{"id":7984,"date":"2022-03-24T21:13:24","date_gmt":"2022-03-25T00:13:24","guid":{"rendered":"https:\/\/mastertrend.info\/?p=7984"},"modified":"2026-04-28T16:38:25","modified_gmt":"2026-04-28T19:38:25","slug":"geforce-rtx-3090-ti-la-top","status":"publish","type":"post","link":"https:\/\/mastertrend.info\/en\/geforce-rtx-3090-ti-la-top\/","title":{"rendered":"GeForce RTX 3090 Ti Technical Data"},"content":{"rendered":"<p><strong>GeForce RTX 3090 Ti Specifications<\/strong><\/p>\n<h2>GeForce RTX 3090 Technical Information.<\/h2>\n<p>The GeForce RTX 3090 Ti, the dedicated graphics card <strong>considerably stronger (and more expensive) produced by NVIDIA for the consumer market<\/strong>, will be free on March 29 according to the latest clarifications.<\/p>\n<p>At the same time, images of the many customized models that will use this graphics chip have appeared on the Internet.<\/p>\n<div>\n<p>Following its official announcement at CES and an expected launch in January, NVIDIA has been fairly quiet about its next flagship in dedicated markets.<\/p>\n<p>We do not understand if the delay is due to the lack of semiconductors, cryptomining or as some comments mention, to &quot;problems&quot; with the <a href=\"https:\/\/mastertrend.info\/en\/camm2-new-standard-of-ram\/\" data-wpil-monitor-id=\"1089\" target=\"_blank\">memory<\/a> GDDR6X that this card uses.<\/p>\n<h3>GeForce RTX 3090 Ti, free way<\/h3>\n<p>It seems that the moment has arrived and the <strong>March 29 is the new date<\/strong> that media like <em>Videocardz<\/em> They manage the availability of the card and also the expert analyses that will be published.<\/p>\n<p>If you follow us you will know her.<\/p>\n<p>It is the new hall of <a href=\"https:\/\/mastertrend.info\/en\/27-oled-gaming-monitor-2\/\" data-wpil-monitor-id=\"496\" target=\"_blank\">range<\/a> from NVIDIA and the entire client graphics market.<\/p>\n<p>And it&#039;s a considerable step up from today&#039;s RTX 3090 with clarifications that specify what we can expect, the ability to move <a href=\"https:\/\/mastertrend.info\/en\/tips-for-measuring-performance-in-games-2\/\" target=\"_blank\" rel=\"noopener\" data-wpil-monitor-id=\"904\">games<\/a> with resolutions up to 8K.<\/p>\n<ul>\n<li>GA102 GPU manufactured using 8nm development.<\/li>\n<li>ten,752 shaders.<\/li>\n<li>336 texturing entities.<\/li>\n<li>112 raster entities.<\/li>\n<li>336 <a href=\"https:\/\/mastertrend.info\/en\/neural-texture-compression\/\" target=\"_blank\" rel=\"noopener\"  data-wpil-monitor-id=\"3823\">tensor cores<\/a>.<\/li>\n<li>84 RT cores.<\/li>\n<li>384-bit collective.<\/li>\n<li>24 GB of GDDR6X memory at 21 Gbps.<\/li>\n<li>1.018 TB\/s bandwidth.<\/li>\n<li>FP32 capability: Around 40 TFLOPS.<\/li>\n<li>TGP: 450 watts.<\/li>\n<li>You would need an 850 watt power supply, and an additional 12 pin power connector (we could connect the adapter to 2 8 pin cables).<\/li>\n<\/ul>\n<h4>Boomstar GeForce RTX 3090 Ti<\/h4>\n<p>It belongs to the customized models that will be marketed.<\/p>\n<p>This is a unique series for China, but it does let us understand elements like the new 12+4 (or 16-pin) PCIe Gen5 connector that allows us to offer up to <strong>600 watts of capacity<\/strong>.<\/p>\n<p><img decoding=\"async\" class=\"aligncenter wp-image-305507\" title=\"GeForce RTX 3090 Ti, the &quot;beast&quot; arrives on March 29 33\" src=\"https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/GeForce-RTX-3090-Ti-la-bestia-llega-el-29-de.jpg\" alt=\"GeForce RTX 3090 Ti\" width=\"1000\" height=\"426\" \/><\/p>\n<p>The images of this card claim exactly that, meaning we&#039;re looking at the next-gen GPU capacity distribution that will likely be adopted by all recent cards in the coming years.<\/p>\n<h2>RTX 3090 vs 3090 Ti specifications<\/h2>\n<div class=\"table-is-responsive\">\n<table width=\"439\">\n<tbody>\n<tr>\n<td width=\"170\"><\/td>\n<td width=\"131\"><strong>NVIDIA RTX 3090<\/strong><\/td>\n<td width=\"130\"><strong>NVIDIA RTX 3090 Ti<\/strong><\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>GPU<\/strong><\/td>\n<td width=\"131\">GA102-300-A1<\/td>\n<td width=\"130\">GA102-350-A1<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Architecture<\/strong><\/td>\n<td colspan=\"2\" width=\"263\">Ampere, 8nm<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Die size<\/strong><\/td>\n<td colspan=\"2\" width=\"263\">628 mm2<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>VRAM memory<\/strong><\/td>\n<td colspan=\"2\" width=\"263\">24GB GDDR6X<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Memory bus<\/strong><\/td>\n<td colspan=\"2\" width=\"263\">384-bit<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Bandwidth<\/strong><\/td>\n<td width=\"131\">936.2 GB\/s<\/td>\n<td width=\"130\">1008 GB\/s<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Memory speed<\/strong><\/td>\n<td width=\"131\">19.5 Gbps<\/td>\n<td width=\"130\">21 Gbps<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>CUDA Cores\/shaders<\/strong><\/td>\n<td width=\"131\">10496<\/td>\n<td width=\"130\">10752<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Tensor Cores<\/strong><\/td>\n<td width=\"131\">328<\/td>\n<td width=\"130\">336<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>ROPs<\/strong><\/td>\n<td width=\"131\">112<\/td>\n<td width=\"130\">112<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>RT Cores<\/strong><\/td>\n<td width=\"131\">82<\/td>\n<td width=\"130\">84<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Base frequency<\/strong><\/td>\n<td width=\"131\">1395 MHz<\/td>\n<td width=\"130\">1560 MHz<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Turbo frequency<\/strong><\/td>\n<td width=\"131\">1695 MHz<\/td>\n<td width=\"130\">1860 MHz<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>TDP<\/strong><\/td>\n<td width=\"131\">350 W<\/td>\n<td width=\"130\">450 W<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Recommended source<\/strong><\/td>\n<td width=\"131\">750 W<\/td>\n<td width=\"130\">850 W<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Ports<\/strong><\/td>\n<td colspan=\"2\" width=\"263\">1 x HDMI 2.1<\/p>\n<p>3 x DisplayPort 1.4<\/p>\n<p>PCIe 1\u00d712\/1\u00d716 pin<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Dimensions (reference designs)<\/strong><\/td>\n<td colspan=\"2\" width=\"263\">336 x 140 x 61 mm<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Departure date<\/strong><\/td>\n<td width=\"131\">1\/9\/20<\/td>\n<td width=\"130\">29\/3\/22<\/td>\n<\/tr>\n<tr>\n<td width=\"170\"><strong>Starting price<\/strong><\/td>\n<td width=\"131\"><strong>$1499<\/strong><\/td>\n<td width=\"130\"><strong>$1999<\/strong><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><img decoding=\"async\" class=\"alignnone wp-image-9584 size-full\" title=\"rtx 3090 ti vs 3090 4\" src=\"https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4.jpg\" alt=\"RTX 3090 Ti\" width=\"1000\" height=\"558\" srcset=\"https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4.jpg 1000w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-300x167.jpg 300w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-768x429.jpg 768w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-18x10.jpg 18w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-128x72.jpg 128w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-120x67.jpg 120w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-180x100.jpg 180w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-360x201.jpg 360w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-470x262.jpg 470w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-750x419.jpg 750w, https:\/\/mastertrend.info\/wp-content\/uploads\/2022\/03\/rtx-3090-ti-vs-3090-4-600x335.jpg 600w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/><\/p>\n<p>We have a timid upgrade in the RTX 3090 Ti, which improves little to the RTX 3090 and\u00a0<strong>increases consumption by 100 W<\/strong>\u2026somewhat questionable.<\/p>\n<p>The chassis changes: PCB and <a href=\"https:\/\/mastertrend.info\/en\/rtx-5080-fe-instability\/\" target=\"_blank\" rel=\"noopener\"  data-wpil-monitor-id=\"2997\">GPU<\/a> different on the RTX 3090 Ti, they were quick to speculate that it is a mule for the RTX 4090!<\/p>\n<div id=\"lx_679961\" class=\"__lxG__multi __lxG__multi_lx_679961 __lxG__loading_block\" data-cnt=\"2\">\n<div id=\"div-gpt-ad-22684225552-2\">However, we have the same die size, same VRAM, same memory bus, but the main changes are:<\/div>\n<\/div>\n<ul>\n<li>More bandwidths.<\/li>\n<li>Faster memory speed (with 21 Gbps).<\/li>\n<li><strong>Increased CUDA Cores<\/strong>\u00a0(256 more),\u00a0<strong>Tensor Cores<\/strong>\u00a0(8 more) and\u00a0<strong>RT Cores<\/strong>\u00a0(2 more).<\/li>\n<li>Increased frequencies:\u00a0<strong>165 MHz more<\/strong>\u00a0in both.<\/li>\n<li><strong>100 W more<\/strong>\u00a0consumption and, therefore, greater requirement of <a href=\"https:\/\/mastertrend.info\/en\/armando-mi-pc\/\" target=\"_blank\" rel=\"noopener\"  data-wpil-monitor-id=\"2765\">power supply<\/a>.<\/li>\n<li><strong>12-pin PCIe connector<\/strong>.<\/li>\n<li><strong>$500 difference<\/strong>.<\/li>\n<\/ul>\n<p>Therefore, we have a\u00a0<strong>33% price increase<\/strong>\u00a0compared to the RTX 3090, so we&#039;ll see if it&#039;s worth paying the extra $500 for the Ti model.<\/p>\n<p>Compared to the RTX 3090 Boomstar, the Ti model will only use a single 16-pin connector instead of two 8-pin ones.<\/p>\n<p>As for expenses, what can I say?<\/p>\n<p>Knowing the situation today we do not believe that they will be able to buy under the barrier of <strong>2,000 US dollars<\/strong>.<\/p>\n<p>Read Also: <a href=\"https:\/\/mastertrend.info\/en\/nvidia-and-amd-graphics-card-comparison\/\" target=\"_blank\">NVIDIA and AMD graphics cards comparison<\/a><\/p>\n<\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>RTX3090TI Technical Information. Learn more now!<\/p>","protected":false},"author":1,"featured_media":7985,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","iawp_total_views":21,"jnews-multi-image_gallery":[],"jnews_single_post":[],"jnews_primary_category":[],"jnews_social_meta":[],"jnews_review":[],"enable_review":"","type":"","name":"","summary":"","brand":"","sku":"","good":[],"bad":[],"score_override":"","override_value":"","rating":[],"price":[],"jnews_override_counter":[],"footnotes":""},"categories":[74],"tags":[1709,1603],"class_list":["post-7984","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-hardware","tag-gpu","tag-nvidia"],"_links":{"self":[{"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/posts\/7984","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/comments?post=7984"}],"version-history":[{"count":2,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/posts\/7984\/revisions"}],"predecessor-version":[{"id":104002,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/posts\/7984\/revisions\/104002"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/media\/7985"}],"wp:attachment":[{"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/media?parent=7984"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/categories?post=7984"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/mastertrend.info\/en\/wp-json\/wp\/v2\/tags?post=7984"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}