Say hey to the retail GeForce RTX 2080 Ti, packed into its field. Sam Machkovech RTX time. Sam Machkovech System necessities. Sam Machkovech ... discover a extreme discount in warmth exhaust from that panel. It is a main shift from the previous few Nvidia Founders Version GPUs, so if you happen to're utilizing a rigorously packed, thermally challenged PC case, be warned. The 2080 Ti with its fan array eliminated. From the skin, the RTX 2080 seems awfully much like its larger 2080 Ti brother. Internally, the RTX 2080 seems fairly totally different. Specs at a look: Nvidia GeForce RTX 2080 Ti Founders Version CUDA CORES 4352 TEXTURE UNITS 272 ROPS 88 CORE CLOCK 1,350MHz BOOST CLOCK 1,635MHz MEMORY BUS WIDTH 352 bits MEMORY BANDWIDTH 616GB/s MEMORY SIZE 11GB GDDR6 Outputs 3x DisplayPort 1.four, 1x HDMI 2.0b, 1x USB Sort-C (VirtualLink VR) Launch date September 20, 2018 PRICE Founders Version (as reviewed): $1,199. Associate playing cards priced at: $1,169. Like every piece of pricey know-how, a top-of-the-line graphics card comes with all method of lingo and abbreviation. You may want a glossary to wade by the stuff inside (processors, CUDA cores, ROPs), the speeds measured (reminiscence bandwidth, enhance clocks, TeraFLOPS), and the outcomes you need from a superb card (anti-aliasing, body charges, increased resolutions). Due to Nvidia's latest merchandise, the RTX 2080 and RTX 2080 Ti, that required glossary is simply getting larger. The 2 largest corporations within the house, Nvidia and AMD, take totally different approaches to numerous GPU components (and people approaches additional differ within the realms of lower-cost and heat-efficient GPUs). However on the very high of the heap, Nvidia guidelines. And never by a slim margin. This far into 2018, the corporate's final wave of PC graphics playing cards nonetheless leads the market in sheer gaming efficiency, together with March 2017's GTX 1080 Ti . This $700-ish card stays the de facto 4K graphics answer within the shopper sector, and AMD has but to the touch it when it comes to efficiency. The truth is, its late-2017 RX Vega 64 (roughly $550) would not even definitively high the older GTX 1080 (non-Ti, roughly $490). That state of affairs put Nvidia in an fascinating place in August when the corporate introduced a brand new wave of graphics playing cards with much more terminology and lingo connected. We weren't shocked to see one in all Nvidia's favourite matters, ray tracing, emphasised at size. However after the mud and bulletins settled, one in all Nvidia's decisions of phrase caught in our thoughts: "RTX-OPs." You learn that appropriately. Nvidia invented a brand new type of graphics measurement (and one which piggybacks over the extra typical "FLOPS," aka "floating-point operations per second"). What's extra, the corporate's measurement assumes an ideal 3D-acceleration pipeline: conventional rasterization mixed with an intentional tie to a number of Nvidia-only APIs. (Conveniently sufficient, Nvidia's new RTX playing cards characteristic three processor cores that mix forces to just do that.) Put all of these collectively, and voilà, Nvidia swears you will get a ton of RTX-OPs. RTX 2080 Ti RTX 2080 GTX Titan XP GTX 1080 Ti GTX 1080 GTX 1070 GTX 1060 GTX 980 Ti CUDA Cores four,352 2,944 three,584 three,584 2,560 1,920 1,280 2,816 Texture Items 272 184 224 224 160 120 80 176 ROPs 88 64 96 88 64 64 48 96 Core Clock 1,350MHz 1,515MHz 1,417MHz 1,480MHz 1,607MHz 1,506MHz 1,506MHz 1,000MHz Enhance Clock 1,635MHz 1,800MHz 1,531MHz 1,582MHz 1,733MHz 1,683MHz 1,708MHz 1,050MHz Reminiscence Bus Width 352-bit 256-bit 384-bit 352-bit 256-bit 256-bit 192-bit 384-bit Reminiscence Velocity 14GHz 14GHz 10GHz 11GHz 10GHz 8GHz 8GHz 7GHz Reminiscence Bandwidth 616GB/s 448GB/s 480GB/s 484GB/s 320GB/s 256GB/s 192GB/s 336GB/s Reminiscence Dimension 11GB GDDR6 8GB GDDR6 12GB GDDR5X 11GB GDDR5X 8GB GDDR5X 8GB GDDR5 6GB GDDR5 6GB GDDR5 TDP 260W 250W 250W 250W 180W 150W 120W 250W On its face, that is fairly wealthy. So what if Nvidia invented its personal measurement customary and declared itself the winner? That is as if McDonald's measured the flavour and worth of quick meals with a time period like "Grimaces." Methinks Wendy's would lose. Nevertheless it's not a wholly loopy transfer on Nvidia's half. With out AMD apparently nipping at its high-end PC heels, Nvidia enjoys some bandwidth (pun meant) to strive some loopy issues. On this case, these come within the type of proprietary requirements for anti-aliasing, ray tracing, and rasterization. For probably the most half, none of these new Nvidia requirements depends on the standard "extra pace, extra CUDA cores" components of previous. Some spectacular tech is clearly going into Nvidia's efforts, together with a shift to a 7nm course of—and thus making room for brand spanking new discrete processor cores on the RTX collection' chips. Nonetheless, Nvidia is rolling the GPU cube with an RTX-fueled pivot. We're not simply going to be quicker than AMD, the RTX collection trumpets. We'll get our hooks into PC-gaming builders by lower-cost particular results that simply occur to be unique to Nvidia playing cards. The catch, after all, is the longer term tense. In reviewing "Founders Editions" of the RTX 2080 ($799) and RTX 2080 Ti ($1,199), I ran the same old gamut of benchmarks, and I did the same old double-checks of varied video games. However all of the whereas, I turned Nvidia's gross sales pitch again and again in my thoughts—as a result of that is kind of all I might do with the gross sales pitch. Itemizing picture by Sam Machkovech