Introduction
The Nvidia GeForce GTX 1660 Ti and RTX 2060 are two close competitors in the mid-range graphics card market. Released one year apart in early 2019, they occupy a similar space in Nvidia‘s product stack.
With street prices often falling around $250 and $300 respectively, both cards target high frame rate 1080p gaming while offering passable 1440p performance as well. The GTX 1660 Ti fetches strong value through lower cost while the RTX 2060 features more advanced GPU architecture and additional capabilities.
This article will compare the GTX 1660 Ti and RTX 2060 across a range of key specifications, real-world gaming benchmarks, prices, unique features and other relevant considerations to declare an outright winner.
Nvidia GeForce GTX 1660 Ti Overview
Released February 2019, the GeForce GTX 1660 Ti slots below Nvidia‘s high-end RTX cards as a value-oriented offering. Built on the Turing architecture but lacking dedicated RT and tensor cores, the 1660 Ti focuses purely on conventional rasterization graphics.
With a 1,500MHz base clock and 1,770MHz boost clock backed by fast 12Gbps GDDR6 memory, it brings strong 1080p gaming performance at an affordable price point. The reference variant with Nvidia‘s dual fan cooler launched at $279 MSRP. Custom cards from partners like Asus and EVGA quickly followed, offering superior cooling solutions and moderate factory overclocks.
Key GTX 1660 Ti specifications:
- Turing GPU architecture (TU116)
- 1,536 CUDA cores
- 6GB GDDR6 memory
- 192-bit memory bus
- 12Gbps memory speed
- 1,500MHz base / 1,770MHz boost clock
- 120W TDP
With a 120W TDP, compact PCB size and single 8-pin power connector, the GTX 1660 Ti works well in smaller form factor builds or pre-built systems. It performs roughly 10-15% behind Nvidia‘s ex-flagship GTX 1070 and on par with AMD‘s RX 590, landing it firmly in the mid tier for its generation.
Nvidia GeForce RTX 2060 Overview
The RTX 2060 hit the market in January 2019 as the most affordable entry point into Nvidia‘s cutting-edge Turing RTX product stack. It brought dedicated hardware for real-time ray tracing and AI-powered DLSS technology down to the $350 price segment for the first time.
With a boost clock reaching 1,680MHz out of the box and 448GB/s of memory bandwidth from its 14Gbps GDDR6, the 2060 easily outmuscles older 10-series cards like the GTX 1070 Ti and solidly beats AMD‘s best from this era.
Key RTX 2060 specifications:
- Turing architecture (TU106)
- 1,920 CUDA cores
- 6GB GDDR6 memory
- 192-bit bus
- 14Gbps memory speed
- 1,365MHz base / 1,680MHz boost clock
- 160W TDP
The jump to a 160W TDP makes the RTX 2060 less ideal for small form factor systems but delivers that extra compute horsepower over the 1660 Ti. It needs a single 8-pin connector as well. Overall the 2060 brings around a 20% performance uplift in traditional gaming over its lower-cost rival at stock settings.
GTX 1660 Ti vs RTX 2060 Head-to-Head Comparison
| Specification | GTX 1660 Ti | RTX 2060 |
| GPU Architecture | Turing | Turing |
| Lithography | 12nm | 12nm |
| CUDA Cores | 1,536 | 1,920 |
| RT Cores | None | 30 |
| Tensor Cores | None | 240 |
| Texture Units | 96 | 120 |
| ROPs | 48 | 48 |
| Base Clock | 1,500MHz | 1,365MHz |
| Boost Clock | 1,770MHz | 1,680MHz |
| Memory Size | 6GB | 6GB |
| Memory Type | GDDR6 | GDDR6 |
| Memory Bus | 192-bit | 192-bit |
| Memory Bandwidth | 288GB/s | 448GB/s |
| TDP | 120W | 160W |
| Power Connectors | 1 x 8-pin | 1 x 8-pin |
| MSRP | $279 | $349 |
Comparing the raw specifications makes the RTX 2060‘s advantages clear. It has 25% more CUDA cores and texture units thanks to Nvidia enabling more SMs on its larger TU106 die. Combined with higher sustained clocks from GPU Boost and faster 14Gbps memory, the 2060 holds over 50% more memory bandwidth for feeding the extra cores.
Despite identical 6GB capacity and bus width, actual real-world bandwidth equates to performance here. More memory throughput lets the 2060 stretch its legs sooner before getting bound in games.
These hardware improvements give the RTX 2060 its ~20% performance lead in conventional rendering. Then there‘s the specialized RT and tensor cores that enable cutting-edge graphics capabilities exclusive to Nvidia‘s RTX series.
Performance Benchmarks: Gaming and Applications
In real-world game testing, the RTX 2060 consistently outperforms the GTX 1660 Ti by around a 20% margin on average across 1080p and 1440p resolutions:
Hardware Unboxed gaming benchmark data
The gap between both cards does close slightly at 4K resolution as performance becomes more bound by memory capacity. Still, the 2060‘s advantage remains thanks to its much higher memory bandwidth. AMD‘s newly released RX 6600 XT now outpaces both cards quite handily though.
Outside of games, the 2060 also carries a strong lead in compute-heavy workloads that leverage its extra CUDA cores like video editing, 3D modeling and GPU compute tasks:
If your workload involves Nvidia‘s latest graphics APIs like ray tracing or DLSS however, the choice becomes clear. The GTX 1660 Ti lacks any RT core or tensor core hardware to run these advanced features. DLSS can help boost frame rates considerably in supporting games by upscaling intelligently from lower resolutions.
Price to Performance Comparison
Based on current retail pricing, the RTX 2060 offers notably better value per dollar over the 1660 Ti in terms of performance:
Tom‘s Hardware value analysis
Nvidia‘s own site actually lists a $100 premium for the RTX 2060 Founder‘s Edition but most third party models sell for under $350. With the introduction of RTX 3060 cards near $330 as well, both these older GPUs have seen price cuts.
Given its larger die size and more complex manufacturing though, margins remain tighter on the 2060. So if every dollar counts in your budget, the 1660 Ti ekes out a win as the value leader of this matchup. Clock speeds do reach similar ranges when overclocked too.
Unique Capabilities – Ray Tracing, DLSS and NVENC
The GeForce RTX 2060‘s defining features are its support for real-time ray tracing (RTX) using dedicated RT cores and Deep Learning Super Sampling (DLSS) powered by tensor cores.
These specialized processing units allow the Turing architecture to accelerate advanced graphics and upscaling algorithms far beyond what shader-based approaches can achieve. RTX and DLSS debuted on Nvidia‘s highest-end 2080 Ti at first before trickling down the stack.
Though early implementations showed plenty of promise, the performance hit made playing with ray tracing effects enabled challenging even on stronger GPUs. Partnering games have since further optimized and expanded RT effects though while DLSS handles the intensive upscaling process to recover lost FPS.
Ray traced global illumination and shadows in Control with DLSS
Additionally, Nvidia‘s Turing NVENC encoder unlocks vastly improved game streaming quality and performance through dedicated silicon onboard. This allows running high bitrate streams in OBS/Twitch/YouTube without any FPS loss for buttery smooth footage.
The GTX 1660 Ti misses out on all these capabilties. It lacks RT, tensor and newer encoding hardware entirely as a more budget focused option. Unless you specifically need features exclusive to GeForce RTX cards, saving money with the 1660 Ti makes sense.
Other Considerations – Thermals, Noise and Power
The RTX 2060‘s higher 160W TDP requires more powerful cooling and airflow to maintain its boosted clock speeds compared to the 1660 Ti‘s 120W power budget. Open air cards generally run warmer as heat gets exhausted directly into the case instead of an external vent.
More premium 2060 models use thicker heatsinks, triple fans and idle fan stop to solve these issues. But acoustically they are likely to generate more noticeable noise under heavy loads. Smaller cases with restricted intake can exacerbate heat buildup as well.
Matching your GPU‘s thermal design to your specific system build plays an important role too. Cramming a massive three slot, triple fan 2060 into a tiny SFF chassis makes little sense for example.
As for power draw, total board power tops out 40W higher on the 2060 but most mid tier and up PSUs have plenty of overhead there. Just make sure supplemental 8-pin PCIe connectors are available, especially on prebuilt systems.
Expected Longevity
The GeForce GTX 1660 Ti and RTX 2060 are likely to see continued driver support and remain game capable at 1080p and 1440p for at least 2-3 more years. However RTX cards should fare better long term given their advanced graphics, compute and AI acceleration capabilities.
Dating back to Nvidia‘s 1000 series Pascal GPUs, the GTX 1080 Ti for example still largely keeps up with today‘s offerings six years from launch. Extra VRAM capacity plays a role there as well though. The RTX 2060‘s unique tensor and RT cores may prove more future proof.
