The Best sub-$200 graphics card – GTX 1050 Ti vs. RX 470 / RX 460 4GB
Looking at buying a sub $200 US graphics card? Well good, I might be able to help steer you in the right direction. On hand we have the newly released GTX 1050 Ti which will be going head to head with AMD’s RX 470 and RX 460 4GB graphics cards. On the menu we have 25 games and that should be more than enough to give us a clear picture of where things are at.
Once again as this is a budget GPU comparison we will be using our Core i3-6100T test machine. This humble little rig should represent the kind of processing power the vast majority of gamers spending less than $200 US on their GPU will be wielding. The latest AMD and Nvidia drivers have been used for testing and updates to existing results have been made where necessary. The results are based on an average of three runs and all testing takes place at 1080p.
For those wondering the RX 470 will be represented by the Asus’s ROG STRIX 4GB OC model while the RX 460 4GB card being used is the Sapphire Nitro model. Meanwhile the GTX 1050 Ti card being used is the little MSI 4G OC model. So with that let’s get to the benchmarks…
Starting with Assassin’s Creed Syndicate, a title that generally favors the GeForce GPUs, we find that the 1050 Ti is 15% slower than the 470. Still with an average of 64 fps the 1050 Ti was still able to deliver very playable performance.
The RX 470 hands the 1050 Ti a much more significant blow in Batman Arkham Knight, here the GeForce GPU was 27% slower despite averaging 62 fps. Meanwhile the 470 never dipped below 55 frames per second at 1080p using the ultra quality settings.
For now we prefer the DX11 API for Battlefield 1 but we will check the DX12 results next. Still using DX11 the 1050 Ti was still 26% slower than the RX 470, so another decent win for AMD here then.
This is why we prefer DX11 for now, while the average frame rate looks great DX12 suffers from constant frame drops.
The newest game in our battery of benchmarks is Call of Duty Infinite Warfare, a game universally loved by all. Here the GTX 1050 Ti was 36% slower than the RX 470 averaging just 49 fps to the Radeon’s 76 fps.
Here we have an old favorite for testing, Crysis 3. Can the GTX 1050 Ti play Crysis? Well yes but only just, dipping as low as 33 fps at times. Meanwhile with a 38 fps average the 1050 Ti was 27% slower than the RX 470.
The 1050 Ti is just 3% faster than the RX 460 4GB in Deus Ex: Mankind Divided, so unsurprisingly it trails the 470 by a rather massive 42% margin. That said it is interesting to note that the 470 does dip quite low in relation to the average frame rate. As a result comparing the minimum frame rate sees the 1050 Ti trail by just a 16% margin.
DiRT Rally is the polar opposite to Mankind Divided and I don’t mean because, one is a racing simulator and the other is a first person shooter. Rather I am referring to how AMD and Nvidia GPUs handle these titles. For the first time the 1050 Ti moves ahead of the RX 470, delivering 21% more performance. In fact even the standard 1050 is faster than the 470 here, this title very clearly favors the green team.
Doom using Vulkan places nicely into the hands of AMD. Here the 1050 Ti was 44% slower than the RX 470 though I should note that both delivered well over 60 frames per second at all times. Also worth mentioning is the fact that the 1050 Ti was just 5% faster than the 4GB 460 in this test.
The RX 470 remains ahead for the F1 2016 test as the 1050 Ti finds itself trailing by an 18% margin. That said the 1050 Ti was also 15% faster than the 4GB 460, so not a bad result for Nvidia overall.
Testing Fallout 4 using the high quality preset with TAA enabled we find that the GTX 1050 Ti is good for an average of 70 fps at 1080p. Despite that impressive result it was almost 20% slower than the 470.
Here the 1050 Ti enjoys a 29% performance advantage over the 4GB 460, while it was just 6% faster than the vanilla 1050. As was the case when testing Mankind Divided we see that although the 1050 Ti is considerably slower than the RX 470 when comparing the average frame rate, the minimum result is much closer. The 1050 Ti was 23% slower when comparing the average frame rates but just 4% slower for the minimum frame rate, a curious result from Far Cry Primal indeed.
Moving to Gears of War 4 we find some pretty competitive results, largely helped by the fact that the 1050 Ti is over 30% faster than the standard 1050 here. This meant that the 1050 Ti was just 12% slower than the RX 470, so a strong result for Nvidia here.
Continuing the strong run the 1050 Ti takes charge in Grand Theft Auto V, rendering 3fps more than the RX 470 to make it 5% faster. The fact that the 1050 Ti maintained at least 60 fps at all times in our test was impressive.
Okay guys rule number 3 of Benchmarkland, always stay hydrated. I have forgotten this rule from time to time and really paid for it the next day, so make sure you take a moment to stop and get some water. DRINKS WATER … Okay let’s get back at it.
The extra 33% performance provided by the RX 470 doesn’t go unnoticed in Mafia III, in fact almost reaching a 60fps average means we might even be able to use the medium quality settings, exciting stuff. Anyway the 1050 Ti was 25% slower than the 470 but, 30% faster than the 4GB 460.
The GTX 1050 Ti was 20% slower than the RX 470 in Shadow of Mordor when comparing the average frame rate.
Benchmarking with Mirror’s Edge Catalyst we find that the 1050 Ti is just 11% slower than the RX 470 as it averaged 66 fps. It was also 32% faster than the 4GB RX 460.
The 1050 Ti is already pushing over 100 fps in Overwatch at 1080p using the ultra quality settings so for most there won’t be a real need to go faster. Nevertheless the RX 470 was almost 20% faster averaging 134 fps.
The Radeon graphics cards don’t really like Steam’s DirectX 11 version of Quantum Break and here we see the minimum frame rate of the RX 470 dipping below that of the standard 1050. Despite that the average frame rate was still very strong.
Yet again we find another game where the RX 470 appears to dominate the 1050 Ti, until you look at the minimum frame rates. The 1050 Ti was on average 25% slower but just 4% or 1 fps slower when checking the minimum frame rate.
The 1050 Ti was good for at least 60 fps in Star Wars Battlefront and yet this meant it was still 27% slower than the RX 470 which easily cracked the 70 fps barrier.
We couldn’t forget The Witcher 3: Wild Hunt, here the 1050 Ti was 27% slower than the 470 but 28% faster than the 4GB RX 460.
Wow, AMD dominates Rainbow Six Siege, here the 1050 Ti wasn’t just 39% slower than the RX 470, but also 4% slower than the 4GB RX 460.
For those not yet bored with The Division the weapon of choice here looks to be the RX 470 as it was 45% faster than the 1050 Ti.
AMD lays the smack down in Titanfall 2. Here the 1050 Ti was 38% slower than the RX 470 averaging just 50 fps to the Radeon’s 81 fps. That said the 1050 Ti was 32% faster than the 4GB RX 460.
Well we made it, the last game benchmark. Total War: Warhammer was tested using DirectX 12 and here the 1050 Ti was just 6% slower than the 470 in what is a very surprising result. Not only that but the GeForce GPU was actually 4% faster when comparing the minimum frame rate.
Well there you have, an indepth look at how the GTX 1050 Ti stacks up again both the RX 470 and RX 460 4GB cards. Now to wrap things up let’s take a look at the results from all the games tested in one big graph.
First up we are checking out how the GTX 1050 Ti stacks up against the cheaper RX 460 4GB. In terms of performance it is pretty much a landslide. The only game to favor the 460 was Rainbow Six Siege by a small margin which performance was also close in Mankind Divided and Doom using Vulkan.
However overall the 1050 Ti was on average 29% faster and this figure was no doubt helped by massive wins in titles such as Quantum Break, Overwatch, Gears of War 4 and DiRT Rally. Given the small difference in price the 1050 Ti certainly looks like the smarter option here but we will get to that soon in the cost per frame analysis.
Okay so here the bars are all pretty much going the other way, that is to say for the most part the 1050 Ti was slower than the RX 470.. No surprises there. The 1050 Ti was on average 22% slower, findings it’s way out in front for just two titles, DiRT Rally and Grand Theft Auto V.
The 1050 Ti was crushed in all the usual suspects though we can now add Titanfall 2 and Call of Duty Infinite Warfare to the list.
Before we jump to the cost per frame figures let’s just take a quick look at efficiency by comparing power consumption against the average frame rates.
Efficiency has been where many of these GPU battle’s has been won and lost over the past few years. As you can see on average the RX 470 was 29% faster than the GTX 1050 Ti, however that extra performance came at a cost. Entire system consumption increased by over 70% from 118 watts to a much more hungry 202 watts. Worse still the RX 460 was 23% slower on average by pushed system consumption 27% higher.
These figures won’t be hugely impactful for everyone. If you have a standard desktop rig with a 350 watt or greater power supply, aren’t worried about a little extra heat but do care about absolute performance then the RX 470 still represents exceptional value. However if operating temps, volume and overall consumption are things you very much care about, then the 1050 Ti might be the more desirable option.
Starting with the scatter plot we see that the GTX 1050 Ti and RX 470 are actually quite similar in terms of value, despite the GeForce GPU being over 20% slower. Meanwhile the RX 460 4GB looks to be pretty poor value while the 2GB version is roughly on par with the GTX 1050.
Moving to the bar graph which I am told many of you prefer, arranging the data by frame rate we see the RX 470 way out in from with the second lowest cost per frame.
Arranging the same graph by cost per frame we find the 1050 2GB provides the best bang for your buck as it just edged out the RX 470. Meanwhile the 1050 Ti was 6% more costly per frame when compared to the RX 470.
In short the GTX 1050 Ti was on average 22% slower than the RX 470, meanwhile the $140 MSRP means it is just 18% cheaper. There were a few games where the RX 470 didn’t look nearly as impressive when taking the minimum frame rate into account but we hope AMD can address these performance issues in future driver updates. It has to be said they have been doing an exceptional job with their drivers lately.
At first picking between the 1050 Ti and RX 470 doesn’t seem that easy, and well that’s because they are both great options. Once again as I see it, there really isn’t a wrong choice. The 1050 Ti is clearly the more efficient GPU but the RX 470 is faster and should work out being the better value choice.
For me the jury is still out on the whole DX12 thing and honestly, will either of these GPUs really be fighting the DX12 battle? I seriously doubt it. For upcoming games that support both DX11 and DX12, we will likely find a similar situation to that of Battlefield 1 and by that I mean a situation where GeForce graphics cards should use the DX11 API, while AMD will probably work best with DX12.
Going forward we might even test Battlefield 1 on Nvidia GPUs using DX11, while testing AMD GPUs using DX12. Anyway let’s not get into that here.
Getting back to the matter at hand I am sticking with our original findings back in August. If you are going to get the RX 460, get the 2GB model. For those chasing 60 fps performance, the 4GB version has little to offer in most games, so it’s $20 US you didn’t need to spend. Not only that but for $10 less we prefer the GTX 1050 anyway.
On the other side of it rather than spending $120 on the RX 460 4GB we would highly recommend spending $140 on the GTX 1050 Ti. That 17% increase in cost will net you almost 30% more performance and a considerably more efficient graphics card.
However, if you can afford to go as high as $170 then the RX 470 becomes very appealing and even opens up the option of 1440p gaming. So in short if we had around $100 to spend we would get the GTX 1050 2GB, for between $120 – $150 we would get the GTX 1050 Ti and for between $150 – $190 the RX 470 seems like the way to go.
For those spending $200 or more we previously went with the GTX 1060 3GB but it might be time to revisit that comparison. With the RX 470’s revised price and AMD’s new found form the 1060 3GB may no longer be the best option here.
Alright I am going to stop there before I get into another comparison and drag this video out even longer. I hope the testing here was useful and it has given you guys the information you need to draw your own conclusions. We will of course continue to monitor the situation as it develops with the arrival of new and exciting gaming titles. As it stands I am confident you guys will truly enjoy gaming on which ever GPU you choose, so at least there is that. Until next time, stay classy!