Pith. sign in

REVIEW 4 major objections 7 minor 37 references

Multi-primitive in-memory computing for Monte Carlo tree search

T0 review · 4 major / 7 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A multi-primitive in-memory design executes all four phases of Monte Carlo tree search at about 60 mW for 9-by-9 Go, with energy efficiency 96x better than a CPU and 65x–2,059x better than an H100 GPU.

desk verdict Phase-to-primitive decomposition is a genuine contribution, but the headline 60 mW / 96x numbers are simulation projections whose weakest link is the synthesized digital logic, not the analog crossbar. read the letter →

arxiv 2607.22869 v1 pith:OHGINNZN submitted 2026-07-24 cs.AR cs.AIcs.ET

Tergel Molom-Ochir , Benjamin F. Morris III , Yintao He , Archit Gajjar , Giacomo Pedretti , Hai Helen Li , Yiran Chen , Jim Ignowski
show 1 more author
Aishwarya Natarajan
This is my paper · ORCID
classification cs.ARcs.AIcs.ET
keywords in-memorycomputingMonteCarlotreesearchedgeAIneuromorphichardware-softwareco-designRRAMcrossbarcontent-addressablememoryphase-to-primitivedecomposition
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

Monte Carlo tree search has been considered too irregular for in-memory computing because its four phases—selection, expansion, rollout, and backpropagation—demand qualitatively different operations. This paper claims that apparent incompatibility is not fundamental, introducing phase-to-primitive decomposition, which reformulates each phase as a hardware-native in-memory primitive and composes them into an on-chip pipeline. On 9-by-9 Go, the resulting IMC-MCTS design runs sustained 5,000-iteration-per-move search at roughly 60 mW, with 96x better energy efficiency than a CPU and 65x–2,059x over an H100 GPU depending on batch size. In a 1,050-game tournament at matched compute, it plays at about 1727 EGF, within sample-size uncertainty of the open-source engines Pachi-UCT and Michi-C. If the simulated energy figures survive silicon, the result would move search-based AI from data-center power budgets to milliwatt-scale embedded devices, and the same substrate already runs eight grid-based tasks across four AI domains.

What carries the argument

The carrying object is phase-to-primitive decomposition, a four-step methodology—profile, match, reformulate, compose—that turns a phase-heterogeneous algorithm into a multi-primitive in-memory workload. For MCTS, it pairs each phase with a distinct hardware primitive: a 6T-SRAM content-addressable memory for $\mathcal{O}(1)$ associative node lookup during selection, a combinational Game Logic Unit for single-cycle legal-move expansion, an analog RRAM crossbar with 8-bit ADC and binary 1-bit DAC encodings for the rollout evaluation, and SRAM with co-located saturating counter logic for in-place read-modify-write backpropagation. A TCAM-based finite-state machine orchestrates the four stages with deterministic timing, and the interlocking latencies (constant-time CAM lookup, fixed 768-ns crossbar read, in-place SRAM updates) are what make the end-to-end pipeline stay on chip. The decomposition is load-bearing: removing any one primitive would collapse the end-to-end gain.

What would settle it

Fabricate the 9-by-9 core at a modern node and measure sustained power and per-move energy during 5,000-iteration-per-move search; if measured power is substantially above the simulated ~60 mW or per-move energy is substantially above the simulated ~124 mJ, the 96x efficiency claim falls. A cheaper check would be to measure the CAM lookup latency and the 768-ns crossbar read in a test chip, since the whole pipeline depends on those two timing guarantees.

Watch

Extended reading notes

Core claim

The core claim is that the obstacle to mapping irregular multi-phase algorithms onto in-memory computing is the choice of operating point, not a fundamental limit. The paper shows that Monte Carlo tree search can be executed entirely on co-located memory primitives: content-addressable memory performs selection by returning all children of a node in a single cycle; combinational logic generates legal moves for expansion; an RRAM crossbar carries out the neural rollout at 0.308 pJ per multiply-accumulate; and SRAM performs in-place backpropagation updates. The cycle-accurate, synthesis-based evaluation reports about 60 mW sustained power at 9-by-9 Go, 96x energy efficiency over the CPU baseline, and a rating of roughly 1727 EGF that sits between Pachi-UCT and Michi-C at matched 500-simulation compute. It also reports that the search ensemble absorbs RRAM conductance noise up to $\sigma=10\%$ without degrading win rate, because hundreds of rolled-out evaluations average out per-position perturbations.

Load-bearing premise

The headline energy and power figures assume that the 22-nm numbers from cycle-accurate simulation—digital logic synthesized from a 65-nm library, SRAM modeled by a cache tool, and RRAM timing scaled from a fabricated 180-nm array—accurately predict a real chip.

Editorial extensions

If this is right

  • Search-based decision-making can run at roughly 60 mW for 9-by-9 Go, a power class that supports drones, microcontroller-class planners, and duty-cycled wearable devices.
  • The same unmodified substrate, with only crossbar weights reprogrammed, executes eight grid-based tasks across strategy games, navigation, protein-folding optimization, and puzzles, all showing better play than random-rollout MCTS as measured by the paper's Search Guidance Gain.
  • Energy efficiency over conventional hardware holds across board sizes, with 96x over CPU and 65x–2,059x over the H100 GPU, plus a 3.4x latency advantage over the batched-256 H100 operating point at 9-by-9.
  • The search ensemble makes MCTS robust to analog-substrate noise: conductance noise up to 10% leaves win rates essentially unchanged even though single-position evaluations diverge.
  • Phase-to-primitive decomposition generalizes as a methodology to other control-flow-heavy algorithms whose phases admit primitive-compatible reformulations, such as constraint satisfaction, planning under partial observability, and neuro-symbolic reasoning.

Reading between the lines

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

  • One testable extension the authors leave implicit is whether the same phase-wise mapping works for continuous state spaces; the paper only demonstrates grid-representable tasks, so a natural follow-up is a non-grid domain such as continuous robot motion planning.
  • If the simulated 22-nm numbers hold on silicon, the technology's real ceiling may be training data rather than the analog substrate: the gap between the weak and strong evaluators suggests playing strength tracks evaluator quality, so stronger play should track larger training corpora and larger crossbars.
  • The broader editorial reading is that algorithmic decomposition, rather than ever-larger general-purpose accelerators, may be the scalable path to milliwatt-scale decision hardware for planning and neuro-symbolic workloads.
  • A direct silicon test of the two timing guarantees—single-cycle CAM lookup and the 768-ns crossbar read—would show whether the whole pipeline's fixed-latency composition survives real device variation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper introduces phase-to-primitive decomposition, a four-step methodology (profile, match, reformulate, compose) for mapping phase-heterogeneous algorithms onto co-located in-memory computing primitives. Applied to Monte Carlo tree search, the selection, expansion, rollout, and backpropagation phases are mapped to content-addressable memory, a combinational Game Logic Unit, an RRAM crossbar neural evaluator, and in-place SRAM updates, respectively, under a TCAM-based finite-state-machine controller. At a simulated 22 nm technology node, the resulting IMC-MCTS accelerator is reported to consume roughly 60 mW of sustained power at 9x9 Go, to achieve 96x energy efficiency over a CPU and 65x-2,059x over an H100 GPU, and to reach an EGF tournament rating of about 1727, within sample-size uncertainty of the open-source engines Pachi-UCT and Michi-C at 500 simulations per move. The same parameterized RTL is evaluated on eight grid-based applications across four domains. All IMC energy and power figures come from cycle-accurate SST simulation with RTL synthesized at 65 nm and scaled to 22 nm via Stillmaker, using crossbar parameters from fabricated 180 nm arrays; no 22 nm chip was fabricated.

Significance. If the quantitative results hold, this is a significant contribution: it is, to my knowledge, the first demonstration that an irregular, multi-phase algorithm such as MCTS can be mapped end-to-end onto a multi-primitive in-memory computing substrate, and the paper supports this with unusually detailed RTL, synthesis, simulation, and verification methodology. The decision-quality comparison against external engines, the explicit noise-robustness analysis, the ablation study, and the plan for public code release are all strengths. The central architectural claim, that phase-to-primitive decomposition can make MCTS IMC-compatible, is well supported by the design and by the tournament results at matched 500-simulation compute. However, the headline energy and power figures rest on digital-logic power estimates that contribute roughly 97% of the reported 60.26 mW and have not been validated by silicon measurement; the quantitative efficiency claims should therefore be read as projections until sensitivity analysis or measured data are provided.

major comments (4)
  1. [Methods, 'IMC-MCTS architecture and synthesis'; Supplementary Table S3; Extended Data Table 2] The headline sustained power of 60.26 mW and the resulting 96x energy-efficiency claim depend almost entirely on digital-logic power estimates that have not been validated by measurement: Supplementary Table S3 assigns 14.82 mW to CAM selection, 28.11 mW to GLU expansion, 12.50 mW to FSM control, and 3.03 mW to backprop SRAM out of the 60.26 mW total, with only 1.54 mW attributed to the RRAM crossbar, and Extended Data Table 2 reports all IMC energy values from a single deterministic simulation run with no variance. Because the digital numbers come from Synopsys synthesis at 65 nm scaled to 22 nm via Stillmaker and the crossbar timing comes from a fabricated 180 nm array scaled to 22 nm, an underestimate of digital power or peripheral energy would directly shrink the headline 96x advantage. I request either end-to-end silicon validation or, at minimum, a systematic sensitivity analysis covering PVT corners, alternative scaling assumptions, and a plausible range of digital-power error, with uncertainty intervals reported on all IMC energy and power numbers.
  2. [Methods, 'Tournament evaluation'; Results, 'Decision quality on Go'] The claim that IMC-strong plays at the level of Pachi-UCT is weakened by a training/evaluation overlap: IMC-strong's evaluator was trained by iteratively sparring against Pachi-UCT (10 iterations of about 35,000 positions each, per the Tournament evaluation section), and the same Pachi-UCT is then used as a head-to-head reference in the 1,050-game tournament. This makes the comparison at least partly a measure of how well the hardware mimics its training opponent rather than an independent strength estimate. I recommend either training IMC-strong without Pachi-UCT and comparing against engines not used in training, or explicitly discussing this circularity as a limitation of the relative-ordering claim.
  3. [Results, 'Substrate reusability across grid-based decision tasks'; Methods, 'IMC-MCTS architecture and synthesis' and…] The statement that 'the same hardware' runs all eight applications is stronger than what the methods support: per-application customization is described as a single-pass digital synthesis in which board size and encoding parameters re-size the CAM, GLU, SRAM, and crossbar arrays, and the cross-domain evaluation protocol states that hardware area and energy were computed using the parameterized synthesis model. The accurate claim is that the same parameterized RTL and pipeline are reused across applications, not that one instantiated chip executes all eight tasks. Please reword the text accordingly, including the abstract's 'same substrate' phrasing, and reconcile the statement that only crossbar weights are reprogrammed with the statement that per-application synthesis-time parameters are changed.
  4. [Methods, 'Crossbar device modeling' and 'Tournament evaluation'] There is an inconsistency in the reported training-set sizes: the crossbar modeling section says networks are trained on 2,000 labeled positions per board size, while the tournament section describes IMC-strong as trained over 10 iterations of about 35,000 positions each, which totals roughly 350,000 positions. The distinction matters because the paper attributes IMC-strong's Elo gap over IMC-weak to training-data quality. Please clarify which budget applies to IMC-strong, whether the 2,000-position budget in Extended Data Table 3 refers only to the cross-domain demonstration networks, and how the 2,000-position networks relate to the 60% three-class accuracy reported for 9x9 Go.
minor comments (7)
  1. [Abstract] The phrase '10 170 legal positions' should read '10^170 legal positions'; the superscript formatting is missing in the arXiv rendering.
  2. [Fig. 2e caption] The Low, Medium, and High play-strength labels in Fig. 2e are not defined in the caption; please state that they correspond to the iteration budgets listed in Supplementary Table S1.
  3. [Methods, 'Baselines and energy measurement'; Extended Data Table 2 caption] The energy comparison uses a 5,000-iteration-per-move budget while the tournament uses 500 simulations per move; please make this explicit in the Extended Data Table 2 caption and in the Fig. 3 legend so readers do not confuse the two budgets.
  4. [Supplementary Note 9.3, Eq. S3] The SGG threshold of +0.05 is appropriately described as an indicative policy-lift cutoff rather than a statistical significance boundary, and the smaller lifts are flagged as directional; however, consider reporting per-application game-level variance or confidence intervals so readers can see the uncertainty in values such as +0.09 for Go.
  5. [Abstract and Methods, 'Crossbar device modeling'] The abstract's phrase 'fabricated RRAM-array parameters' could mislead readers into thinking the RRAM arrays were fabricated at 22 nm; please clarify that the arrays were fabricated at 180 nm and scaled to 22 nm, with the scaling assumptions stated.
  6. [Data and code availability] The manuscript promises a Zenodo deposit upon publication but provides no anonymous review link or DOI; since the energy claims are simulation-based, providing the simulator, RTL, and trained weights to reviewers during the review process would materially support verification.
  7. [Results, 'Decision quality on Go'] The phrase 'at matched 500-simulation compute' applies to the MCTS engines but not to KataGo, which uses its default playout schedule; please specify that KataGo is not iso-compute in the main text or figure caption.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the energy, power, and play-strength claims rest on external benchmarks and itemized simulation, not on self-referential fits.

full rationale

The derivation chain is not circular. The central claims are each grounded in external measurements or independent baselines: CPU and GPU energies are instrumented with RAPL and nvidia-smi and reported with variance; the IMC energy is a deterministic cycle-accurate synthesis-based simulation whose component powers are itemized in Supplementary Table S3; Go strength is measured in a 1,050-game round-robin against external engines (KataGo, Pachi-UCT, Michi-C, GnuGo) with Elo calibrated to an external anchor; substrate reusability is quantified by SGG values computed from head-to-head games against a random-rollout baseline. The two concerns flagged in the reader's take are transparency and statistical-power caveats, not circular reductions: the IMC-strong evaluator was trained while sparring against Pachi-UCT before being compared with Pachi-UCT, but the training objective is three-class classification accuracy, not the tournament Elo, and the published Elo is a measured outcome rather than a fitted quantity; the +0.05 SGG cutoff is explicitly described as an indicative policy-lift threshold and not a statistical-significance boundary. No equation in the paper is defined in terms of the quantity it is said to predict, no fitted parameter is renamed as a prediction, and no load-bearing claim rests solely on a self-citation. The crossbar timing and device parameters are cited from a fabricated 180-nm 64x64 RRAM array, which is external, falsifiable evidence rather than a self-referential assertion. The paper is therefore self-contained against external benchmarks, and no specific reduction of a result to its inputs can be exhibited.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

No new physical entities are proposed; the design reuses known primitives (6T-SRAM CAM, 16T TCAM, 1T1R RRAM crossbar, SRAM banks) plus a game-specific combinational unit (GLU). The only free parameter listed is the SGG threshold, which is a hand-chosen cutoff, not a fitted physical constant.

free parameters (1)
  • Search Guidance Gain significance threshold = +0.05
    Chosen as an indicative policy-lift cutoff; the paper states it is not a statistical significance boundary, and at 10-game matchups the standard error of a win-rate lift is roughly 0.15, so the threshold is below sampling noise.
assumptions (5)
  • standard math UCB1 with C=sqrt(2) is an appropriate selection policy for MCTS
    Invoked in the selection phase (Supplementary Note 1.2); standard bandit result from the cited literature.
  • domain assumption A two-layer feedforward MLP of 162-96-3 can evaluate Go positions well enough to play strong kyu level under 500 MCTS simulations
    Supported only by the paper's own simulated tournament, which is within sample-size uncertainty; the claimed 96% validation accuracy for this small network is not independently verified.
  • domain assumption Stillmaker scaling from TSMC 65nm to 22nm and CACTI SRAM modeling give reliable absolute energy and power estimates
    All headline energy figures depend on this; the IMC-NN values come from a single deterministic simulation with no error bars (Methods; Extended Data Table 2).
  • domain assumption Gaussian conductance noise with sigma in {2,5,10}% captures RRAM device non-idealities
    Used for the noise-robustness analysis in Supplementary Note 5; based on prior memristor work and validated only in simulation.
  • domain assumption An 8-bit ADC and binary DAC datapath preserves MCTS decision quality
    Stated as the standard IMC operating point; robustness is shown via simulated noise injection, not silicon measurement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-primitive in-memory computing for Monte Carlo tree search." pith.science (2026). https://pith.science/paper/OHGINNZN

@misc{pith2026260722869,
  author       = {Pith},
  title        = {Pith review of: Multi-primitive in-memory computing for Monte Carlo tree search},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OHGINNZN}},
  note         = {Machine review of arXiv:2607.22869}
}
read the original abstract

Monte Carlo tree search (MCTS) enables artificial intelligence (AI) decision-making, but requires 55-300 W on conventional processors, limiting edge deployment. In-memory computing (IMC) is energy-efficient on regular workloads but has been considered incompatible with irregular multi-phase algorithms. We introduce phase-to-primitive decomposition, which reformulates each algorithmic phase as a hardware-native IMC primitive. Applied to MCTS, selection, expansion, rollout and backpropagation map to content-addressable memory, combinational logic, a resistive random-access memory (RRAM) crossbar and static random-access memory, keeping search on chip. At 22 nm with fabricated RRAM-array parameters, IMC-MCTS consumes ~60 mW for 9x9 Go, achieving 96x energy efficiency over a central processing unit (CPU) and 65x-2,059x over an H100 graphics processing unit (GPU). It reaches a European Go Federation rating within sample-size uncertainty of open-source Go engines (Pachi-UCT and Michi-C). The same substrate runs eight applications across four AI domains.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 24 canonical work pages

  1. [1]

    In: International Conference on Computers and Games, pp

    Coulom,R.:EfficientselectivityandbackupoperatorsinMonte-Carlotreesearch. In: International Conference on Computers and Games, pp. 72–83. Springer, Berlin, Heidelberg (2006)

    work page 2006

  2. [2]

    In: Euro- pean Conference on Machine Learning (ECML), pp

    Kocsis, L., Szepesvári, C.: Bandit based Monte-Carlo planning. In: Euro- pean Conference on Machine Learning (ECML), pp. 282–293. Springer, Berlin, Heidelberg (2006)

    work page 2006

  3. [3]

    In: Computers and Games (CG 2016)

    Tromp, J.: The number of legal Go positions. In: Computers and Games (CG 2016). Lecture Notes in Computer Science, vol. 10068, pp. 183–190. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-50935-8_17

    work page doi:10.1007/978-3-319-50935-8_17 2016

  4. [4]

    Nature529(7587), 484–489 (2016) https://doi.org/10.1038/nature16961

    Silver, D., Huang, A., Maddison, C.J., Guez, A., Sifre, L., Driessche, G., Schrit- twieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., Dieleman, S., Grewe, D., Nham, J., Kalchbrenner, N., Sutskever, I., Lillicrap, T., Leach, M., Kavukcuoglu, K., Graepel, T., Hassabis, D.: Mastering the game of Go with deep neural networks and tree search. Nature...

    doi:10.1038/nature16961 2016

  5. [5]

    Science362(6419), 1140–1144 (2018) https: //doi.org/10.1126/science.aar6404

    Silver, D., Hubert, T., Schrittwieser, J., Antonoglou, I., Lai, M., Guez, A., Lanc- tot, M., Sifre, L., Kumaran, D., Graepel, T., Lillicrap, T.P., Simonyan, K., Hassabis, D.: A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play. Science362(6419), 1140–1144 (2018) https: //doi.org/10.1126/science.aar6404

    doi:10.1126/science.aar6404 2018

  6. [6]

    Integration58, 74–81 (2017)

    Stillmaker, A., Baas, B.: Scaling equations for the accurate prediction of CMOS device performance from 180nm to 7nm. Integration58, 74–81 (2017)

    work page 2017

  7. [7]

    Technical Report HPL-2009-85, HP Labs (2009)

    Muralimanohar, N., Balasubramonian, R., Jouppi, N.P.: CACTI 6.0: A tool to model large caches. Technical Report HPL-2009-85, HP Labs (2009)

    work page 2009

  8. [8]

    IEEE Journal of Solid-State Circuits 41(3), 712–727 (2006)

    Pagiamtzis, K., Sheikholeslami, A.: Content-addressable memory (CAM) circuits and architectures: A tutorial and survey. IEEE Journal of Solid-State Circuits 41(3), 712–727 (2006)

    work page 2006

Show all 37 references
  1. [9]

    In: 2012 Symposium on VLSI Technology (VLSIT), pp

    Auth, C.,et al.: A 22nm high performance and low-power CMOS technology featuring fully-depleted tri-gate transistors, self-aligned contacts and high density MIM capacitors. In: 2012 Symposium on VLSI Technology (VLSIT), pp. 131–132 (2012). https://doi.org/10.1109/VLSIT.2012.6242496

    2012

  2. [10]

    IEEE Journal of Solid-State Circuits51(4), 1009–1021 (2016) https://doi.org/ 10.1109/JSSC.2016.2515510 27

    Jeloka, S., Akesh, N.B., Sylvester, D., Blaauw, D.: A 28nm configurable memory (TCAM/BCAM/SRAM) using push-rule 6T bit cell enabling logic-in-memory. IEEE Journal of Solid-State Circuits51(4), 1009–1021 (2016) https://doi.org/ 10.1109/JSSC.2016.2515510 27

    2016

  3. [11]

    SIGMETRICS Performance Evaluation Review38(4), 37–42 (2011)

    Rodrigues, A.F., Hemmert, K.S., Barrett, B.W., Kersey, C., Oldfield, R., Weston, M., Risen, R., Cook, J., Rosenfeld, P., Cooper-Balis, E., Jacob, B.: The structural simulation toolkit. SIGMETRICS Performance Evaluation Review38(4), 37–42 (2011)

    2011

  4. [12]

    Advanced Materials30(9), 1705914 (2018)

    Hu, M., Graves, C.E., Li, C., Li, Y., Ge, N., Montgomery, E., Dávila, N., Jiang, H., Williams, R.S., Yang, J.J., Xia, Q., Strachan, J.P.: Memristor-based ana- log computation and neural network classification with a dot product engine. Advanced Materials30(9), 1705914 (2018)

    2018

  5. [13]

    Advanced Electronic Materials5(9), 1800876 (2019)

    Sheng, X., Graves, C.E., Kumar, S., Li, X., Buchanan, B., Zheng, L., Lam, S., Li, C., Strachan, J.P.: Low-conductance and multilevel CMOS-integrated nanoscale oxide memristors. Advanced Electronic Materials5(9), 1800876 (2019)

    2019

  6. [14]

    In: 2020 IEEE International Memory Work- shop (IMW), pp

    Li, C., Ignowski, J., Sheng, X., Wessel, R., Jaffe, B., Ingemi, J., Graves, C., Strachan, J.P.: CMOS-integrated nanoscale memristive crossbars for CNN and optimization acceleration. In: 2020 IEEE International Memory Work- shop (IMW), pp. 1–4. IEEE, Piscataway, NJ (2020). http...

    2020

  7. [15]

    Nature 608, 504–512 (2022) https://doi.org/10.1038/s41586-022-04992-8

    Wan, W., Kubendran, R., Schaefer, C., Eryilmaz, S.B., Zhang, W., Wu, D., Deiss, S., Raina, P., Qian, H., Gao, B., Joshi, S., Wu, H., Wong, H.-S.P., Cauwenberghs, G.: A compute-in-memory chip based on resistive random-access memory. Nature 608, 504–512 (2022) https://doi.org/10...

    2022 doi

  8. [16]

    ACM Transactions on Modeling and Performance Evaluation of Computing Systems3(2), 9 (2018) https://doi.org/10.1145/3177754

    Khan, K.N., Hirki, M., Niemi, T., Nurminen, J.K., Ou, Z.: RAPL in action: Experiences in using RAPL for power measurements. ACM Transactions on Modeling and Performance Evaluation of Computing Systems3(2), 9 (2018) https://doi.org/10.1145/3177754

    2018 doi

  9. [17]

    SoftwareX30, 102139 (2025) https: //doi.org/10.1016/j.softx.2025.102139

    Klęsk, P.: MCTS-NC: A thorough GPU parallelization of Monte-Carlo tree search implemented in python via numba.cuda. SoftwareX30, 102139 (2025) https: //doi.org/10.1016/j.softx.2025.102139

    2025

  10. [18]

    Shafiee, A., Nag, A., Muralimanohar, N., Balasubramonian, R., Strachan, J.P., Hu, M., Williams, R.S., Srikumar, V.: ISAAC: A convolutional neural network acceleratorwithin-situanalogarithmeticincrossbars.In:Proceedingsofthe43rd International Symposium on Computer Architecture ...

    2016

  11. [19]

    IEEE Journal of Solid-State Circuits52(1), 127–138 (2017) https://doi.org/10.1109/JSSC.2016

    Chen, Y.-H., Krishna, T., Emer, J.S., Sze, V.: Eyeriss: An energy-efficient recon- figurable accelerator for deep convolutional neural networks. IEEE Journal of Solid-State Circuits52(1), 127–138 (2017) https://doi.org/10.1109/JSSC.2016. 2616357

    2017 doi

  12. [20]

    ACM SIGARCH Computer Architecture News45(2), 1–12 (2017) https://doi.org/10.1145/3140659.3080246

    Jouppi, N.P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A.,et al.: In-datacenter performance analysis 28 of a tensor processing unit. ACM SIGARCH Computer Architecture News45(2), 1–12 (2017) https://doi.org/10.1145...

    2017

  13. [21]

    IEEE Transactions on Cir- cuits and Systems I: Regular Papers (2025) https://doi.org/10.1109/TCSI.2025

    Abdelmaksoud, A.J., Agwa, S., Prodromakis, T.: DiP: A scalable, energy-efficient systolic array for matrix multiplication acceleration. IEEE Transactions on Cir- cuits and Systems I: Regular Papers (2025) https://doi.org/10.1109/TCSI.2025. 3591960

    2025 doi

  14. [22]

    In: Proceedings of the 2023 ACM/SIGDA International Symposium on Field Programmable Gate Arrays (FPGA), pp

    Meng, Y., Kannan, R., Prasanna, V.: A framework for Monte-Carlo tree search on CPU–FPGA heterogeneous platform via on-chip dynamic tree management. In: Proceedings of the 2023 ACM/SIGDA International Symposium on Field Programmable Gate Arrays (FPGA), pp. 235–245. ACM, New Yor...

    2023

  15. [23]

    In: 2022 32nd International Confer- ence on Field-Programmable Logic and Applications (FPL), pp

    Meng, Y., Kannan, R., Prasanna, V.K.: Accelerating Monte-Carlo tree search on CPU–FPGA heterogeneous platform. In: 2022 32nd International Confer- ence on Field-Programmable Logic and Applications (FPL), pp. 298–305. IEEE, Piscataway, NJ, USA (2022)

    2022

  16. [24]

    Master’s thesis, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal (2019)

    Carvalho, R.R.: SoC–FPGA Monte Carlo tree search processor. Master’s thesis, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, Portugal (2019)

    2019

  17. [25]

    In: 2014 International Conference on Field- Programmable Technology (FPT), pp

    Qasemi, E., Samadi, A., Shadmehr, M.H., Azizian, B., Mozaffari, S., Shirian, A., Alizadeh, B.: Highly scalable, shared-memory, Monte-Carlo tree search based Blokus Duo solver on FPGA. In: 2014 International Conference on Field- Programmable Technology (FPT), pp. 370–373. IEEE,...

    2014

  18. [26]

    In: International Con- ference on Learning and Intelligent Optimization (LION)

    Buzer, L., Cazenave, T.: GPU for Monte Carlo search. In: International Con- ference on Learning and Intelligent Optimization (LION). Lecture Notes in Computer Science, vol. 14286, pp. 179–193. Springer, Cham (2023). https://doi. org/10.1007/978-3-031-44505-7_13

    2023 doi

  19. [27]

    In: 2022 3rd International Conference on Artificial Intelligence, Information Processing and Cloud Computing (AIIPCC), pp

    Sun, Y., Yuan, D., Gao, M., Zhu, P.: GPU acceleration of Monte Carlo tree search algorithm for Amazon chess and its evaluation function. In: 2022 3rd International Conference on Artificial Intelligence, Information Processing and Cloud Computing (AIIPCC), pp. 434–440. IEEE, Pi...

    2022

  20. [28]

    Available online (2022)

    Thomas, G.: A Full-GPU Implementation of MCTS in Julia: The Key to Gumbel MuZero? Google Summer of Code 2022 project report (AlphaZero.jl). Available online (2022)

    2022

  21. [29]

    In: 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp

    Huang, B., Boularias, A., Yu, J.: Parallel Monte Carlo tree search with batched rigid-body simulations for speeding up long-horizon episodic robot planning. In: 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 1–8. IEEE, Piscataway, NJ, USA ...

    2022

  22. [30]

    In: Latifi, S

    Johnson, C., Barford, L., Dascalu, S.M., Harris, F.C.: CUDA implementation of computer Go game tree search. In: Latifi, S. (ed.) Information Technology: New Generations, pp. 339–350. Springer, Cham (2016)

    2016

  23. [31]

    In: 2011 IEEE International Symposium on Parallel and Distributed Processing Workshops and PhD Forum, pp

    Rocki, K., Suda, R.: Large-scale parallel Monte-Carlo tree search on GPU. In: 2011 IEEE International Symposium on Parallel and Distributed Processing Workshops and PhD Forum, pp. 2034–2037. IEEE, Piscataway, NJ, USA (2011)

    2011

  24. [32]

    In: 2024 International Conference on Neuromorphic Systems (ICONS), pp

    Ho, Y., Carbajal, A., Escamilla, L., Pinar, A.: Neuromorphic Monte Carlo tree search methods for shortest path interdiction. In: 2024 International Conference on Neuromorphic Systems (ICONS), pp. 307–311. IEEE, Piscataway, NJ, USA (2024). https://doi.org/10.1109/ICONS62911.2024.00053

    2024

  25. [33]

    IEEE Transactions on Games13(3), 315–320 (2021) https://doi.org/10

    Steinmetz, E., Gini, M.: More trees or larger trees: Parallelizing Monte Carlo tree search. IEEE Transactions on Games13(3), 315–320 (2021) https://doi.org/10. 1109/TG.2020.3048331

    2021

  26. [34]

    Fujiwara, H., Mori, H., Zhao, W.-C., Chuang, M.-C., Naous, R., Chuang, C.-K., Hashizume, T., Sun, D., Lee, C.-F., Akarvardar, K., Adham, S., Chou, T.-L., Sinangil, M.E., Wang, Y., Chih, Y.-D., Chen, Y.-H., Liao, H.-J., Chang, T.-Y.J.: A5-nm254-TOPS/W221-TOPS/mm 2 fully-digital...

    2022

  27. [35]

    In: Proceedings of the 41st International Conference on Machine Learning (ICML)

    Geng, Z., Wang, J., Liu, Z., Xu, S., Tang, Z., Yuan, M., Hao, J., Zhang, Y., Wu, F.: Reinforcement learning within tree search for fast macro placement. In: Proceedings of the 41st International Conference on Machine Learning (ICML). Proceedings of Machine Learning Research, v...

    2024

  28. [36]

    In: 2016 IEEE Intelligent Vehicles Symposium (IV), pp

    Lenz, D., Kessler, T., Knoll, A.: Tactical cooperative planning for autonomous highway driving using Monte-Carlo tree search. In: 2016 IEEE Intelligent Vehicles Symposium (IV), pp. 447–453. IEEE, Piscataway, NJ, USA (2016). https://doi. org/10.1109/IVS.2016.7535424

    2016

  29. [37]

    MIT Press, Cambridge, MA, USA (2005) 30

    Thrun, S., Burgard, W., Fox, D.: Probabilistic Robotics. MIT Press, Cambridge, MA, USA (2005) 30

    2005

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.