Introduction

NVIDIA today launched the GeForce RTX 5060 Ti, and we have with us the Palit RTX 5060 Ti Infinity 3 16 GB. The company today introduced both the 8 GB and 16 GB memory variants of the RTX 5060 Ti, with the 8 GB variant starting at $375, and the 16 GB variant at $425. Both variants have identical specs besides the memory sizes. The new Infinity line of graphics cards from Palit are designed to be priced at NVIDIA baseline pricing, and meet the company's recent SFF Ready specs that call for custom-design graphics cards to be reasonably small, to fit into some of the tighter mid-tower cases and SFF boxes. Infinity 3 denotes a triple-fan variant of this design, Palit also has a dual-fan version of this card that's even smaller. The new GeForce RTX 5060 Ti from NVIDIA is perhaps the company's most important GPU from this generation, as it targets the largest segments of the market—1080p and the aspirational section of it that likes to max out their gameplay, and possibly even upgrade to 1440p displays.



Just like the GeForce RTX 5070 Ti is technically classified as a 1440p-class GPU but is found to be capable of 4K UHD gaming, letting it occupy a gray-area between the performance- and enthusiast segments; the new RTX 5060 Ti is being offered as a 1080p-class GPU, but we put it in a similar gray-area between the mainstream- and performance segments. The RTX 5060 Ti is designed to max out your games at 1080p, including with ray tracing; although if you know your way around your game's settings, can get the NVIDIA App to find the right ones, or want to leverage the awesome capabilities of DLSS 4 with Multi Frame Generation, which the RTX 5060 Ti and the Blackwell graphics architecture have to offer, then 1440p becomes a viable use-case. The 16 GB memory should help with lowering the performance cost of ray tracing, too.

The GeForce RTX 5060 Ti is powered by the Blackwell graphics architecture, which introduces the concept of Neural Rendering—to leverage the power of generative AI to create in-game visual assets in real time, and blend it with the classic raster 3D just like ray traced objects are. NVIDIA even worked with Microsoft to lay the API-level groundwork for Neural Rendering. What makes this exclusive to the RTX 50-series is the new hardware scheduler NVIDIA introduced, called the AI Management Processor (AMP), which lets the GPU accelerate generative AI models and render graphics in tandem. Microsoft's API-level standardization with DirectX 12 lets games directly access the Tensor cores on the GPU.

With this generation, NVIDIA is also introducing DLSS 4 and Multi Frame Generation. DLSS 4 sees NVIDIA replace the convoluted neural networks (CNN) based AI models handling super resolution, ray reconstruction, and frame generation; with newer Transformer-based models, which are more accurate, and offer image quality improvements at all performance tiers. Multi Frame Generation is a step up from DLSS 3 Frame Generation introduced with the RTX 40-series; which sees the GPU generate up to 3 frames entirely using AI, for every conventionally-rendered frame. This effectively quadruples frame-rates, but when used in conjunction with super resolution, it could mean 1 conventionally rendered pixel resulting in up to 15 pixels drawn entirely by AI. Multi Frame Generation relies on hardware flip-metering, which is exclusive to Blackwell GPUs.

The new Blackwell generation CUDA core not just introduce generational IPC increases, but are also redesigned to accelerate neural shaders. The 4th generation RT core, besides continuing to lower the performance cost of ray tracing compared to the previous generation, comes with optimization for mega geometry—ray traced objects with much more complex geometry (more surfaces for rays to interact with). The 5th generation Tensor cores lay the groundwork for not just for neural rendering, but also introduce support for newer data formats, including FP4. Updates are also introduced to the display engine, which supports DisplayPort 2.1 with UHBR20, and HDMI 2.1. The latest generation video encode and decode accelerators support 4:2:2 color formats, and AV1 B-frames, which vastly reduce video stream bitrates and file sizes.

The GeForce RTX 5060 Ti introduces the new GB206 silicon, NVIDIA's fourth new silicon from this generation. This chip, like the rest of the Blackwell family, is built on the same exact NVIDIA 4N foundry node as the RTX 40-series Ada generation, and so all efficiency improvements you see are purely a function of the architecture. The RTX 5060 Ti maxes out the GB206 silicon on which it is based. This chip comes with 36 streaming multiprocessors (SM), which work out to 4,608 CUDA cores, 144 Tensor cores, 36 RT cores, 144 TMUs, and 48 ROPs. The chip also features 32 MB of L2 cache, which the RTX 5060 Ti maxes out. The memory interface is 128-bit wide GDDR7. This upgrade has the greatest impact on memory bandwidth compared to the previous generation RTX 4060 Ti 16 GB. With a memory speed of 28 Gbps (GDDR7-effective), the RTX 5060 Ti enjoys 448 GB/s of memory bandwidth, a massive 55% increase over the 18 Gbps GDDR6 that the RTX 4060 Ti used. The RTX 5060 Ti comes with an updated PCI-Express 5.0 x8 host interface. On all platforms since 2021—Intel 12th Gen Alder Lake and Ryzen 7000 Zen 4, this means comparable host interface bandwidth to PCI-Express 4.0 x16.

The Palit RTX 5060 Ti Infinity 3 16 GB is very close to what would have been an NVIDIA reference design. It is priced at NVIDIA's baseline price, has reference clock speeds, meets SFF Ready product dimensions, and is one of only two cards we're reviewing today that comes with a classic 8-pin PCIe power connector. Given the 180 W TGP of the RTX 5060 Ti, a single 8-pin PCIe (150 W) with PCIe slot power (75 W), meet the power budget of the GPU. If you recall, the RTX 4070 had similar power specs. Most other custom design RTX 5060 Ti cards we're reviewing today implement the newer 16-pin 12V-2x6 power connector, with an included adapter. Palit is pricing this card at $430.

NVIDIA Blackwell Architecture

NVIDIA does not provide a block diagram for the GB206 GPU (we asked), so we had to quickly hack one out from the GB202 diagram. This is accurate just not as pretty.

The GeForce Blackwell graphics architecture heralds NVIDIA's 4th generation of RTX, the late-2010s re-invention of the modern GPU that sees a fusion of real time ray traced objects with conventional raster 3D graphics. With Blackwell, NVIDIA is helping add another dimension, neural rendering, the ability for the GPU to leverage a generative AI to create portions of a frame. This is different from DLSS, where an AI model is used to reconstruct details in an upscaled frame based on its training date, temporal frames, and motion vectors. Today we are reviewing NVIDIA's fifth GPU from this generation, the RTX 5060 Ti. At the heart of this graphics card is the new GB206 silicon. This chip has a unique die-size and SM count that doesn't have a predecessor from the previous Ada generation. With the GeForce Blackwell generation, NVIDIA tailored the SM counts of its GPUs closer to those of the actual SKUs, to minimize wasted die-space. The GB206 is maxed out by the RTX 5060 Ti. This is NVIDIA's fourth gaming GPU based on this architecture. It is built on the same NVIDIA 4N foundry node as the previous generation AD106. It is a tiny, somewhat square die measuring 181 mm², with a transistor count of 21.9 billion.

The GB206 silicon is laid out essentially in the same component hierarchy as past generations of NVIDIA GPUs, but with a few notable changes. The GPU features a PCI-Express 5.0 x8 host interface. PCIe Gen 5 has been around since Intel's 12th Gen Core "Alder Lake" and AMD's Ryzen 7000 "Zen 4," so there is a sizable install-base of systems that can take advantage of it. The GPU is of course compatible with older generations of PCIe. The GB206 also features the new GDDR7 memory interface that's making its debut with this generation. The chip features a 128-bit wide memory bus. NVIDIA is using this to drive 16 GB of memory at 28 Gbps speeds, yielding 448 GB/s of memory bandwidth, which is a 55% increase over the RTX 4060 Ti and its 18 Gbps GDDR6.

The GigaThread Engine is the main graphics rendering workload allocation logic on the GB206, but there's a new addition, a dedicated serial processor for managing all AI acceleration resources on the GPU, NVIDIA calls this AMP (AI management processor). Other components at the global level are the Optical Flow Processor, a component involved in older versions of DLSS frame generation and for video encoding; and an updated media acceleration engine consisting of one each of NVDEC and NVENC video accelerators. The new 9th Gen NVENC video encode accelerators come with 4:2:2 AV1 and HEVC encoding support. The central region of the GPU has the single largest common component, the 32 MB L2 cache, which the RTX 5060 Ti maxes out.


There are three graphics processing clusters (GPC) on the GB206. Each of these contains 12 streaming multiprocessors (SM) across 6 texture processing clusters (TPCs), resulting in 36 SM. Every GPC also contains a ROP partition with 12 ROPs, each, working out to 48 ROPs on the silicon. Within each SM, you'll find 128 CUDA cores. Unlike the Ada generation SM that each had 64 FP32+INT32 and 64 purely-FP32 SIMD units, the new Blackwell generation SM features concurrent FP32+INT32 capability on all 128 SIMD units. These 128 CUDA cores are arranged in four slices, each with a register file, a level-0 instruction cache, a warp scheduler, two sets of load-store units, and a special function unit (SFU) handling some special math functions such as trigonometry, exponents, logarithms, reciprocals, and square-root. The four slices share a 128 KB L1 data cache, and four TMUs. The most exotic components of the Blackwell SM are the four 5th Gen Tensor cores, and a 4th Gen RT core.

With 3 GPCs containing 6 TPCs each, there are a total of 36 SM, worth 4,608 CUDA cores, 144 Tensor cores, 36 RT cores, and 144 TMUs, on the GB206 silicon. The GB206 silicon is endowed with 48 ROPs, all of which are enabled on the RTX 5060 Ti. The RTX 5060 Ti also maxes out all 32 MB of L2 cache present on the die.


Perhaps the biggest change to the way the SM handles work, introduced with Blackwell, is the concept of neural shaders—treating portions of the graphics rendering workload done by a generative AI model as shaders. Microsoft has laid the groundwork for standardization of neural shaders with its Cooperative Vectors API, in the latest update to DirectX 12. The Tensor cores are now accessible for workloads through neural shaders, and the shader execution reordering (SER) engine of the Blackwell SM is able to more accurately reorder workloads for the CUDA cores and the Tensor core in an SM.


The new 5th Gen Tensor core introduces support for FP4 data format (1/8 precision) to fast moving atomic workloads, providing 32 times the throughput of the very first Tensor core introduced with the Volta architecture. Over the generations, AI models leveraged lesser precision data formats, and sparsity, to improve performance. The AI management processor (AMP) is what enables simultaneous AI and graphics workloads at the highest levels of the GPU, so it could be simultaneously rendering real time graphics for a game, while running an LLM, without either affecting the performance of the other. AMP is a specialized hardware scheduler for all the AI acceleration resources on the silicon. This plays a crucial role for DLSS 4 multi-frame generation to work.


The 4th Gen RT core not just offers a generational increase in ray testing and ray intersection performance, which lowers the performance cost of enabling path tracing and ray traced effects; but also offers a potential generational leap in performance with the introduction of Mega Geometry. This allows for ray traced objects with extremely high polygon counts, increasing their detail. Poly count and ray tracing present linear increases in performance costs, as each triangle has to intersect with a ray, and there should be sufficient rays to intersect with each of them. This is achieved by adopting clusters of triangles in an object as first-class primitives, and cluster-level acceleration structures. The new RT cores introduce a component called a triangle cluster intersection engine, designed specifically for handling mega geometry. The integration of a triangle cluster compression format and a lossless decompression engine allows for more efficient processing of complex geometry.


The GB206 and the rest of the GeForce Blackwell GPU family is built on the exact same TSMC "NVIDIA 4N" foundry node, which is actually 5 nm, as previous-generation Ada, so NVIDIA directed efforts to finding innovative new ways to manage power and thermals. This is done through a re-architected power management engine that relies on clock gating, power gating, and rail gating of the individual GPCs and other top-level components. It also worked on the speed at which the GPU makes power-related decisions.


The quickest way to drop power is by adjusting the GPU clock speed, and with Blackwell, NVIDIA introduced a means for rapid clock adjustments at the SM-level.


NVIDIA updated both the display engine and the media engine of Blackwell over the previous generation Ada, which drew some flack for holding on to older display I/O standards such as DisplayPort 1.4, while AMD and Intel had moved on to DisplayPort 2.1. The good news is that Blackwell supports DP 2.1 with UHBR20, enabling 8K 60 Hz with a single cable. The company also updated NVDEC and NVENC, which now support AV1 UHQ, double the H.264 decode performance, MV-HEVC, and 4:2:2 formats.

Neural Rendering

Neural Rendering promises to be as transformative to modern graphics as programmable shaders itself. 3D Graphics rendering evolved from fixed-function over the turn of the century, to programmable shaders, HLSL, geometry shaders, compute shaders, and ray tracing, over the past couple of decades. In 2025, NVIDIA is writing the next chapter in this journey with Blackwell neural shaders. This allows for a host of neural-driven effects, including neural materials, neural volumes, and even neural radiance fields. Microsoft introduced the new Cooperative Vectors API for DirectX in a recent update, making it possible to access Tensor cores within a graphics API. Combined with a new shading language, Slang, this breakthrough enables developers to integrate neural techniques directly into their workflows, potentially replacing parts of the traditional graphics pipeline. Slang splits large, complex functions into smaller pieces that are easier to handle. Given that this is a DirectX standard API feature, there is nothing that stops AMD and Intel from integrating Neural Rendering (Cooperative Vectors) into their graphics drivers.

RTX Neural Materials works to significantly reduce the memory footprint of materials in 3D scenes. Under conventional rendering, the memory footprint of a material is bloated from complex shader code. Neural materials convert shader code and texture layers into a compressed neural representation. This results in up to a 7:1 compression ratio and enables small neural networks to generate stunning, film-like materials in real-time. For example, silk rendered with traditional shaders might lack the multicolored sheen seen in real life. Neural materials, however, capture intricate details like color variation and reflections, bringing such surfaces to life with unparalleled realism—and at a fraction of the memory cost.


The new Neural Radiance Cache, which dynamically trains a neural network during gameplay using the user's GPU, allowing light transport to be cached spatially, enabling near-infinite light bounces in a scene. This results in realistic indirect lighting and shadows with minimal performance impact. NRC partially traces 1 or 2 rays before storing them in a radiance cache, and infers an infinite amount of rays and bounces for a more accurate representation of indirect lighting in the game scene.

DLSS 4 and Multi Frame Generation

DLSS 4 introduces a major leap in image quality and performance. It isn't just a version bump with the introduction of a new feature, namely Multi Frame Generation, but introduces updates to nearly all DLSS sub-features. DLSS from its very beginning relied on AI to reconstruct details in super resolution, and with DLSS 4, NVIDIA is introducing a new transformer-based AI model to succeed the convolutional neural networks previous used, for double the parameters, four times the compute performance, and significantly improved image quality. Ray Reconstruction, introduced with DLSS 3.5, gets a significant image quality update with the new transformer-based model.


To understand Multi Frame Generation, you need to understand how DLSS Frame Generation, introduced with GeForce Ada, works. An Optical Flow Accelerator component gives the DLSS algorithm data to generate an entire frame using a neural network, using information from a previous rendered frame, effectively doubling frame rate. In Multi Frame Generation, AI takes over the functions of optical flow, to predict up to three frames following a conventionally rendered frame, effectively drawing four frames form the rendering effort of one.


Now, assuming this rendered frame is a product of Super Resolution, with the maximum performance setting generating 4x the pixels from a single rendered pixel, you're looking at a possibility where the rendering effort of 1/4th a frame goes into drawing 4 frames, or 15 in every 16 pixels being generated entirely by DLSS. When generating so many frames, Frame Pacing becomes a problem—irregular frame intervals impact smoothness. DLSS 4 addresses these issues by using a dedicated hardware unit inside Blackwell, which takes care of flip metering, reducing frame display variability by 5-10x. The Display Engine of Blackwell contains the hardware for flip metering.

NVIDIA Reflex 2

The original NVIDIA Reflex brought about a significant improvement to the responsiveness of maxed out graphics in competitive online gameplay, by compacting the rendering queue with the goal of reducing the whole system latency by up to 50%. Reflex is mandatory in DLSS 3 Frame Generation, given the latency cost imposed by the technology. Multi-frame generation calls for an equally savvy piece of technology, so we hence have Reflex 2. NVIDIA claims to have achieved a 75% reduction in latency with Frame Warp, which updates the camera (viewport) positions based on user inputs in real-time, and then uses temporal information to reconstruct the frame to display.

Packaging

The Card

The Palit Infinity 3 styling uses a mostly black color theme with small white highlights. Both the front cooler and backplate are made from plastic. The backplate has cutouts for air to flow through.


Dimensions of the card are 29.0 x 12.5 cm, and it weighs 709 g.


Installation requires two slots in your system. We measured the card's width to be 41 mm.


Display connectivity includes three standard DisplayPort 2.1b and one HDMI 2.1b.

All GeForce RTX 50-series Blackwell cards feature a new display engine supporting DisplayPort 2.1b with UHBR20, as well as HDMI 2.1a. Both interfaces support DSC (display stream compression). With DSC enabled, a single DisplayPort on this card can drive 4K 12-bit HDR at 480 Hz; or 8K 12-bit HDR at up to 165 Hz. The RTX 5060 Ti features an updated media acceleration engine with support for 4:2:2 video formats, AV1 UHQ, and MV-HEVC. Unlike the bigger RTX 50 models, which have two, there is a single NVENC and NVDEC unit each.


The card uses a single 8-pin connector, which, theoretically allows a maximum power draw of 150 W, plus 75 W from the slot. The card's power limit is set to 180 W, with no manual increases allowed.


There is no lighting on the Palit RTX 5060 Ti Infinity 3.