{"id":1188,"date":"2025-12-22T09:58:32","date_gmt":"2025-12-22T00:58:32","guid":{"rendered":"https:\/\/rtlearner.com\/?p=1188"},"modified":"2025-12-22T09:58:34","modified_gmt":"2025-12-22T00:58:34","slug":"rram-7-stdp-spike-timing-dependent-plasticity","status":"publish","type":"post","link":"https:\/\/rtlearner.com\/en\/rram-7-stdp-spike-timing-dependent-plasticity\/","title":{"rendered":"About RRAM \u2013 7 STDP (Spike-Timing-Dependent Plasticity)"},"content":{"rendered":"<p class=\"wp-block-paragraph\">When studying, we redline important information (strengthening) and forget unimportant information (weakening). The same goes for the synapses in our brains. So, what exactly do synapses use as a criterion to strengthen or weaken connections? The secret rule discovered by neuroscientists is STDP (Spike-Timing-Dependent Plasticity).<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">In this article, we'll explore the principles of implementing this complex biological learning rule in hardware using RRAM devices. This is the foundation of the Spiking Neural Network (SNN).<\/p>\n\n\n<style>.kb-table-of-content-nav.kb-table-of-content-id1188_cace77-80 .kb-table-of-content-wrap{padding-top:var(--global-kb-spacing-sm, 1.5rem);padding-right:var(--global-kb-spacing-sm, 1.5rem);padding-bottom:var(--global-kb-spacing-sm, 1.5rem);padding-left:var(--global-kb-spacing-sm, 1.5rem);box-shadow:0px 0px 14px 0px rgba(0, 0, 0, 0.2);}.kb-table-of-content-nav.kb-table-of-content-id1188_cace77-80 .kb-table-of-contents-title-wrap{padding-top:0px;padding-right:0px;padding-bottom:0px;padding-left:0px;}.kb-table-of-content-nav.kb-table-of-content-id1188_cace77-80 .kb-table-of-contents-title{font-weight:regular;font-style:normal;}.kb-table-of-content-nav.kb-table-of-content-id1188_cace77-80 .kb-table-of-content-wrap .kb-table-of-content-list{font-weight:regular;font-style:normal;margin-top:var(--global-kb-spacing-sm, 1.5rem);margin-right:0px;margin-bottom:0px;margin-left:0px;}@media all and (max-width: 767px){.kb-table-of-content-nav.kb-table-of-content-id1188_cace77-80 .kb-table-of-contents-title{font-size:var(--global-kb-font-size-md, 1.25rem);}.kb-table-of-content-nav.kb-table-of-content-id1188_cace77-80 .kb-table-of-content-wrap .kb-table-of-content-list{font-size:var(--global-kb-font-size-sm, 0.9rem);}}<\/style>\n\n<style>.kadence-column1188_d00b8e-dd > .kt-inside-inner-col{box-shadow:inset 0px 0px 14px 0px rgba(0, 0, 0, 0.2);border-top:0px solid transparent;border-right:0px solid transparent;border-bottom:0px solid transparent;border-left:0px solid transparent;}.kadence-column1188_d00b8e-dd > .kt-inside-inner-col,.kadence-column1188_d00b8e-dd > .kt-inside-inner-col:before{border-top-left-radius:0px;border-top-right-radius:0px;border-bottom-right-radius:0px;border-bottom-left-radius:0px;}.kadence-column1188_d00b8e-dd > .kt-inside-inner-col{column-gap:var(--global-kb-gap-sm, 1rem);}.kadence-column1188_d00b8e-dd > .kt-inside-inner-col{flex-direction:column;}.kadence-column1188_d00b8e-dd > .kt-inside-inner-col > .aligncenter{width:100%;}.kadence-column1188_d00b8e-dd > .kt-inside-inner-col:before{opacity:0.3;}.kadence-column1188_d00b8e-dd{position:relative;}@media all and (max-width: 1024px){.kadence-column1188_d00b8e-dd > .kt-inside-inner-col{border-top:0px solid transparent;border-right:0px solid transparent;border-bottom:0px solid transparent;border-left:0px solid transparent;flex-direction:column;justify-content:center;}}@media all and (max-width: 767px){.kadence-column1188_d00b8e-dd > .kt-inside-inner-col{border-top:0px solid transparent;border-right:0px solid transparent;border-bottom:0px solid transparent;border-left:0px solid transparent;flex-direction:column;justify-content:center;}}<\/style>\n<div class=\"wp-block-kadence-column kadence-column1188_d00b8e-dd\"><div class=\"kt-inside-inner-col\">\n<p class=\"wp-block-paragraph\"><strong>Related articles<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-1-mechanism\/\">About RRAM - 1 Operation mechanism<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-2-property\/\">About RRAM - 2 Properties<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-3-crossbar-array\/\">About RRAM - 3 Crossbar array and Sneak Path Current<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-4-forming-compliance-current\/\">About RRAM \u2013 4 Forming and Compliance Current<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-5-impedance-matching\/\">About RRAM \u2013 5 Pulse measurement and impedance matching<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-6-oxram-cbram\/\" data-type=\"post\" data-id=\"1173\">About RRAM \u2013 6 Filament Materials: OxRAM vs. CBRAM<\/a><\/p>\n<\/div><\/div>\n\n\n\n<h2 class=\"wp-block-heading\">1. Hebbian Learning<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">To understand STDP, you must first understand \u2018Hebb\u2019s Rule\u2019.<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"wp-block-paragraph\"><em>\u201cCells that fire together, wire together.\u201d<\/em><\/p>\n<\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">That is, when the cause (pre-synapse, input) and the result (post-synapse, output) are activated at similar times, the brain judges, \u201cOh, these two are related!\u201d and strengthens the weight.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">STDP adds a sophisticated concept called \u2018Timing\u2019 to this.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">2. What is STDP?: Timing is everything.<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">STDP adjusts synaptic weights based on the order in which signals come in, i.e., \u0394t (time difference).<\/p>\n\n\n\n<div class=\"wp-block-math\"><math display=\"block\"><semantics><mrow><mrow><mi mathvariant=\"normal\">\u0394<\/mi><\/mrow><mi>t<\/mi><mo>=<\/mo><msub><mi>t<\/mi><mrow><mi>p<\/mi><mi>o<\/mi><mi>s<\/mi><mi>t<\/mi><\/mrow><\/msub><mo>\u2212<\/mo><msub><mi>t<\/mi><mrow><mi>p<\/mi><mi>r<\/mi><mi>e<\/mi><\/mrow><\/msub><\/mrow><annotation encoding=\"application\/x-tex\">\\Delta t = t_{post} &#8211; t_{pre}<\/annotation><\/semantics><\/math><\/div>\n\n\n\n<h3 class=\"wp-block-heading\">\u2460 LTP (Long-Term Potentiation): <strong>\u0394<\/strong>t &gt; 0<\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Situation:<\/strong> The output signal (Post) occurs <strong>immediately after<\/strong> the input signal (Pre).<\/li>\n\n\n\n<li><strong>Analysis:<\/strong> \u201cThe input causes the output!\u201d (Causality established)<\/li>\n\n\n\n<li class=\"translation-block\"><strong>Result:<\/strong> Strengthens synaptic connections (reduces component resistance)<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\">\u2461 LTD (Long-Term Depression): <strong>\u0394<\/strong>t &lt; 0<\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Situation:<\/strong> The output signal (Post) goes out first, and the input signal (Pre) <strong>comes in later<\/strong> .<\/li>\n\n\n\n<li><strong>Analysis:<\/strong> \"This input has nothing to do with the result?\" (No causal relationship)<\/li>\n\n\n\n<li class=\"translation-block\"><strong>Result:<\/strong> Disconnects synaptic connections (increases device resistance)<\/li>\n<\/ul>\n\n\n<style>.kb-image1188_fa7b4a-3e.kb-image-is-ratio-size, .kb-image1188_fa7b4a-3e .kb-image-is-ratio-size{max-width:700px;width:100%;}.wp-block-kadence-column > .kt-inside-inner-col > .kb-image1188_fa7b4a-3e.kb-image-is-ratio-size, .wp-block-kadence-column > .kt-inside-inner-col > .kb-image1188_fa7b4a-3e .kb-image-is-ratio-size{align-self:unset;}.kb-image1188_fa7b4a-3e figure{max-width:700px;}.kb-image1188_fa7b4a-3e .image-is-svg, .kb-image1188_fa7b4a-3e .image-is-svg img{width:100%;}.kb-image1188_fa7b4a-3e .kb-image-has-overlay:after{opacity:0.3;}@media all and (max-width: 767px){.kb-image1188_fa7b4a-3e.kb-image-is-ratio-size, .kb-image1188_fa7b4a-3e .kb-image-is-ratio-size{max-width:280px;width:100%;}.kb-image1188_fa7b4a-3e figure{max-width:280px;}}<\/style>\n<div class=\"wp-block-kadence-image kb-image1188_fa7b4a-3e\"><figure class=\"aligncenter size-large\"><img data-dominant-color=\"f9f8f9\" data-has-transparency=\"false\" style=\"--dominant-color: #f9f8f9;\" loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"761\" src=\"https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-2-1024x761.jpg\" alt=\"STDP\" class=\"kb-img wp-image-1190 not-transparent\" srcset=\"https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-2-1024x761.jpg 1024w, https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-2-300x223.jpg 300w, https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-2-768x571.jpg 768w, https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-2.jpg 1200w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><figcaption>STDP<\/figcaption><\/figure><\/div>\n\n\n\n<h2 class=\"wp-block-heading\">3. Implementing STDP with RRAM: Waveform Overlapping Technique<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">RRAM is a two-terminal device. It doesn't have a complex timing calculator like the brain. So how does it detect who arrived first and adjust its resistance? The secret lies in the pulse shape (waveform shaping).<\/p>\n\n\n<style>.kb-image1188_434388-6a.kb-image-is-ratio-size, .kb-image1188_434388-6a .kb-image-is-ratio-size{max-width:600px;width:100%;}.wp-block-kadence-column > .kt-inside-inner-col > .kb-image1188_434388-6a.kb-image-is-ratio-size, .wp-block-kadence-column > .kt-inside-inner-col > .kb-image1188_434388-6a .kb-image-is-ratio-size{align-self:unset;}.kb-image1188_434388-6a figure{max-width:600px;}.kb-image1188_434388-6a .image-is-svg, .kb-image1188_434388-6a .image-is-svg img{width:100%;}.kb-image1188_434388-6a .kb-image-has-overlay:after{opacity:0.3;}@media all and (max-width: 767px){.kb-image1188_434388-6a.kb-image-is-ratio-size, .kb-image1188_434388-6a .kb-image-is-ratio-size{max-width:280px;width:100%;}.kb-image1188_434388-6a figure{max-width:280px;}}<\/style>\n<div class=\"wp-block-kadence-image kb-image1188_434388-6a\"><figure class=\"aligncenter size-full\"><a href=\"https:\/\/www.researchgate.net\/figure\/llustration-of-the-STDP-rule-and-how-the-resistive-state-of-the-RRAM-device-changes_fig3_319131584\" class=\"kb-advanced-image-link\" target=\"_blank\" rel=\"noopener\"><img data-dominant-color=\"e8e8e9\" data-has-transparency=\"false\" style=\"--dominant-color: #e8e8e9;\" loading=\"lazy\" decoding=\"async\" width=\"850\" height=\"751\" src=\"https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-3.jpg\" alt=\"\" class=\"kb-img wp-image-1191 not-transparent\" srcset=\"https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-3.jpg 850w, https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-3-300x265.jpg 300w, https:\/\/rtlearner.com\/wp-content\/uploads\/2025\/12\/image-10-3-768x679.jpg 768w\" sizes=\"auto, (max-width: 850px) 100vw, 850px\" \/><\/a><figcaption>STDP pulse<\/figcaption><\/figure><\/div>\n\n\n\n<h3 class=\"wp-block-heading\">Operation mechanism (Superposition)<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The actual voltage across the device (V<sub>net<\/sub>) is the difference between the upper electrode voltage (V<sub>top<\/sub>) and the lower electrode voltage (V<sub>bottom<\/sub>).<\/p>\n\n\n\n<div class=\"wp-block-math\"><math display=\"block\"><semantics><mrow><msub><mi>V<\/mi><mrow><mi>n<\/mi><mi>e<\/mi><mi>t<\/mi><\/mrow><\/msub><mo>=<\/mo><msub><mi>V<\/mi><mrow><mi>t<\/mi><mi>o<\/mi><mi>p<\/mi><\/mrow><\/msub><mo>\u2212<\/mo><msub><mi>V<\/mi><mrow><mi>b<\/mi><mi>o<\/mi><mi>t<\/mi><mi>t<\/mi><mi>o<\/mi><mi>m<\/mi><\/mrow><\/msub><\/mrow><annotation encoding=\"application\/x-tex\">V_{net} = V_{top} &#8211; V_{bottom}\n<\/annotation><\/semantics><\/math><\/div>\n\n\n\n<ol start=\"1\" class=\"wp-block-list\">\n<li><strong>When the time difference is appropriate (Pre -&gt; Post):<\/strong>\n<ul class=\"wp-block-list\">\n<li>When two pulses overlap, the voltage difference exceeds the set voltage (V<sub>set<\/sub>). -&gt; The element resistance decreases (LTP).<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>When the time difference is reversed (Post -&gt; Pre):<\/strong>\n<ul class=\"wp-block-list\">\n<li>When two pulses overlap, the voltage difference exceeds the Reset Voltage (V<sub>reset<\/sub>). -&gt; Increase in element resistance (LTD).<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>When the time difference is too large:<\/strong>\n<ul class=\"wp-block-list\">\n<li>The voltage is low because the pulses do not meet each other -&gt; no change in element resistance.<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n\n\n\n<p class=\"wp-block-paragraph translation-block\">With just these simple laws of physics, RRAM learns on its own without any separate CPU computation.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">4. Why is it important? (Unsupervised Learning)<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Conventional deep learning (DNN) is supervised learning, which provides the correct answer (label) and calculates and updates the error. This requires enormous computational effort and consumes significant power.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">However, RRAM (SNN) unsupervised learning with STDP is possible.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Even without being told the correct answer, it automatically strengthens the connection between \u201csignals that frequently come in together\u201d by simply looking at the temporal patterns of the data.<\/li>\n\n\n\n<li>This is most similar to how the human brain learns about the world as a baby, and is a key technology for implementing <strong>ultra-low-power edge AI<\/strong> .<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">5. Conclusion: The most brain-like memory<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The real reason RRAM is being called the next-generation AI device isn't simply because it has analog resistance. It's because it offers the most elegant and simple hardware implementation of the <strong>STDP learning rule<\/strong>, which adjusts connection strength solely by pulse timing.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">As RRAM research deepens, we will begin to blur the lines between studying semiconductors and studying the brain. <strong>Neuromorphic engineering<\/strong>lies at the very edge of this divide.<\/p>\n\n\n<style>.kadence-column1188_6f16ce-63 > .kt-inside-inner-col{box-shadow:inset 0px 0px 14px 0px rgba(0, 0, 0, 0.2);border-top:0px solid transparent;border-right:0px solid transparent;border-bottom:0px solid transparent;border-left:0px solid transparent;}.kadence-column1188_6f16ce-63 > .kt-inside-inner-col,.kadence-column1188_6f16ce-63 > .kt-inside-inner-col:before{border-top-left-radius:0px;border-top-right-radius:0px;border-bottom-right-radius:0px;border-bottom-left-radius:0px;}.kadence-column1188_6f16ce-63 > .kt-inside-inner-col{column-gap:var(--global-kb-gap-sm, 1rem);}.kadence-column1188_6f16ce-63 > .kt-inside-inner-col{flex-direction:column;}.kadence-column1188_6f16ce-63 > .kt-inside-inner-col > .aligncenter{width:100%;}.kadence-column1188_6f16ce-63 > .kt-inside-inner-col:before{opacity:0.3;}.kadence-column1188_6f16ce-63{position:relative;}@media all and (max-width: 1024px){.kadence-column1188_6f16ce-63 > .kt-inside-inner-col{border-top:0px solid transparent;border-right:0px solid transparent;border-bottom:0px solid transparent;border-left:0px solid transparent;flex-direction:column;justify-content:center;}}@media all and (max-width: 767px){.kadence-column1188_6f16ce-63 > .kt-inside-inner-col{border-top:0px solid transparent;border-right:0px solid transparent;border-bottom:0px solid transparent;border-left:0px solid transparent;flex-direction:column;justify-content:center;}}<\/style>\n<div class=\"wp-block-kadence-column kadence-column1188_6f16ce-63\"><div class=\"kt-inside-inner-col\">\n<p class=\"wp-block-paragraph\"><strong>Related articles<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-1-mechanism\/\">About RRAM - 1 Operation mechanism<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-2-property\/\">About RRAM - 2 Properties<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-3-crossbar-array\/\">About RRAM - 3 Crossbar array and Sneak Path Current<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-4-forming-compliance-current\/\">About RRAM \u2013 4 Forming and Compliance Current<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-5-impedance-matching\/\">About RRAM \u2013 5 Pulse measurement and impedance matching<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2705<a href=\"https:\/\/rtlearner.com\/en\/rram-6-oxram-cbram\/\" data-type=\"post\" data-id=\"1173\">About RRAM \u2013 6 Filament Materials: OxRAM vs. CBRAM<\/a><\/p>\n<\/div><\/div>\n\n\n\n<p class=\"wp-block-paragraph\">References: <em><a href=\"https:\/\/pubs.acs.org\/doi\/10.1021\/nl904092h\" target=\"_blank\" rel=\"noopener\">Nanoscale Memristor Device as Synapse in Neuromorphic Systems<\/a><\/em><\/p>","protected":false},"excerpt":{"rendered":"<p>When we study, we highlight important content (reinforce it) and underline unimportant content\u2026<\/p>","protected":false},"author":1,"featured_media":1190,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_kadence_starter_templates_imported_post":false,"_kad_post_transparent":"","_kad_post_title":"","_kad_post_layout":"","_kad_post_sidebar_id":"","_kad_post_content_style":"","_kad_post_vertical_padding":"","_kad_post_feature":"","_kad_post_feature_position":"","_kad_post_header":false,"_kad_post_footer":false,"_kad_post_classname":"","footnotes":""},"categories":[114],"tags":[20,21],"class_list":["post-1188","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-rram-research","tag-memristor","tag-rram"],"_links":{"self":[{"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/posts\/1188","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/comments?post=1188"}],"version-history":[{"count":2,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/posts\/1188\/revisions"}],"predecessor-version":[{"id":1217,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/posts\/1188\/revisions\/1217"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/media\/1190"}],"wp:attachment":[{"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/media?parent=1188"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/categories?post=1188"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/rtlearner.com\/en\/wp-json\/wp\/v2\/tags?post=1188"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}<!-- This website is optimized by Airlift. Learn more: https://airlift.net. Template:. Learn more: https://airlift.net. Template: 6a588749d36f73a90a601ae0. Config Timestamp: 2026-07-16 07:24:54 UTC, Cached Timestamp: 2026-07-18 03:07:47 UTC, Optimization Time: 3.38ms -->