Gtx 1070 ti vs gtx 1080: Nvidia GTX 1070-Ti vs 1080

Nvidia GTX 1070-Ti vs 1080

  • CPU
  • GPU
  • SSD
  • HDD
  • RAM
  • USB

VS

YouTube*NEW*

About

Real World Speed

Performance profile from 1,448,635 user samples

Benchmark your GPU here

1,087,024 User Benchmarks

Best Bench: 117% MSI(1462 3364) ≥ 4GB
Worst Bench: 94% Gigabyte(1458 3730)

Poor: 94%

Great: 117%

SPEED RANK: 38th / 688

361,611 User Benchmarks

Best Bench: 106% MSI(1462 C300)
Worst Bench: 88% Gigabyte(1458 3792)

Poor: 88%

Great: 106%

SPEED RANK: 47th / 688

Effective 3D Speed

Effective 3D Gaming GPU Speed
106 % Slightly faster effective speed.
+10%
96.3 %
Lighting

Avg. Locally-deformable PRT (Bat)
130 fps Slightly better lighting effects.
+9%
119 fps
Reflection

Avg. High dynamic range lighting (Teapot)
146 fps Slightly better reflection handling.
+7%
136 fps
MRender

Avg. Render target array GShader (Sphere)
116 fps Faster multi rendering.
+23%
94.1 fps
Gravity

Avg. NBody particle system (Galaxy)
135 fps Slightly faster NBody calculation.
+9%
124 fps
Lighting

Locally-deformable PRT (Bat)
145 fps Slightly better peak lighting effects.
+9%
133 fps
Reflection

High dynamic range lighting (Teapot)
197 fps +3% 191 fps
MRender

Render target array GShader (Sphere)
124 fps Faster peak multi rendering.
+23%
101 fps
Gravity

NBody particle system (Galaxy)
148 fps Slightly faster peak NBody calculation.
+10%
135 fps
CSGO

Avg. Fps @ 1080p on Max
198 Fps 200 Fps +1%
Fortnite

Avg. Fps @ 1080p on Max
140 Fps Slightly Better Fortnite Fps.
+9%
129 Fps
PUBG

Avg. Fps @ 1080p on Max
96 Fps +2% 94 Fps
GTAV

Avg. Fps @ 1080p on Max
94 Fps Better GTAV Fps.
+11%
85 Fps
LoL

Avg. Fps @ 1080p on Max
215 Fps Slightly Better LoL Fps.
+5%
204 Fps
Overwatch

Avg. Fps @ 1080p on Max
124 Fps 125 Fps +1%
BF1

Avg. Fps @ 1080p on Max
111 Fps Better BF1 Fps.
+15%
97 Fps
The Witcher 3: Wild Hunt

Avg. Fps @ 1080p on Max
94 Fps +2% 92 Fps
Minecraft

Avg. Fps @ 1080p on Max
239 Fps 259 Fps Slightly Better Minecraft Fps.
+8%
DOTA 2

Avg. Fps @ 1080p on Max
163 Fps Slightly Better DOTA 2 Fps.
+8%
150 Fps
WoW

Avg. Fps @ 1080p on Max
96 Fps 97 Fps +1%
WoT

Avg. Fps @ 1080p on Max
133 Fps +4% 128 Fps

Market Share

Based on 55,827,326 GPUs tested.

See market share leaders

Market Share

Market Share (trailing 30 days)
1.43 % Hugely higher market share.
+198%
0.48 %
Value

Value For Money
62.7 % 63.6 % +2%
User Rating

UBM User Rating
69 % 69 %
Price

Price (score)
$400 $359 Cheaper.
+10%
Age

Newest
76 Months 59 Months More recent.
+22%
Parallax

Parallax occlusion mapping (Stones)
147 fps Slightly better peak texture detail.
+7%
137 fps
Splatting

Force Splatted Flocking (Swarm)
127 fps Slightly faster peak complex splatting.
+6%
120 fps
Parallax

Avg. Parallax occlusion mapping (Stones)
131 fps Slightly better texture detail.
+8%
121 fps
Splatting

Avg. Force Splatted Flocking (Swarm)
118 fps Slightly faster complex splatting.
+6%
111 fps

ADVERTISEMENT

Custom PC Builder (Start a new build)
Build your perfect PC: compare component prices, popularity, speed and value for money.

CHOOSE A COMPONENT:

CPU GPU SSD HDD RAM MBD

Graphics Card Rankings (Price vs Performance)
October 2022 GPU Rankings.

We calculate effective 3D speed which estimates gaming performance for the top 12 games. Effective speed is adjusted by current prices to yield value for money. Our figures are checked against thousands of individual user ratings. The customizable table below combines these factors to bring you the definitive list of top GPUs. [GPUPro]

ADVERTISEMENT

Group Test Results

  • Best user rated — User sentiment trumps benchmarks for this comparison.
  • Best value for money — Value for money is based on real world performance.
  • Fastest real world speed — Real World Speed measures performance for typical consumers.

How Fast Is Your GPU? (Bench your build)
Size up your PC in less than a minute.

Welcome to our freeware PC speed test tool. UserBenchmark will test your PC and compare the results to other users with the same components. You can quickly size up your PC, identify hardware problems and explore the best upgrades.

UserBenchmark of the month

Gaming

Desktop

ProGaming

CPUGPUSSDHDDRAMUSB

How it works
  • — Download and run UserBenchmark.
  • — CPU tests include: integer, floating and string.
  • — GPU tests include: six 3D game simulations.
  • — Drive tests include: read, write, sustained write and mixed IO.
  • — RAM tests include: single/multi core bandwidth and latency.
  • — SkillBench (space shooter) tests user input accuracy.
  • — Reports are generated and presented on userbenchmark. com.
  • — Identify the strongest components in your PC.
  • — See speed test results from other users.
  • — Compare your components to the current market leaders.
  • — Explore your best upgrade options with a virtual PC build.
  • — Compare your in-game FPS to other users with your hardware.

 Frequently Asked Questions

 Best User Rated

  •   

    Nvidia RTX 3060-Ti

  •   

    Nvidia RTX 3070

  •   

    Nvidia RTX 3050

  •   

    Nvidia GTX 1660S (Super)

  •   

    Nvidia RTX 3080

  •   

    Nvidia RTX 2070S (Super)

  •   

    Nvidia GTX 1650S (Super)

  •   

    Nvidia GTX 1060-6GB

  •   

    Nvidia RTX 2060

  •   

    AMD RX 6600-XT

  •   

    AMD RX 5600-XT

  •   

    Nvidia RTX 2060S (Super)

About  •  User Guide  •  FAQs  •  Email  •  Privacy  •  Developer  •  YouTube
Feedback

GeForce GTX 1070 Ti vs.

GTX 1080

Nvidia’s RTX 20-series graphics cards have left many feeling underwhelmed. Yes, they’re all faster than the ‘equivalent’ previous generation models, and things like DLSS and ray tracing could widen the gap in the future, but every tier of the RTX line moved up in pricing. The RTX 2070 effectively takes over the 1080 price, the RTX 2080 inherits the 1080 Ti price, and the RTX 2080 Ti sits in the Titan Xp’s gold-plated throne. Understandably, many are wondering if they can just get by with the previous generation, and lest there was any doubt, the answer is unequivocally yes. You do not need the best graphics cards available to play games, and even four year old cards like the GTX 980 are still doing fine at reduced settings.

Things are further complicated by the lack of GTX 1080 and 1080 Ti cards. Nvidia has stated that these cards are being phased out, and that makes sense—their replacements are here in the form of the RTX lineup. But what if you want to spend less than $500? The best candidate right now is the GTX 1070 Ti, and with prices on the 1080 on the way up, it may soon be the only choice. But just how much performance do you give up by going with a GTX 1070 Ti vs. the GTX 1080? And what are the major differences?

Let’s start with the differences first, as that’s pretty straightforward. As discussed in our GTX 1070 Ti review, the features are a mix between the GTX 1070 and 1080. The 1070 Ti inherits the 1070’s use of GDDR5 memory, as opposed to the 1080’s GDDR5X. That gives it the same bandwidth as the 1070, and 20 percent less bandwidth than the 1080—or 27 percent less bandwidth than the 11Gbps 1080 models, which were never very common.

You do not need the fastest GPUs available to play games, and even four year old cards are still doing fine at reduced settings.

Nvidia offsets the loss of memory bandwidth by increasing the number of GPU cores. The GTX 1070 Ti disables one SM on the GP104 core, resulting in 2,432 CUDA cores, which is 27 percent more than the vanilla GTX 1070. Compared to the 1080’s 2,560 CUDA cores, it’s a five percent drop in core count. Clockspeeds are also a mix between the 1070 and 1080. The base clock (which in my experience you almost never see) is the same 1,607MHz as the 1080, while the boost clock is the same 1,683MHz as the 1070—50MHz less than the 1080’s 1,733MHz boost clock. But that’s only the reference clocks, and in many cases there are factory overclocked models that all end up running at 1,800-1,850MHz while playing games, regardless of which GTX model you’re using.

If you combine those three changes—memory bandwidth, core counts, and GPU clocks—you end up with a card that should slot in right between the GTX 1070 and GTX 1080. And that’s precisely what happens when we run the benchmarks. Here’s the overall summary for all recently tested graphics cards on 14 games, at four different setting/resolution combinations. I’ve highlighted the 1070 Ti and 1080, but you can also compare performance with the new RTX cards, or AMD cards, or anything else.

Image 1 of 4

At 1080p medium quality, the GTX 1080 is only five percent faster than the GTX 1070 Ti. That’s because memory bandwidth is less of a factor here. There are a few games where the gap is wider, and others where it’s smaller, but it’s very close overall. Moving to 1080p ultra doesn’t change things—the overall gap is only five percent in favor of the GTX 1080. And at 1440p ultra, it’s still only 5.5 percent. Notice the trend yet? But finally, at 4k ultra, the 1080 takes a commanding lead of… 6.7 percent.

In other words, there’s almost no case where the GTX 1080 will handle games significantly better than the GTX 1070 Ti. Across the entire test suite of 14 games and 4 resolutions—56 comparison points—the GTX 1080’s largest lead is in Strange Brigade, with a 10-14 percent advantage depending on the resolution. And if you’re curious, the GTX 1070 Ti on average is about 15 percent faster than the GTX 1070, so the boost in core counts and clockspeed has a bigger impact than the lack of memory bandwidth.

That’s good news, because the GTX 1080 is on the way out, and for the time being the GTX 1070 Ti remains readily available, with a current starting price of $380 after $20 rebate card. What’s that look like in terms of overall value, and FPS per monetary unit?

Image 1 of 6

If you’re looking strictly at the GPU prices and performance, the midrange RX 570/580 and GTX 1060 cards take top honors, regardless of currency. Faster but more expensive GPUs deliver diminishing returns. On the flipside, include the base system price in the matrix, overall value skews heavily toward faster GPUs. The RTX 2080/2070 and GTX 1080 Ti/1070 Ti claim the top four spots, with the 1080 ranking between 5th and 8th depending on the market—and the 1070 sitting at 5th to 9th, if you’re wondering. (For reference, I used a midrange build that includes an i5-8600K, modest liquid cooling, Z370 motherboard, 16GB DDR4, 480GB M.2 SSD, NZXT H500 case, and a 650W Gold PSU.)

So why did Nvidia even release the 1070 Ti in the first place—isn’t it basically the same market as the GTX 1070? If you look at the charts and focus on AMD’s Vega cards, you’ll find the answer. For raw performance, the GTX 1070 Ti manages to consistently beat the RX Vega 56, and even claims a few victories over the Vega 64 in select games—a feat the vanilla 1070 couldn’t quite manage. The GTX 1070 is still a good option as well, since it’s typically priced about $50 less than the 1070 Ti, but with the post-crypto-bubble GPU prices it’s not quite as competitive.

There’s a second important aspect as well, however. The GTX 1080 and 1080 Ti are the only GPUs to use GDDR5X, and with such limited demand the prices tend to be higher. By combining a slightly slower core with the more readily available GDDR5, Nvidia ends up with the best of both worlds. GDDR5X is likely a major part of why Nvidia has opted to discontinue production of the 1080 and 1080 Ti, while for the time being the 1070 and 1070 Ti remain easily available. (Don’t be surprised if that changes in the next month or two with the anticipated arrival of an RTX 2060, however.)

Bottom line: if you’re in the market for a higher performance graphics card but can’t quite stomach the pricing on the RTX 2070 and RTX 2080, the GTX 1070 Ti is the best graphics card for around $350 right now.