Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

SynDCIM: A Performance-Aware Digital Computing-in-Memory Compiler with Multi-Spec-Oriented Subcircuit Synthesis

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims a compiler can turn performance targets into digital compute-in-memory macro layouts, and backs it with a 40nm test chip.

desk verdict A real DCIM compiler with a real tape-out, worth refereeing despite an unverified prediction-to-silicon link and a missing comparison table. read the letter →

arxiv 2411.16806 v2 pith:HORNXZFE submitted 2024-11-25 cs.AR

Kunming Shao , Fengshi Tian , Xiaomeng Wang , Jiakun Zheng , Jia Chen , Jingyu He , Hui Wu , Jinbo Chen
show 7 more authors
Xihao Guan Yi Deng Fengbin Tu Jie Yang Mohamad Sawan Tim Kwang-Ting Cheng Chi-Ying Tsui
This is my paper · ORCID
classification cs.AR
keywords digitalcomputing-in-memorymemorycompileragileEDAdesignautomationmulti-spec-orientedsearchsubcircuitsynthesisParetofrontier40nmCMOStestchip
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that a digital computing-in-memory (DCIM) macro—a circuit that performs multiply-accumulate math directly inside an SRAM array—can be generated automatically from user performance targets instead of being hand-crafted. It presents SynDCIM, a compiler that searches a library of subcircuits (SRAM cells, multipliers, adders, alignment and fusion units) to satisfy specifications on frequency, throughput, power, and area, then produces a layout through a standard digital flow with structured placement. The claim is supported by a 40nm test macro that reached 1.1 GHz and 1921 TOPS/W under sparse INT4 inputs, competitive with manually designed DCIM macros. If the claim holds, adapting DCIM to a new AI workload would shift from circuit engineering to writing a specification.

What carries the argument

The load-bearing mechanism is the subcircuit library paired with the multi-spec-oriented searcher. The library holds PPA lookup tables for seven subcircuit types across topologies, dimensions, and timing constraints; the searcher evaluates candidate assemblies against the user's frequency and PPA preferences and applies retiming, column splitting, register removal, and component substitution as fine-tuning moves. A second mechanism is the bit-wise carry-save adder built from a mix of 4-2 compressors and full adders, which trades area and power against critical-path delay, plus a structured-data-path placement script that keeps the SRAM array regular during automatic place-and-route.

What would settle it

Set the compiler to the 64x64, MCR=2, INT4/8, FP4/8 specification used in the Pareto-front experiment, fabricate the resulting macro, and measure its maximum clock frequency and energy at 0.9 V; if the measured shmoo falls well short of the 800 MHz target, or measured power greatly exceeds the library estimate, the library models are over-optimistic.

Watch

Extended reading notes

Core claim

The central discovery is that architectural synthesis of DCIM macros can be made specification-driven. SynDCIM accepts array dimensions, precision mix, memory-compute ratio, and user PPA preferences as inputs, and a multi-spec-oriented heuristic searcher assembles subcircuits from a library whose entries carry power, timing, and area lookup tables. The searcher checks critical paths in the MAC chain, retimes or splits columns when timing is violated, and emits a Pareto front of candidate designs. One candidate is taken through synthesis, structured placement and routing, and post-layout verification to yield a manufacturable layout. The paper presents fabrication results showing that the generated macro measures 1.1 GHz and 1921 TOPS/W, and claims this makes SynDCIM the first performance-aware DCIM compiler.

Load-bearing premise

The whole performance-to-layout promise rests on the PPA lookup tables in the subcircuit library predicting post-layout silicon behavior accurately enough that the search can trust them.

Editorial extensions

If this is right

  • Designers can explore DCIM macro alternatives for a target application by rerunning the compiler with different PPA preferences instead of hand-editing cells.
  • Support for FP8 and BF16 formats costs roughly 10% to 20% more power than INT4/INT8 at the same array size, according to post-layout evaluation.
  • Larger arrays improve energy efficiency because the MAC logic and peripherals are amortized over more bits.
  • The measured 1.1 GHz and 1921 TOPS/W place the compiler-generated macro among state-of-the-art manual DCIM designs.
  • A single flow covers RTL generation, synthesis, placement, routing, DRC/LVS, and post-layout simulation, so fabricated macros can be produced from a specification.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural test is to run SynDCIM on the exact specifications of published manual DCIM macros and compare area, frequency, and energy directly; the paper reports competitive numbers but does not show a like-for-like head-to-head.
  • The same library-plus-search recipe could be reused for other memory technologies, such as ReRAM-assisted or analog CIM, if PPA models for those cells were added.
  • Closing the loop by re-characterizing the library after placement and routing, and feeding measured silicon data back in, would turn the performance-alignment claim into a self-correcting process.
  • The compiler's Pareto front could also feed a system-level design-space explorer, so an entire AI chip and its embedded DCIM macros are co-optimized.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper presents SynDCIM, an end-to-end compiler for digital computing-in-memory (DCIM) macros that takes user-defined architecture and performance specifications (dimensions, precisions, MCR, MAC frequency, weight-update frequency, PPA preferences) and automatically produces a DCIM macro layout. The compiler consists of a subcircuit library with PPA lookup tables, a multi-spec-oriented (MSO) heuristic searcher that selects and optimizes subcircuits to meet timing and PPA constraints, and a synthesis/APR flow using Cadence Innovus SDP for structured placement and routing. The authors report post-layout evaluations across array sizes and precisions, a Pareto-frontier design space exploration for a 64x64 MCR=2 macro, and silicon validation from a 40nm test chip containing eight macros. The measured test macro reaches 1.1 GHz at 1.2V and 1921 TOPS/W at 12.5% input sparsity and 50% weight sparsity in INT4, with an area efficiency of 80.5 TOPS/mm2 scaled to 1b-1b. The central claims are that SynDCIM is the first performance-aware DCIM compiler, that it automatically generates optimal or Pareto-frontier DCIM macro designs aligned with user-defined performance specifications, and that the generated designs are competitive with state-of-the-art manually designed DCIM macros.

Significance. If the central claims hold, SynDCIM addresses a real gap in DCIM design automation: reducing the manual effort in custom memory cells and MAC logic while explicitly optimizing for user-defined power, performance, and area targets. The silicon validation at 40nm is a concrete strength, as is the reported post-layout evaluation flow and the use of a characterized subcircuit library. The paper also proposes a practical mixed compressor/full-adder CSA design and structured-data-path APR methodology, which are useful contributions independent of the compiler framework. However, the significance is currently tempered by missing supporting evidence: the key comparison table is absent from the manuscript, the accuracy of the PPA lookup tables is not directly validated against measured silicon, and the optimality/Pareto-frontier claims rest on a heuristic with no demonstrated bound or comparison. These issues are fixable but must be addressed before the paper's headline assertions can be accepted.

major comments (4)
  1. [Section IV-B, Table II] Table II, which is the central comparison with state-of-the-art DCIM designs, is not present in the manuscript: only the caption appears. The claim that the SynDCIM-generated macro exhibits competitive performance cannot be verified without the table contents, including the comparison points, process nodes, precisions, sparsity conditions, and normalization rules. In addition, the reported 1921 TOPS/W is measured at 12.5% input sparsity and 50% weight sparsity; the text does not state the dense (100%/100%) baseline or clarify whether the energy-efficiency number is normalized to 1b-1b operation as the area-efficiency number is. Please provide the full table and a dense or normalized comparison so that the competitive-performance claim is actually testable.
  2. [Section III-B and Section IV-B] The paper never provides a predicted-versus-measured comparison for the fabricated macro. The performance-to-layout alignment claim requires that the SCL lookup tables, built from custom-cell characterization and, for uncharacterized configurations, from estimated and scaled synthesis data, accurately predict final silicon behavior. Section III-B states that PPA for other configurations is 'estimated and scaled from synthesis data,' but there is no report of what the SCL or post-layout simulation predicted for the 64x64 MCR=2 test macro in terms of frequency, power, and area, nor a comparison with the measured 1.1 GHz, 1921 TOPS/W, and 0.112 mm2. Without this validation, a reader cannot determine whether the MSO searcher's selected design truly meets user specifications or whether the result is an artifact of LUT optimism or pessimism. Please add a predicted-versus-measured table or scatter plot for the fabricated macro, and discuss any discrepancies.
  3. [Section III-C and Abstract] The abstract and Section III-A describe the generated designs as 'optimal DCIM macro designs' and the searcher as producing designs 'at the Pareto frontier,' but Algorithm 1 is a heuristic hierarchical search with no optimality guarantee and no comparison against exhaustive search, random search, or an alternative multi-objective optimizer. The claim of Pareto optimality is load-bearing for the paper's contribution, yet the evidence in Figure 8 only shows that several design points occupy a plausible trade-off region. Please either soften the claims to 'near-optimal' or 'Pareto-approximate,' or provide a validation study on a small design space where exhaustive enumeration is feasible to demonstrate that the heuristic does not miss substantially better designs.
  4. [Table I (Introduction)] Table I, which is used to position SynDCIM as the first performance-aware DCIM compiler, also appears without its contents in the manuscript. The comparison dimensions that justify the 'first' claim—such as whether prior compilers support multi-spec optimization, automated layout generation, silicon validation, or flexible precision—are not visible. Without this table, the novelty claim is unsupported. Please include the full table with explicit comparison criteria and entries for AutoDCIM, the ISLPED structured-macro work, EasyACIM, and Arctic, among others.
minor comments (5)
  1. [Section II-A] The acronym MCR is used in 'MCR-aware designs' but is not expanded at first use; please define Memory-Compute Ratio when it first appears.
  2. [Section IV-A, Figure 7] The sentence 'FP8 and BF16 consume around 10% and 20% more power than INT4 and INT8, respectively' is ambiguous because it is unclear whether the comparison is at the same array dimensions and whether the 10% and 20% refer to FP8 vs INT4 and BF16 vs INT8, or to both FP formats vs both INT formats. Please rephrase for clarity.
  3. [Section IV-B] The phrase 'test DCIM macro chip generated by SynDCIM' is awkward; it would be clearer as 'a test chip containing a DCIM macro generated by SynDCIM.'
  4. [Section III-D and IV-A] The manuscript uses both 'post-simulation' and 'post-layout simulation' in different places; please standardize the terminology to avoid confusion about which simulation stage is being reported.
  5. [References] Reference [11] is listed as ICCAD 2024 with a DOI but the citation year appears inconsistent with the conference year in the text; please verify the bibliographic details for all references.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the compiler's design-selection chain is self-contained and the central claim is independently checked by silicon measurements.

full rationale

SynDCIM's derivation chain is: user specifications plus a characterized subcircuit library (Section III-B) drive a multi-spec-oriented searcher (Section III-C) that selects a design, which is then synthesized, placed, routed, and verified by post-layout simulation (Section III-D), with a fabricated 40nm test chip providing independent silicon measurements (Section IV-B: 1.1 GHz and 1921 TOPS/W). None of these steps defines the claimed output in terms of the input: the PPA lookup tables are empirical characterizations of custom cells, not parameters fitted to the target performance, and the final measured frequency and efficiency are external to the search. The self-citations (AutoDCIM [5], REDCIM [9], ResCIM [11]) are background descriptions, not load-bearing justifications for SynDCIM's novelty or correctness. The absence of a predicted-versus-measured comparison and the missing Table II contents are verifiability and completeness gaps, not circularity, because they do not make the derivation equivalent to its own inputs.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No numeric free parameters are fitted in the paper; the PPA lookup tables are empirical characterization data, not parameters tuned to make the central claim pass. No new physical entities are introduced. SynDCIM is a software compiler; the mixed CSA is a circuit topology, not a postulated entity needing external evidence.

assumptions (4)
  • domain assumption A DCIM macro can be decomposed into seven subcircuit categories whose PPA behavior can be characterized independently and composed into an accurate macro-level estimate.
    Section II-B and III-B define the search space; if interactions between subcircuits are not captured by independent LUTs, the search may select designs that fail to meet targets.
  • domain assumption The characterized PPA lookup tables remain valid across the explored dimensions, precisions, and timing constraints.
    Section III-B builds LUTs for selected configurations and estimates others by scaling from synthesis data; no predicted-versus-measured validation is shown.
  • ad hoc to paper The heuristic hierarchical search finds designs at the Pareto frontier, or close enough to it.
    Algorithm 1 is a heuristic search with no optimality guarantee, yet the abstract and Section I claim 'optimal architectures'.
  • domain assumption The 40nm test chip measurements represent the performance of designs produced under the general flow.
    Section IV-B validates a single 64x64 MCR=2 macro; other generated designs are only post-layout simulated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SynDCIM: A Performance-Aware Digital Computing-in-Memory Compiler with Multi-Spec-Oriented Subcircuit Synthesis." pith.science (2026). https://pith.science/paper/HORNXZFE

@misc{pith2026241116806,
  author       = {Pith},
  title        = {Pith review of: SynDCIM: A Performance-Aware Digital Computing-in-Memory Compiler with Multi-Spec-Oriented Subcircuit Synthesis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HORNXZFE}},
  note         = {Machine review of arXiv:2411.16806}
}
read the original abstract

Digital Computing-in-Memory (DCIM) is an innovative technology that integrates multiply-accumulation (MAC) logic directly into memory arrays to enhance the performance of modern AI computing. However, the need for customized memory cells and logic components currently necessitates significant manual effort in DCIM design. Existing tools for facilitating DCIM macro designs struggle to optimize subcircuit synthesis to meet user-defined performance criteria, thereby limiting the potential system-level acceleration that DCIM can offer. To address these challenges and enable agile design of DCIM macros with optimal architectures, we present SynDCIM, a performance-aware DCIM compiler that employs multi-spec-oriented subcircuit synthesis. SynDCIM features an automated performance-to-layout generation process that aligns with user-defined performance expectations. This is supported by a scalable subcircuit library and a multi-spec-oriented searching algorithm for effective subcircuit synthesis. The effectiveness of SynDCIM is demonstrated through extensive experiments and validated with a test chip fabricated in a 40nm CMOS process. Testing results reveal that designs generated by SynDCIM exhibit competitive performance when compared to state-of-the-art manually designed DCIM macros.

Figures

Figures reproduced from arXiv: 2411.16806 by the authors.

Figure 1
Figure 1. Illustration of emerging DCIM architectures and key subcir￾cuits. results across columns. Several DCIM macros have imple￾mented memory-compute ratio (MCR)-aware designs to en￾hance on-macro memory density and facilitate efficient weight updates and MAC operations through specialized multiplier and multiplexer circuits. B. DCIM Subcircuits A typical DCIM macro consists of seven key subcircuits: FP&INT Alignment Unit,… view at source ↗
Figure 2
Figure 2. Overall framework of the proposed performance-aware DCIM compiler, SynDCIM. an OAI22 gate as a fused multiplier and multiplexer, saving wiring overhead but becoming less scalable when the memory￾compute ratio (MCR) exceeds 2. (3) [2] uses a 2T transmission gate for selection and a NOR gate for multiplication, which is a commonly adopted approach. Adder tree: The adder tree accumulates partial sums across rows of a c… view at source ↗
Figure 3
Figure 3. Illustration of the proposed DCIM subcircuit library for synthesis [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Proposed bit-wise carry-save-adders: principle and circuit details. compressors, full adders, and half adders. Previous work has shown that bit-wise 4-2 compressor-based CSAs are faster, smaller, and more energy-efficient than conventional signed RCA-based adder trees.…
Figure 7
Figure 7. Figure 7: Post-layout energy efficiency of SynDCIM-generated DCIM macros across different precisions and dimensions. with standard cells, allowing integration into the standard digital flow. During placement by APR tools, cells may be scattered, affecting macro performance. To g…
Figure 6
Figure 6. Figure 6: The proposed implementation and evaluation flow of SynD￾CIM. ing DCIM macros are optimized for the specified performance constraints and PPA preferences. D. Implementation and Evaluation Flow SynDCIM executes the layout implementation through a standard digital flow in…
Figure 10
Figure 10. Figure 10: Die photo of the fabricated 40nm test DCIM macro chip generated by SynDCIM. TABLE II COMPARISON OF THE TEST DCIM MACRO CHIP GENERATED BY SYNDCIM WITH STATE-OF-THE-ART DCIM DESIGNS [PITH_FULL_IMAGE:figures/full_fig_p006_10.png]
Figure 9
Figure 9. Figure 9: Shmoo plot of the test DCIM macro generated by SynDCIM. points at the Pareto frontier. We define the specifications as follows: H=W=64, MCR=2, precisions = INT4/8, FP4/8, MAC and weight update frequency @0.9V = 800MHz. The generated and implemented designs are shown an…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. A Flexible Precision Scaling Deep Neural Network Accelerator with Efficient Weight Combination

    cs.AR 2025-02 conditional novelty 6.0 of 10

    A precision-scalable DNN accelerator uses 2-bit and 3-bit weight decomposition, bit-serial activations, and a split carry-save adder tree to reach 68.94 TOPS/W peak efficiency at 2/2-bit.

Reference graph

Works this paper leans on

14 extracted references · 12 canonical work pages · cited by 1 Pith paper

  1. [1]

    16.4 an 89tops/w and 16.3 tops/mm 2 all-digital sram-based full-precision compute-in memory macro in 22nm for machine-learning edge applications

    Yu-Der Chih, Po-Hao Lee, Hidehiro Fujiwara, Yi-Chun Shih, Chia-Fu Lee, Rawan Naous, Yu-Lin Chen, Chieh-Pu Lo, Cheng-Han Lu, Haruki Mori, et al. 16.4 an 89tops/w and 16.3 tops/mm 2 all-digital sram-based full-precision compute-in memory macro in 22nm for machine-learning edge applications. In 2021 IEEE International Solid-State Circuits Conference (ISSCC),...

    2021

  2. [2]

    A 5-nm 254-tops/w 221-tops/mm 2 fully- digital computing-in-memory macro supporting wide-range dynamic- voltage-frequency scaling and simultaneous mac and write operations

    Hidehiro Fujiwara, Haruki Mori, Wei-Chang Zhao, Mei-Chen Chuang, Rawan Naous, Chao-Kai Chuang, Takeshi Hashizume, Dar Sun, Chia-Fu Lee, Kerem Akarvardar, et al. A 5-nm 254-tops/w 221-tops/mm 2 fully- digital computing-in-memory macro supporting wide-range dynamic- voltage-frequency scaling and simultaneous mac and write operations. In 2022 IEEE Internatio...

    work page 2022

  3. [3]

    A 4nm 6163-tops/w/b 4790-tops/mm2/b sram based digital-computing-in-memory macro supporting bit-width flexibility and simultaneous mac and weight update

    Haruki Mori, Wei-Chang Zhao, Cheng-En Lee, Chia-Fu Lee, Yu-Hao Hsu, Chao-Kai Chuang, Takeshi Hashizume, Hao-Chun Tung, Yao-Yi Liu, Shin-Rung Wu, et al. A 4nm 6163-tops/w/b 4790-tops/mm2/b sram based digital-computing-in-memory macro supporting bit-width flexibility and simultaneous mac and weight update. In 2023 IEEE International Solid-State Circuits Con...

    work page 2023

  4. [4]

    Hidehiro Fujiwara, Haruki Mori, Wei-Chang Zhao, Kinshuk Khare, Cheng-En Lee, Xiaochen Peng, Vineet Joshi, Chao-Kai Chuang, Shu- Huan Hsu, Takeshi Hashizume, et al. 34.4 a 3nm, 32.5 tops/w, 55.0 tops/mm 2 and 3.78 mb/mm 2 fully-digital compute-in-memory macro supporting int12 × int12 with a parallel-mac architecture and foundry 6t-sram bit cell. In 2024 IE...

    work page 2024

  5. [5]

    Autodcim: An automated digital cim compiler

    Jia Chen, Fengbin Tu, Kunming Shao, Fengshi Tian, Xiao Huo, Chi- Ying Tsui, and Kwang-Ting Cheng. Autodcim: An automated digital cim compiler. In 2023 60th ACM/IEEE Design Automation Conference (DAC), pages 1–6. IEEE, 2023

    work page 2023

  6. [6]

    Auto- matic generation of structured macros using standard cells–application to cim

    Christian Lanius, Jie Lou, Johnson Loh, and Tobias Gemmeke. Auto- matic generation of structured macros using standard cells–application to cim. In 2023 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED) , pages 1–6. IEEE, 2023

    work page 2023

  7. [7]

    EasyACIM: An End-to-End Automated Analog CIM with Synthesizable Architecture and Agile Design Space Exploration

    Haoyi Zhang, Jiahao Song, Xiaohan Gao, Xiyuan Tang, Yibo Lin, Runsheng Wang, and Ru Huang. Easyacim: An end-to-end automated analog cim with synthesizable architecture and agile design space exploration. arXiv preprint arXiv:2404.13062 , 2024

    work page Pith review arXiv 2024

  8. [8]

    Arc- tic: Agile and robust compute-in-memory compiler with parameterized int/fp precision and built-in self test

    Hongyi Zhang, Haozhe Zhu, Siqi He, Mengjie Li, Chengchen Wang, Xiankui Xiong, Haidong Tian, Xiaoyang Zeng, and Chixiao Chen. Arc- tic: Agile and robust compute-in-memory compiler with parameterized int/fp precision and built-in self test. In 2024 Design, Automation & Test in Europe Conference & Exhibition (DATE) , pages 1–6. IEEE, 2024

    work page 2024

Show all 14 references
  1. [9]

    Redcim: Reconfigurable digital computing-in-memory processor with unified fp/int pipeline for cloud ai acceleration

    Fengbin Tu, Yiqi Wang, Zihan Wu, Ling Liang, Yufei Ding, Bongjin Kim, Leibo Liu, Shaojun Wei, Yuan Xie, and Shouyi Yin. Redcim: Reconfigurable digital computing-in-memory processor with unified fp/int pipeline for cloud ai acceleration. IEEE Journal of Solid-State Circuits, 58...

    2022

  2. [10]

    All-digital time-domain compute-in-memory engine for binary neural networks with 1.05 pops/w energy efficiency

    Jie Lou, Christian Lanius, Florian Freye, Tim Stadtmann, and Tobias Gemmeke. All-digital time-domain compute-in-memory engine for binary neural networks with 1.05 pops/w energy efficiency. In ESSCIRC 2022-IEEE 48th European Solid State Circuits Conference (ESSCIRC) , pages 149...

    2022

  3. [11]

    Rescim: Variation- resilient high weight-loading bandwidth in-memory computation based on fine-grained hybrid integration of multi-level reram and sram cells

    Xiaomeng Wang, Jingyu He, Kunming Shao, Jiakun Zheng, Fengshi Tian, Tim Kwang-Ting Cheng, and Chi-Ying Tsui. Rescim: Variation- resilient high weight-loading bandwidth in-memory computation based on fine-grained hybrid integration of multi-level reram and sram cells. In Procee...

    2024

  4. [12]

    Cream: computing in reram- assisted energy and area-efficient sram for neural network acceleration

    Liukai Xu, Songyuan Liu, Zhi Li, Dengfeng Wang, Yiming Chen, Yanan Sun, Xueqing Li, Weifeng He, and Shi Xu. Cream: computing in reram- assisted energy and area-efficient sram for neural network acceleration. In Proceedings of the 59th ACM/IEEE Design Automation Conference , pa...

    2022

  5. [13]

    Tl-nvsram-cim: Ultra- high-density three-level reram-assisted computing-in-nvsram with dc- power free restore and ternary mac operations

    Dengfeng Wang, Liukai Xu, Songyuan Liu, Zhi Li, Yiming Chen, Weifeng He, Xueqing Li, and Yanan Sun. Tl-nvsram-cim: Ultra- high-density three-level reram-assisted computing-in-nvsram with dc- power free restore and ternary mac operations. In 2023 IEEE/ACM International Conferen...

    2023

  6. [14]

    A 1–8b reconfigurable digital sram compute-in-memory macro for processing neural networks

    Heng You, Weijun Li, Delong Shang, Yumei Zhou, and Shushan Qiao. A 1–8b reconfigurable digital sram compute-in-memory macro for processing neural networks. IEEE Transactions on Circuits and Systems I: Regular Papers , 2024

    2024

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.