Best video card for 250$ candian?

Ultra_Gamer

Honorable
Sep 13, 2015
81
0
10,630
Hey, Looking for a video card that can play gta v fallout 4 battle steep, at medium to high settings with decent frame rates.

Here is my build so far


Motherboard: Asus M5A78L- M/USB3 Micro ATX AM3+

RAM: HyperX Fury 8GB DDR3-1866

Power Supply: Solid Gear 650W ATX Power Supply

Case: Ultra Etorque X1 Mid-Tower Gaming Case

CPU: AMD FX-8300 3.3GHz 8-Core

Thanks,
 
Solution
Listed in order of BF4 performance @ 1080p and with Canadian newegg prices

For $265 GTX 1060 (97 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137037

For $365 GTX 480 (70 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137023

For $230 RX 470 (60 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137050

For $224 GTX 1050 Ti (38 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137054

For $185 - RX 460 (32 fps) - http://www.newegg.ca/Product/Product.aspx?Item=9SIA4FT52B0834

For $179 GTX 1050 (34 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137059

Best option: 1060 if ya can swing extra $15.... investing $15 over budget takes you from 60 to 97 fps ...

RCFProd

Expert
Ambassador
Best value for your build:

Video Card: Zotac GeForce GTX 1050 Ti 4GB Mini Video Card ($189.25 @ Vuugo)
Total: $189.25
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2016-11-18 13:03 EST-0500

Best possible GPU within price range:

Video Card: MSI Radeon RX 470 4GB ARMOR OC Video Card ($229.99 @ Newegg Canada)
Total: $229.99
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2016-11-18 13:04 EST-0500
 
Listed in order of BF4 performance @ 1080p and with Canadian newegg prices

For $265 GTX 1060 (97 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137037

For $365 GTX 480 (70 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137023

For $230 RX 470 (60 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137050

For $224 GTX 1050 Ti (38 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137054

For $185 - RX 460 (32 fps) - http://www.newegg.ca/Product/Product.aspx?Item=9SIA4FT52B0834

For $179 GTX 1050 (34 fps) - http://www.newegg.ca/Product/Product.aspx?Item=N82E16814137059

Best option: 1060 if ya can swing extra $15.... investing $15 over budget takes you from 60 to 97 fps ... proverbial "no brainer" ... if ya can swing the $15

Best Under $250: RX 470 ... however again, the 1060 crushes 470 (62% faster) if ya can add that extra $15 to budget (to be fair $35 more than 470)

Best Buy Under $225: GTX 1050
 
Solution


That's correct, but hardly significant. It's not 97 to 60 ... it's 92 to 60 which eiteher way doesn't exactly put them on an "even keel" . Worth mentioning that the the 3GB still crushes the $120 more ($385) RX 480 in BF4 by a whopping 31 % ... and 10% in GTAV.

The 1060 3GB competes with the 480 not the 470... the RX 480 and 3GB 1060 pretty much run neck and neck "across the board" when both are at stock speeds, but in the OPs listed games, the 480 suffers quite a bit. It also must be mentioned that the MSI 3 GB card manually overclocks another 14.5% ... in comparable tests, the Asus Strix 470 only managed 6.5% which substantially widens the gap.

Another way to look at this is " what does one get for their extra investment" ? ... the 3GB is 52% faster for a price increase of just 15%, in other words, a ROI (Return on Investment) of 346 %.

... or if ya wanna look at it like a "horse race" as the saying goes, to really put in perspective, if it was the Belmont States horse race, in which Secretariat set the record "winning margin ever" by 3 lengths, the RX 470 loses to the 3Gb 1060 by 273 lengths.... 293 lengths is we "juice their oats and "overclock them :)

BTW, if you never seen that... it's incredible... easily the most dominating performance in any championship sports event in history.

https://www.youtube.com/watch?v=AG_27cCW5bw

https://en.wikipedia.org/wiki/1973_Belmont_Stakes
 

RCFProd

Expert
Ambassador
Those percentages seem to be off to quite the extreme.

'' the 3GB is 52% faster for a price increase of just 15%, in other words, a ROI (Return on Investment) of 346 %.

52% faster? That's an insane value and I can't find where this number comes from, unless It's from a single benchmark in Battlefield 4 and here come 2 important factors to play:
- Polaris GPU's are known for specifically running worse in Battlefield 4, and BF4 pretty much alone.
- Several driver updates for Polaris have signficanty (+/- 10%) improved CPU overhead performance in DX11 (Mainly CPU intensive) titles including Battlefield 4.
- Release benchmarks of Polaris GPU's therefor aren't accurate on that fact alone. Not only have they increased CPU overhead performance, they've also consistently started scoring higher in synthetic benchmarks scores over each AMD driver update. There have been plenty since release.


Actually, in most titles, the RX 470 is very close to the GTX 1060 3GB in performance, whilst the RX 480 4GB usually edges on the GTX 1060 3GB. The performance difference overall between the RX 470 and the RX 480 is just ~10 fps on average.

The RX 470 comes as close as you'd wish to an AMD version of a GTX 970. And AMD drivers have shown a decent increase in performance and lowered CPU overhead in several DX11 titles, which AMD GPU's used to suffer from. Games with CPU overhead report up to 10% performance increase.

It must be importantly noted that Polaris GPU's are running a lot worse in Battlefield 4 than their actual capability. They have seen an increase in performance since then, I'm sure your numbers are release benchmark numbers. I think Battlefield 4 is one of the titles where Polaris has seen a 10% increase in CPU overhead performance in Battlefield 4.

As it stands, based on several games that have been released after 2013, is that your numbers aren't close to being correct. Several tests from Digital Foundry and just a few benchmarks (A few, AMD drivers from today are a step up in performance than the ones from 6 months ago, those won't be an accurate representation of current performance) will show just how close the RX 470 4GB and GTX 1060 3GB are to each other performance. Actually to the point I'd say that it comfortably takes the price/performance throne from the GTX 1060 3GB, with the extra 1GB of VRAM in mind for a few particular titles that do need it.

[video="https://www.youtube.com/watch?v=CNNRGH3gMvg"][/video]

Battlefield 1 shows the bigger difference in performance between the GTX 1060 3GB and RX 470 4GB than I've seen in other benchmarks, still pretty marginal when taking price into account.

1440p.png


 


Let's go back and read what I wrote:

Listed in order of BF4 performance @ 1080p

That's correct, but hardly significant. It's not 97 to 60 ... it's 92 to 60 which either way doesn't exactly put them on an "even keel" . Worth mentioning that the the 3GB still crushes the $120 more ($385) RX 480 in BF4 by a whopping 31 % ... and 10% in GTAV.

Actually, in most titles, the RX 470 is very close to the GTX 1060 3GB in performance, whilst the RX 480 4GB usually edges on the GTX 1060 3GB. The performance difference overall between the RX 470 and the RX 480 is just ~10 fps on average.

perfrel_1920_1080.png


470 has an average score of 83% in the test suite
1060 GB and 480 both have 100% average


1060 3GB OC's 14.5% (94.5fps in BF3)
perf_oc.png


Don't wanna clog the board w/ images but
480 8GB manages just 8.6% (83.6%)
480 4 Gb manages just 6.5% (79.0 fps)

The RX 470 comes as close as you'd wish to an AMD version of a GTX 970. And AMD drivers have shown a decent increase in performance and lowered CPU overhead in several DX11 titles, which AMD GPU's used to suffer from. Games with CPU overhead report up to 10% performance increase.

Nope 83% isn't close as I'd wish to the 1060 3Gb or 480 8GB... to be considered equivalent, that means equal. Again using the horse race analogy, losing by 147 lengths is not a tie.

As it stands, based on several games that have been released after 2013, is that your numbers aren't close to being correct.

First off, the subject at hand .. as stated by the OP is "a video card that can play gta v fallout 4". The context of this discussion and each of my posts has therefore been limited to the stated topic. "Several games that have been released since 2013" are off topic.

Second, they aren't "my numbers". If you don't like the published results, take it up with TechPowerUp, Guru3d or any other well respected unbased site which shows the same thing

As I had showed the advantages of the 1060 over the 480 in BF4 (31%) are far greater than GTAV (10%)

Guru3D didn't test BF4, but they did GTAV

Reference GTX 1060 6GB = 90 fps in GTAV
Reference RX 480 8GB = 84 fps
MSI RX 470 w/ factory OC = 81 fps
Reference 470 = 76 sps (90 / 76 = 18%)

They also tested BG Hardline which I guess is in the picture if BF4 is

Reference 1060 = 73 fps
MSI 470 8GB w/ factory overclock = 57 (73 / 57 = 28%)

Can I go out and fine games which show the 480 doing better ? Of course. But a) the OP is looking for which card to choose specifically for those games b) the 480 costs a whole lot more c) the 1060 overclocks a helluva lot better than the 470 / 480 cards.

When i am trying to help someone that is "Looking for a video card that can play gta v fallout 4 battle steep, at medium to high settings with decent frame rates."

I have to say that:

-92 fps in BF4 is a helluva lot more "decent" than 71 fps for the 480 and 60 fps for the 470 ... even more so when overclocked, that goes to 105, 77 and 64 fps respectively for a 36% advantage over the 480 and a 64% advantage over the 470. In the horse race, that's a win by 322 lengths

-95 fps in GTAV is a helluva lot more "decent" than 87 fps for the 480 and 73 fps for the 470 ... even more so when overclocked, that goes to 109, 93 and 78 fps respectively for a 17% advantage over the 480 and a 40% advantage over the 470. In the horse race, that's a win by 243 lengths

EDIT:

I do have to apologize for my misread... Appears I got a momentary case of dyslexia seeing "4 battle" in the OP and seeing BF4 in my head... my bad. Ignore all the BF4 stuff :)

So lets do Fallout 4

-82 fps in Fallout 4 is a helluva lot more "decent" than 70 fps for the 480 and 67 fps for the 470 ... even more so when overclocked, that goes to 94, 76 and 71 fps respectively for a 26% advantage over the 480 and a 32% advantage over the 470. In the horse race, that's a win by 267 lengths

So what we have here for the 2 games specifically noted is in the 2 races the 1060 3GB won by 267 lengths in one race and 243 in the other.

I did a Yahoo search on Battle Steep but couldn't find anything

EDIT 2:

As for the benchmarks you added, there is no BF1 on his agenda








 

RCFProd

Expert
Ambassador
I've been keeping track of AMD performance and this is what the problem I'm stumbling on: I can't back-up how much Polaris GPU's have improved performance wise since release. Why is that? When GPU's are released, they are reviewed right then and there, and that's the performance reference everyone will look at. Certain graphics cards aren't being re-reviewed in a couple of months, which in case of AMD Polaris would be more than just convenient and actually justice too. There has been a genuine improved since the RX 470 and RX 480 have been initially tested in July.

The only way to find recent benchmarks are in new games that have been recently benchmarked. However those few games aren't an actually accurate representation, because the game might not be optimised well or another title might favor another GPU over the other. It's hard for me to put it into a clear perspective, because numbers of recent drivers for AMD GPU's can't actually be found and compared.

The best I could do is find benchmarks from recent games, here is the outcome of those:

index.php

skqz0G.jpg

index.php

 

Ultra_Gamer

Honorable
Sep 13, 2015
81
0
10,630
Sorry for the typo, it was battle feild 4 not "4 battle" thanks guys for the opinions lot of knowledge here i will review it, :). here is the link for steep. http://www.game-debate.com/games/index.php?g_id=23962&game=Steep.(Edit) how will my cpu and ram fair with these video cards, and the rest of my build?
 


The performance of GFX cards increases with every driver release on both sides .... multi card and 3D profiles get upgraded, new games are upgraded and performance gets tweaked. Been following this industry for 30 years and have only once seen an article on a reputable site claiming "new drivers" changing the "balance of power" so to speak. That article was written right after a new set of drivers came out for AMD and 2 days after new drivers came out for nVidia. Of course, with the author having an agenda, he didn't redo the tests despite the fact that he published 2 days after nVidia also had it's own drivers. The article was hence soundly and justifiably trashed....and when compared latest driver to latest driver, all was with only inconsequential differences as it was before the updates.

If, as you say, reviewers don't go back and retest, then ya gotta explain that when each new card is tested, the results of each of the previous cards reviewed change ... look at techpowerup for example and you will see cards getting different scores in each subsequent test

On Aug 10th, 2016 ... the 470 got 59.6 fps
https://www.techpowerup.com/reviews/ASUS/RX_460_STRIX_OC/12.html

On Oct 25th, 2016 ... the 470 got 60.6 fps in COD:BO
https://www.techpowerup.com/reviews/MSI/GTX_1050_Ti_Gaming_X/11.html

On Nov 16th, 2016 ... the 470 got 60.5 fps
https://www.techpowerup.com/reviews/Zotac/GeForce_GTX_1060_Mini_3_GB/11.html
Similarly in BF4, the 470 went from 59.6 to 60.8 fps .... so clearly, contrary to claim, they are being retested and no... driver improvements with the the 4xx series cards are not erasing the difference .... BF4 (one of the OPs interested games is still at 90+ fps ... so it's not like the improvement from 59.6 to 60.8 from "driver improvements" is going to make a difference when the competition is 50% faster

https://www.techpowerup.com/reviews/Zotac/GeForce_GTX_1060_Mini_3_GB/15.html

Fallout 4
Nov 16 = 67.6 (470 4 GB) / 87.5 (1060 3 GB)
Sep 06 = 67.5 (470 4 GB) / 82.3 * (1060 3 GB)

*Note that the MSI has a 4% factory OC so that number is slighty inflated should be around 79 fps w/o factory OC ... but still ... 1060 improved 6%..... 470 improved 0%

GTAV
Nov 16 = 73.7 (470 4 GB) / 92.0 (1060 3 GB)
Sep 06 =73.2 (470 4 GB) / 95.4* (1060 3 GB)
*Note that the MSI has a 4% factory OC so that number is slighty inflated should be around 91.7 fps w/o factory OC ... accounting for the factory OC, both cards remain relatively stagnant w/ < a full fps improvement. The 1060 3Gb is still 25% faster despite the 0.5fps increase in the 470 from "driver improvements"

We can go back and forth and hand pick games but the OPs focus is on 3 games and ... a) it's not close in any of them, b) driver improvements are clearly not changing **anything** and c) the difference in OC ability of the two camps insures that any driver improvements **if they had occurred**, would be wiped out several fold by the inability of the AMD cards to break single digits in their OC %.

We can both run around the web and hand pick games that support either position, but the only relevant ones will be BF4, Fallout 4 and GTAV.
 


I guess "my head" knew what you were looking for :)

As for The game "Steep", the reviewer states the 1060 "wipes the board" with the GTX 480 (3:42 mark) at 1440p (and less).

https://youtu.be/HDFuhRkX6HQ?t=217

 

RCFProd

Expert
Ambassador
Sure, It's just weird that my RX 480 8GB is getting 80-85 fps on average in Battlefield 4 with Maxed settings + 10% Resolution scale increase, whilst reported values are lower, not even because they're reference GPU's as mine is similarly clocked as a reference (1300Mhz Core clock vs. 1266/1288Mhz). And this is with multiplayer, TechPowerUp tested singleplayer which would give me another boost.

You've got the statistics right based on those games, however I'm going to run a test in Battlefield 4 to see how it performs in singleplayer.