[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81939-en":3,"doc-seo-81939-105":30,"detail-sidebar-cat-0-en-105":84},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":20,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},81939,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","NEMESIS NEtlist Driven Modeling and Equation Synthesis with Inversion Aware SPICE Anchoring","NEMESIS is a multimodal framework for operational transconductance amplifier (OTA) design that uses large language models to synthesize performance equations from a given OTA netlist and schematic. The method balances rapid approximate analytical models with SPICE-verified accuracy by progressively generating equations, retrieving structurally similar prior OTA equations when available, and otherwise deriving initial equations directly from circuit inputs. A SPICE-based repair loop iteratively refines models in a 65nm PDK, achieving \u003C7% average relative error and about 4622× speedup over full SPICE evaluation across biasing ranges.","NEMESIS: NEtlist-Driven Modeling and Equation Synthesis with Inversion-Aware SPICE Anchoring  \nSubhadip Ghosh, Ramesh Harjani, and Sachin S. Sapatnekar  \nDepartment of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, USA  \narXiv :2607 .05657v 1 [ cs .AR] 6 Jul 2026  \nAbstract—This work presents NEMESIS, a multimodal framework for operational transconductance amplifier (OTA) design using large language models (LLMs). NEMESIS strikes a balance between fast, approximate analytical models vs. accurate, computationally expensive SPICE evaluations. Given an OTA netlist and schematic, NEMESIS first identifies circuit primitivesand then generates progressively more accurate performance equations. The framework begins with equations retrieved from the prior invocations of NEMESIS to structurally similar OTAs, if available; otherwise, it uses the LLM to derive the initial equations directly from the circuit input. These equations are iteratively refined via a SPICE-based repair loop. In a commercial 65nm PDK, NEMESIS is demonstrated on five OTA topologies, producing SPICE-verified equations across biasing ranges with \u003C 7% average relative error and a post-convergence evaluation speedup of ≈ 4622 × over full SPICE-based evaluation.  \nI. INTRODUCTION  \nOperational transconductance amplifiers (OTAs) are widely used in sensing, amplification, and data conversion, yet their design remains manual and iterative. Improving one performance target often perturbs others, and the performance bottleneck that prevents the design from meeting all specifications can shift with circuit topology, bias, and target requirements. This coupling motivates automated sizing and design-space exploration (DSE), which must evaluate many design candidates and bias points. Since each candidate is assessed across multiple metrics, repeated SPICE-based evaluation becomes expensive throughout the optimization loop. The performance model used in this loop, therefore, directly determines both runtime and design insight. Fast, explicit models can reduce this cost by linking device parameters to circuit-level performance, but generating such models remains difficult due to topology-specific dependencies and bias/small-signal effects.  \nConventional approaches typically use one of two model classes for circuit-performance evaluation. The first uses explicit analytical equations derived from circuit structure. These models are fast and interpretable, but they are handcrafted, topology-specific, and often inaccurate in modern processes. They include equation-driven models used in [1]–[6], symbolic models invoked in [7], and posynomial models used in geometric programming [8]–[10] . The second class directly uses SPICE for performance evaluation. These evaluations are accurate and avoid analytical simplifications, but are computationally expensive and provide limited circuit-level insight. This class has been used in optimizers based on  \nBayesian optimization [11]–[14], machine learning (ML) [15], This work was supported in part by the NSF under award 2212345 .  \n[16], and reinforcement learning [17]–[20] . We aim to bridge these model classes with fast, interpretable equations that are SPICE-verified under a prescribed error threshold.  \nBeyond facilitating automated optimization, a second motivation for our work is that better models could also aid manual DSE. For manual use, a model should be interpretable, with terms traceable to specific devices, nodes, and small-signal contributions rather than being hidden inside a black-box predictor. Improved models link predicted behavior to circuit structure and aid both manual and automated optimization.  \nOur solution is based on using large language models (LLMs), which have been deployed in analog design workflows via copilots [21], search-oriented agents [22], [23], and multi-agent assistants [24]–[26] . LLMs leverage broad engineering knowledge to flexibly infer relationships among device paramet","cbCaigXxEFOd6IEE","https://ap.wps.com/l/cbCaigXxEFOd6IEE","pdf",1588504,7,1,"English","en",105,"# Introduction\n## Motivation and challenge\n## Related work: analytical models vs. SPICE evaluation\n## Role of LLMs and circuit primitive decomposition\n## Goals and contributions of NEMESIS","[{\"question\":\"What results does NEMESIS report in a 65nm PDK evaluation?\",\"answer\":\"Across five OTA topologies and multiple biasing ranges, NEMESIS produces SPICE-verified equations with less than 7% average relative error and an approximately 4622× evaluation speedup versus full SPICE-based evaluation.\"}]","NEMESIS NEtlist Driven Modeling and Equation Synthesis with Inversion Aware SPICE Anchoring | PDF",1784177169,18,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":79,"head_meta":81,"extra_data":83,"updated_unix":28},"nemesis-netlist-driven-modeling-and-equation-synthesis-with-inversion-aware-spice-anchoring","",{"@graph":36,"@context":78},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/nemesis-netlist-driven-modeling-and-equation-synthesis-with-inversion-aware-spice-anchoring/81939/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-07-30","2026-07-16",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72],{"name":73,"@type":74,"acceptedAnswer":75},"What results does NEMESIS report in a 65nm PDK evaluation?","Question",{"text":76,"@type":77},"Across five OTA topologies and multiple biasing ranges, NEMESIS produces SPICE-verified equations with less than 7% average relative error and an approximately 4622× evaluation speedup versus full SPICE-based evaluation.","Answer","https://schema.org",{"og:url":52,"og:type":80,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":82,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":85},[86,90,94,98,103,108,112,115,120,123,127],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":87,"show_sort_weight":88,"slug":89},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":91,"show_sort_weight":92,"slug":93},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Exam",70,"exam",{"id":99,"doc_module":4,"doc_module_name":46,"category_name":100,"show_sort_weight":101,"slug":102},5,"Comic",60,"comic",{"id":104,"doc_module":4,"doc_module_name":46,"category_name":105,"show_sort_weight":106,"slug":107},6,"Technology",50,"technology",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":109,"show_sort_weight":110,"slug":111},"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":113,"slug":114},30,"research-report",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},9,"Religion & Spirituality",20,"religion-spirituality",{"id":118,"doc_module":4,"doc_module_name":46,"category_name":121,"show_sort_weight":118,"slug":122},"World Cup","world-cup",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":124,"slug":126},10,"Lifestyle","lifestyle",{"id":128,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":99,"slug":130},19,"General","general"]