es: Such as FinFETs and mul i-
h eshold CMOS, educing ene gy consump ion
in po able diagnos ic de ices. Neu omo phic
VLSI: Mimicking biological neu al ne wo ks o
accele a e AI asks in eal ime. These
inno a ions make VLSI indispensable o AI-
powe ed bioelec onics, whe e de ice size,
e iciency, and eliabili y a e c i ical.
Bioelec onics me ges elec onics wi h
biological sys ems o sensing, moni o ing, and
in e en ion. Examples include elec ochemical
biosenso s, neu al in e aces, and lab-on-chip
pla o ms. AI enhances hese sys ems by
enabling: Pa e n ecogni ion: De ec ing
bioma ke s, disease signa u es, o abno mal
physiological signals. P edic i e modeling:
Fo ecas ing disease onse o p og ession.
Adap i e con ol: Op imizing pe sonalized
diagnos ics and he apeu ic esponses.
Howe e , AI implemen a ion in bioelec onics
aces hu dles such as limi ed compu a ional
esou ces, da a p i acy, and he need o ul a-
low-la ency pe o mance. VLSI add esses hese
cons ain s by embedding compu a ion close o
he senso (―edge AI‖), educing dependence on
ex e nal cloud in as uc u e.
VLSI unde pins he a chi ec u e and
unc ionali y o AI-d i en diagnos ic ools in
se e al ways:
Signal Acquisi ion and P ep ocessing Cus om-
designed analog on -end ci cui s cap u e and
il e biosenso signals. On-chip ampli ie s,
ADCs, and noise- educ ion uni s ensu e high
ideli y da a o AI algo i hms.
On-Chip AI Accele a o s ASICs
(Applica ion-Speci ic In eg a ed Ci cui s):
Op imized o execu ing AI in e ence asks.
FPGAs (Field P og ammable Ga e A ays):
Flexible pla o ms o p o o yping AI
algo i hms in diagnos ics. Neu omo phic Chips:
Real- ime classi ica ion o biosignals wi h ul a-
low powe consump ion. Low-Powe
A chi ec u es Wea able and implan able
diagnos ic de ices demand ex ended ba e y
li e. VLSI powe -e icien a chi ec u es (clock
ga ing, dynamic ol age scaling) ex end
ope a ional longe i y. Da a Secu i y and Edge
AI VLSI-based secu e enc yp ion modules
ensu e pa ien da a con iden iali y while
enabling on-de ice AI in e ence. Scalabili y
and Minia u iza ion In eg a ion o mul iple
senso a ays wi h AI p ocessing co es on a
single chip educes de ice oo p in , enabling
poin -o -ca e and po able diagnos ics.
Diagnos ic Applica ions The syne gy o
VLSI, AI, and bioelec onics is ans o ming
diagnos ics ac oss domains: Poin -o -Ca e
Tes ing (POCT): Po able diagnos ic ki s o
glucose moni o ing, in ec ious diseases, and
ca diac ma ke s. Neu ological Diagnos ics:
B ain-compu e in e aces using VLSI-AI chips
o epilepsy de ec ion o neu odegene a i e
disease moni o ing. Cance Bioma ke
De ec ion: AI-assis ed signal p ocessing in
mic o luidic and biosenso de ices. Genomics
and P o eomics: High- h oughpu sequencing
da a p ocessed ia AI-accele a ed VLSI
sys ems. Wea ables and Implan ables:
Con inuous moni o ing o i al signs (ECG,
SpO₂, glucose, s ess bioma ke s).
T oubleshoo ing App oaches:
Signal Acquisi ion and In eg i y:
Noise Reduc ion: Use o shielding,
g ounding, and analog il e ing o
minimize elec omagne ic in e e ence
(EMI) and biological signal dis o ion.
Calib a ion: Pe iodic calib a ion o
biosenso s o co ec o d i , baseline
shi s, and c oss-sensi i i y.
Redundancy Checks: Inco po a ion o
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
10
mul iple senso nodes o alida e
measu emen s and de ec aul y
channels.
Ha dwa e-Le el T oubleshoo ing
(VLSI Sys ems)
Simula ion and Ve i ica ion: SPICE,
HDL (VHDL/Ve ilog) simula ions o
de ec iming e o s, logic aul s, and
powe dissipa ion issues be o e
ab ica ion.
Design- o -Tes abili y (DFT): Buil -in
sel - es (BIST) and scan chain
echniques o aul localiza ion in
in eg a ed ci cui s.
The mal Managemen : T adi ional
cooling solu ions (hea sinks, packaging
ma e ials) o p e en o e hea ing in
high-densi y chips.
Powe and Ene gy Cons ain s:
Dynamic Vol age and F equency
Scaling (DVFS): Adjus ing ope a ing
ol ages and equencies o balance
pe o mance wi h ba e y li e.
Clock Ga ing and Powe Ga ing:
Selec i ely disabling inac i e ci cui
blocks o oubleshoo unnecessa y
powe d ain.
Ba e y Moni o ing: Regula es ing o
ba e y heal h and ene gy e iciency in
wea ables/implan ables.
AI Algo i hm T oubleshoo ing:
C oss-Valida ion and Benchma king:
Compa ing AI model ou pu s wi h
g ound u h da ase s o iden i y
misclassi ica ions.
O e i ing Con ol: Applying
adi ional echniques like
egula iza ion, p uning, o ea u e
educ ion o imp o e gene aliza ion.
Fallback Rule-Based Sys ems: Using
de e minis ic algo i hms when AI ails
o da a is insu icien .
Sys em In eg a ion Issues:
Modula Debugging: Isola ing
subsys ems (senso , analog on -end,
digi al p ocesso , AI co e) o es
unc ionali y independen ly.
S anda d In e acing P o ocols: Using
SPI, I²C, o UART es p o ocols o
alida e communica ion be ween
senso s and p ocesso s.
Ha dwa e-in- he-Loop (HIL) Tes ing:
Running simula ed biological signals
h ough he ha dwa e o iden i y
in eg a ion aul s.
Da a Secu i y and Reliabili y:
E o -Co ec ing Codes (ECC):
T adi ional pa i y checks and ECC o
eliable memo y ope a ion in diagnos ic
VLSI chips.
Enc yp ion Tes ing: Ve i ying secu e
da a ansmission p o ocols (AES, RSA)
using s anda d c yp og aphic es
ec o s.
Audi T ails: Manual log checks o
da a consis ency and compliance wi h
egula o y s anda d app oaches ha e
been applied o oubleshoo ing in
a ious domains.
Me hodology:
Da a Acquisi ion and P ep ocessing:
The me hodology o his pape is
designed o sys ema ically explo e he ole o
VLSI echnologies in enabling AI-based
bioelec onic diagnos ic ools. A mul i-s ep
app oach has been adop ed, combining
li e a u e e iew, echnology mapping, case
s udy analysis, and compa a i e e alua ion:
Comp ehensi e Li e a u e Re iew:
Re iewed pee - e iewed jou nals,
con e ence p oceedings, pa en s, and indus y
whi e pape s co e ing VLSI design,
bioelec onic de ices, and AI-enabled
diagnos ics.
Su eyed his o ical de elopmen s in
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
11
VLSI (e.g., CMOS scaling, FinFETs, 3D ICs)
alongside ad ancemen s in AI ha dwa e
accele a o s.
Focused on publica ions om he las
wo decades o cap u e bo h ounda ional
knowledge and eme ging ends.
Technology Mapping and Classi ica ion:
Classi ied VLSI inno a ions ele an
o diagnos ics (e.g., neu omo phic
a chi ec u es, low-powe ci cui s, ASICs,
FPGAs).
Mapped hese echnologies agains
hei applica ions in bioelec onics, including
biosensing, neu al in e aces, lab-on-chip
sys ems, and wea able/implan able
diagnos ics.
C ea ed a amewo k linking VLSI
ha dwa e ea u es (minia u iza ion, powe
e iciency, in eg a ion) wi h AI-enabled
diagnos ic capabili ies ( eal- ime analysis,
p edic ion, adap i e con ol).
Analy ical F amewo k De elopmen :
Adop ed a sys ems-le el pe spec i e
o analyze he AI–bioelec onics in e ace and
he ole o VLSI a mul iple laye s (signal
acquisi ion, p ep ocessing, compu a ion, da a
secu i y).
Applied compa a i e e alua ion o dis inguish
be ween adi ional oubleshoo ing me hods
and eme ging AI-d i en sel -diagnos ic
app oaches.
Case S udy Analysis:
Examined ep esen a i e case s udies
including neu omo phic p ocesso s (IBM
T ueNo h, In el Loihi), lab-on-chip
diagnos ics, and IoBT sys ems.
E alua ed hese sys ems in e ms o diagnos ic
accu acy, p ocessing e iciency, scalabili y,
and powe consump ion.
Challenges and Limi a ions Assessmen :
Iden i ied cu en ba ie s by
analyzing echnical epo s and clinical
alida ion s udies, ocusing on issues o powe
e iciency, ha dwa e-so wa e co-design,
manu ac u ing cos s, and egula o y hu dles.
C oss- alida ed challenges wi h
eedback epo ed in biomedical enginee ing
and elec onics esea ch communi ies.
Fu u e Pe spec i es Syn hesis:
Fo ecas ed he ajec o y o AI–VLSI–
bioelec onics con e gence by analyzing
ends in neu omo phic compu ing, quan um-
inspi ed VLSI, and 6G-enabled edge AI.
In eg a ed insigh s om bo h echnology
oadmaps (e.g., ITRS, IEEE epo s) and
biomedical inno a ion o ecas s.
Challenges and Limi a ions:
Powe E iciency and The mal
Managemen :
Challenge: Implan able and wea able
diagnos ic de ices ace s ic cons ain s on
ene gy consump ion and hea dissipa ion.
Con inuous moni o ing and AI in e ence
inc ease ba e y d ain.
Limi a ions: Cu en low-powe VLSI
echniques (clock ga ing, dynamic ol age
scaling) a e insu icien o high-complexi y
AI asks o e long du a ions.
App oaches: De elopmen o ul a-low-
powe neu omo phic chips, ene gy ha es ing
mechanisms (bio uel cells, body hea , mo ion),
and ad anced cooling ma e ials.
Ha dwa e–So wa e Co-Design:
Challenge: AI algo i hms a e o en de eloped
wi hou ull op imiza ion o ha dwa e
execu ion, leading o ine iciencies in
diagnos ic de ices.
Limi a ions: Lack o s anda dized co-design
amewo ks o AI–VLSI in eg a ion in
biomedical sys ems.
App oaches: Join op imiza ion o AI models
and ha dwa e (e.g., quan iza ion-awa e
aining, p uning), adop ion o domain-speci ic
a chi ec u es, and collabo a i e design
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
12
en i onmen s.
Da a P i acy and Secu i y:
Challenge: Pa ien da a handled by diagnos ic
sys ems mus emain secu e while enabling
eal- ime p ocessing.
Limi a ions: T adi ional enc yp ion
algo i hms consume signi ican powe and
la ency when implemen ed in cons ained
de ices.
App oaches: Ligh weigh c yp og aphic
modules embedded in VLSI chips,
homomo phic enc yp ion o on-de ice AI
in e ence, blockchain-based audi ails.
Scalabili y and Minia u iza ion:
Challenge: In eg a ion o mul iple biosenso s,
AI co es, and communica ion modules on a
single chip wi hou sac i icing eliabili y.
Limi a ions: Physical scaling beyond 5 nm is
cos ly and limi ed by quan um and leakage
e ec s.
App oaches: 3D IC in eg a ion, chiple -based
a chi ec u es, he e ogeneous in eg a ion
(mixing analog, digi al, memo y, and senso
uni s).
Reliabili y and Robus ness:
Challenge: Biosenso s and AI chips in
diagnos ics mus unc ion eliably in dynamic
biological en i onmen s wi h high signal
a iabili y.
Limi a ions: Senso d i , noise in e e ence,
and a iabili y ac oss pa ien s educe
obus ness.
App oaches: Redundancy in senso design,
adap i e AI models ained on di e se
da ase s, e o -co ec ing codes in VLSI
memo y modules.
Manu ac u ing and Cos Cons ain s:
Challenge: Ad anced VLSI p ocesses (5 nm,
3 nm nodes) a e expensi e, limi ing
accessibili y o low- esou ce heal hca e
se ings.
Limi a ions: High ab ica ion cos s hinde
scalabili y and widesp ead clinical
deploymen .
App oaches: Open-sou ce ha dwa e design
pla o ms (e.g., RISC-V), econ igu able
FPGAs o p o o yping, cos -sha ing
collabo a ions be ween academia, indus y,
and heal hca e p o ide s.
Regula o y and E hical Ba ie s:
Challenge: Clinical alida ion, pa ien sa e y,
and egula o y app o al (FDA, CE) o AI-
d i en diagnos ic ha dwa e a e leng hy and
complex.
Limi a ions: Absence o clea amewo ks o
ce i ying VLSI–AI–bioelec onics sys ems in
heal hca e.
App oaches: Ea ly-s age collabo a ion wi h
egula o y agencies, explainable AI models o
anspa ency, e hical go e nance amewo ks.
Conclusion:
The in eg a ion o VLSI echnologies
wi h AI-d i en bioelec onics ep esen s a
pa adigm shi in diagnos ic heal hca e. By
enabling minia u iza ion, low-powe
ope a ion, high-speed compu a ion, and secu e
on-de ice in elligence, VLSI p o ides he
essen ial ounda ion o ansla ing biological
signals in o ac ionable insigh s. Applica ions
anging om poin -o -ca e es ing and cance
bioma ke de ec ion o neu al diagnos ics and
wea able heal h moni o s demons a e he
ans o ma i e po en ial o hese echnologies.
Despi e ema kable p og ess,
signi ican challenges emain, including powe
e iciency, ha dwa e–so wa e co-design,
scalabili y, cos , and egula o y app o al.
Add essing hese limi a ions equi es
in e disciplina y collabo a ion ac oss
elec onics, compu e science, bioenginee ing,
and clinical medicine. Eme ging solu ions
such as neu omo phic p ocesso s, 3D IC
in eg a ion, edge AI secu i y modules, and
quan um-inspi ed a chi ec u es highligh
p omising di ec ions o u u e de elopmen .
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
13
Ul ima ely, he con e gence o VLSI,
AI, and bioelec onics will ede ine
diagnos ics by enabling in elligen , eal- ime,
and pe sonalized heal hca e solu ions. As
inno a ion con inues, hese sys ems a e poised
o mo e om labo a o y p o o ypes o
clinically alida ed ools ha imp o e
accessibili y, e iciency, and pa ien ou comes
wo ldwide.comp ehensi e AI-d i en
amewo k o diagnosing dynamic sys ems.
These me hods a e used wi h machine lea ning
models in he amewo k o p o ide an
e ec i e, lexible way o iden i y and ix
p oblems in eal ime. The sugges ed
a chi ec u e o e s scalable, eal- ime solu ions
o in ica e, dynamic sys ems, and i has he
po en ial o e olu ionize oubleshoo ing
ac oss se e al Fields.
Re e ences:
1. H. P. C. Padinha e, C.-Y. Yang, D. Tu, J.
Y. Ge asimo , A. M. Da , A. A. Mo ei a,
M. Masse i, R. K oon, D. Bliman, R.
Olsson, E. S a inidou, M. Be gg en and
S. Fabiano, ―O ganic elec ochemical
neu ons o neu omo phic pe cep ion,‖
Na . Elec on., ol. 7, no. 7, pp. 525–536,
Jul. 2024. doi:10.1038/s41928-024-
01200-5. DIVA Po a
2. P. Belle i, J. Pons i Ta és, I. McCulloch,
P. W. M. Blom, Z. M. Ko ács-Vajna, P.
Gkoupidenis and F. To icelli,
―Un a elling he ope a ion o o ganic
a i icial neu ons o neu omo phic
bioelec onics,‖ Na . Commun., ol. 15,
A . no. (2024). doi:10.1038/s41467-024-
49668-1. Na u e
3. G. Ligo io, F. San o o and H. Kleemann,
―Ad ancing o ganic neu omo phic
de ices: p og ess in modeling,
ab ica ion, and biosensing,‖ Ad .
Elec on. Ma e ., 2024.
doi:10.1002/aelm.202400873. OpenURL
Connec ion
4. M. Hosseinzadeh Fakh , I. Lopez
Ca asco, D. Belyae , J. Kang, Y. H.
Shin, J. S. Yeo, W. G. Koh, J. Ham, A.
Michaelis, J. Opi z and N. Beshchasna,
―Recen ad ances in wea able
elec ochemical biosenso s owa ds
echnological and ma e ial aspec s,‖
Biosens. Bioelec on. X, ol. 19, A . no.
100503, 2024.
doi:10.1016/j.biosx.2024.100503. Yonsei
Uni e si y
5. G. Xu, M. Zhang, T. Mei, W. Liu, L.
Wang and K. Xiao, ―Nano luidic ionic
mem is o s,‖ ACS Nano, Jul. 18, 2024.
doi:10.1021/acsnano.4c06467. AbleSci
6. J. Kim, S. Han, G. Ko, J.-H. Kim, C.
Lee, T. Kim, C.-H. Youn and J.-Y. Kim,
―EPU: An ene gy-e icien explainable
AI accele a o wi h spa si y- ee
compu a ion and hea -map
comp ession/p uning,‖ IEEE J. Solid-
S a e Ci cui s, ol. 59, no. 3, pp. 830–
841, Ma . 2024.
doi:10.1109/JSSC.2024.3346913. DBLP
7. Y.-H. Tsai, Y.-C. Lin, W.-C. Chen e al.,
―A 28-nm 1.3-mW speech- o- ex
accele a o o edge AI de ices,‖ IEEE J.
Solid-S a e Ci cui s, No . 2024.
doi:10.1109/JSSC.2024.3389965.
Resea chGa e
8. G. O’Lea y, J. Koe ne , M. Kanchwala,
J. S. Filho, J. Xu, T. A. Valian e and R.
Geno , ―B ainFo es : Neu omo phic
mul iplie -less bi -se ial weigh -memo y-
op imized 1024- ee b ain-s a e
classi ica ion p ocesso ,‖ IEEE T ans.
Biomed. Ci cui s Sys ., ol. 19, no. 1, pp.
55–67, Feb. 2025.
doi:10.1109/TBCAS.2024.3481160.
PubMed
IJAAR Vol. 6 No. 38 ISSN – 2347-7075
Rashmi Singh
14
9. H. P. C. Ha ikesh, H.-Y. Wu, S. Fabiano
e al., ―Single o ganic elec ochemical
neu on capable o an icoincidence
de ec ion,‖ Sci. Ad ., 2025.
doi:10.1126/sciad .ad 3194. PMC
10. Y. Choi, S. Jeong, H. Jeong, S. Han, J.
Ko, S. E. Yu, Z. Xu, M. S. Chae, M. Son,
Y. Meng, S. Xu, J.-H. Kang, S. Mun and
S.-H. Bae, ―Ad anced AI compu ing
enabled by 2D ma e ial-based
neu omo phic de ices,‖ npj
Uncon en ional Compu ing, ol. 2, no.
1, Ap . 2025. doi:10.1038/s44335-025-
00023-7.
