Tom's Hardware Verdict
The GeForce RTX 3060 Ti beats last year's RTX 2080 Super and costs $300 less. It's a great option for the $400 GPU market, though once again we expect it will sell out quickly for the next couple of months.
Pros
- +
Great price/performance ratio
- +
Attractive pricing (if it stays in stock)
- +
Efficient, cool, and decent overclocking
- +
DLSS 2.0 makes it even faster
Cons
- -
8GB VRAM can be limiting
- -
Good luck finding one in stock
- -
No LEDs (if you like bling)
Why you can trust Tom's Hardware Our expert reviewers spend hours testing and comparing products and services so you can choose the best for you. Find out more about how we test.
The Nvidia GeForce RTX 3060 Ti brings the entry price for Ampere GPUs down to $400, because apparently the GeForce RTX 3070, GeForce RTX 3080, and GeForce RTX 3090 weren't selling out fast enough. Maybe that's not fair, but round four of the Nvidia Ampere architecture seems destined to repeat recent history and sell out within minutes of launch. The fact that this is now the best price to performance ratio in GPU land makes it one of the best graphics cards, and it currently occupies seventh place on our GPU Benchmarks hierarchy, not counting any Titan cards. Maybe if you've been really good (or you're really lucky), you'll be able to procure one this side of 2021. [Update: Yup, the RTX 3060 Ti sold out fast. Here's our advice on how to buy an RTX 3060 Ti (or other new Nvidia GPU).]
Not a whole lot has changed relative to the RTX 3070. Both the RTX 3060 Ti and 3070 use the same GA104 GPU, but the 3060 Ti has fewer functional units enabled. You get the same 8GB of GDDR6 memory, clocked at the same 14Gbps. Meanwhile, the GPU core has 38 SMs enabled, versus 46 SMs on the 3070, making the 3060 Ti theoretically around 17 percent slower. Clock speeds also factor in a bit, as the boost clock on the 3060 Ti is 3.5 percent lower, so in general, the largest performance gap will be about 20 percent — and as we'll see, it's often quite a bit less than that.
| Graphics Card | RTX 3060 Ti | RTX 3080 | RTX 3070 | RTX 2080 Super | RTX 2060 Super |
|---|---|---|---|---|---|
| Architecture | GA104 | GA102 | GA104 | TU104 | TU106 |
| Process Technology | Samsung 8N | Samsung 8N | Samsung 8N | TSMC 12FFN | TSMC 12FFN |
| Transistors (Billion) | 17.4 | 28.3 | 17.4 | 13.6 | 10.8 |
| Die size (mm^2) | 392.5 | 628.4 | 392.5 | 545 | 445 |
| TPCs | 19 | 34 | 23 | 24 | 17 |
| SMs | 38 | 68 | 46 | 48 | 34 |
| GPCs | 5 | 6 | 6 | 6 | 3 |
| GPU Cores | 4864 | 8704 | 5888 | 3072 | 2176 |
| Tensor Cores | 152 | 272 | 184 | 384 | 272 |
| RT Cores | 38 | 68 | 46 | 48 | 34 |
| Base Clock (MHz) | 1410 | 1440 | 1500 | 1650 | 1470 |
| Boost Clock (MHz) | 1665 | 1710 | 1725 | 1815 | 1650 |
| VRAM Speed (Gbps) | 14 | 19 | 14 | 15.5 | 14 |
| VRAM (GB) | 8 | 10 | 8 | 8 | 8 |
| VRAM Bus Width | 256 | 320 | 256 | 256 | 256 |
| ROPs | 80 | 96 | 96 | 64 | 64 |
| TMUs | 152 | 272 | 184 | 192 | 136 |
| TFLOPS FP32 (Boost) | 16.2 | 29.8 | 20.3 | 11.2 | 7.2 |
| Tensor TFLOPS FP16 (Sparsity) | 65 (130) | 119 (238) | 81 (163) | 89 | 57 |
| RT TFLOPS (approximate) | 39.7 | 58.1 | 39.7 | 31.6 | 23.8 |
| Bandwidth (GBps) | 448 | 760 | 448 | 496 | 448 |
| TDP (watts) | 200 | 320 | 220 | 185 | 185 |
| Dimensions | 242x112x38 mm | 285x112x38 mm | 242x112x38 mm | 267x116x38 mm | 229x113x38 mm |
| Weight | 1031g | 1355g | 1034g | 1278g | 975g |
| Launch Date | Oct-20 | Sep-20 | Oct-20 | Jul-19 | Jul-19 |
| Launch Price | $399 | $699 | $499 | $699 | $399 |
The specs tell the same story relative to the previous generation Nvidia Turing GPUs. The 3060 Ti has quite a bit more theoretical compute performance than the RTX 2080 Super, and more than doubles the compute performance relative to the RTX 2060 Super. We know GPU prices are all kinds of screwed up right now, but it's worth remembering that the 2060 Super had the same official $399 price as the 3060 Ti. Power requirements are slightly higher than the 2060 Super, but lower than the 2080 Super.
Of course, theoretical specs don't always jive with real-world performance. One of the big changes with Ampere is that the INT32 cores can now do FP32 calculations as well. However, there's a decent amount of INT32 work in most games, so a good chunk of the potential FP32 performance is used up elsewhere. Still, with the other architectural changes, we expect the 3060 Ti to perform roughly on par with the 2080 Super, which means it will also be significantly faster than the 2060 Super.
Meet the GeForce RTX 3060 Ti Founders Edition
There are no surprises in regards to the design and aesthetics of the GeForce RTX 3060 Ti Founders Edition: It's identical to the RTX 3070 FE, except for the small text that now says RTX 3060 Ti. It's a reasonably compact card that will fit in most cases, measuring 9.5 inches long. If you're doing a mini-ITX build, plan accordingly, but most ATX and micro-ATX cases should be fine.
With a reduced TDP of 200W, and considering what we saw from the 3070, cooling and noise shouldn't pose any problem. We'll get to the hard numbers later, but let's just say that the 1665 MHz boost clock is very conservative at stock, and there's a decent bit of overclocking headroom as well.
Connectivity consists of the same triple DisplayPort 1.4a outputs and single HDMI 2.1 port that we've seen on other Ampere Founders Edition models. You can use up to four monitors, and there's some ventilation on the rest of the IO bracket.
Somewhat interesting is that even the RTX 3060 Ti Founder Edition continues to use Nvidia's 12-pin power connector. Again, we're not quite sure what the point is, particularly since it just takes a single 8-pin PEG connector on one end and passes that through to the 12-pin connection. Even better is that Nvidia had concept renderings of the 3070 with an 8-pin connector. It doesn't really matter, as the adapter cable is included in the box, but it's also completely unnecessary. The 3080/3090 at least changed the PCB and have the 12-pin connector mounted in such a way that it uses a bit less area. Not surprisingly, none of the AIB partners seem interested in pushing the 12-pin agenda, so if you really don't like it, there will be plenty of other options.
MORE: Best Graphics Cards
MORE: GPU Benchmarks
MORE: All Graphics Content
Arm vs. Qualcomm trial begins — Arm demands that the patent-infringing Nuvia designs be destroyed
China developing critical chipmaking supply chains — photoresist ecosystem emerges for ArF and KrF lasers
Inno3D hints at Neural Rendering and Advanced DLSS technology for Nvidia's next-gen RTX 50-series GPUs — DLSS 4 may debut with Blackwell
