Gadgetory

Ubisoft chivalry clones I guess by surprise when benchmarking the for honor beta over the last few days the game makes use of high detailed textures and relatively complex geometry to orchestrate what comes across as a deeply detailed environment even the animations like running down the stairs are animated to a point of detail we don't often see in multiplayer brawlers today we're benchmarking our Nvidia and AMD Suites of GPUs in the four honored beta that was released just for the last few days before that this coverage is brought to you by catalyst gaming Mint's and energy boost designed for gamers while eliminating the dyes sugar and calories and chemicals that traditional energy drinks contain use code gamers in excess at the link in the description below for 5% off before getting started here the usual disclaimer for honor right now is in beta the game will fully ship on the 14th so it's really not that far away most things are probably pretty final but there's still a few things that are missing from the game for example the lower resolution texture settings so when testing we're basically limited to high resolution textures that means we don't have a good idea of how to gigabyte cards will handle when you have the ability to reduce the texture quality further there's the question of drivers Nvidia presently has some beta ready drivers for the game for honor but AMD we did email them they don't have any ready yet we asked them if they had a pressed version that would be ready for this benchmark and the answer was no not yet they will be ready on day one launch day and we'll have to revisit at that time to see how their drivers improve performance but for now we're still testing anyway overall the performance wasn't bad it's just that driver tuning may improve some of the low-end performance in terms of the frame times that we see here in this benchmark so keep all of that in mind the graphics settings and fro on ER are straightforward the game again is locked to high texture quality and defaults to using temporal anti-aliasing for a frame to frame AAA that reduces shimmering and marching ants effect and the rest is your usual mix of geometric detail dynamic shadows and environment detail to learn specifically what each setting does check out our article link to the description below it has some definitions of all the settings in the game for the usual research before doing the testing we found that the game's built-in benchmark is really not reliable it makes the game look a lot worse than it is in actuality what benchmarking even with just with bot matches we see generally the same performance in a bot match as in a multiplayer match so that's a good thing they're very consistent between each other the maps are all pretty consistent within a couple FPS of each other and then the built-in benchmark is just it's a lot worse and primarily in the frame time its frame times are lower and it's also not really built in a way that you're ever gonna be playing the game it's effectively a spectator mode that's on speed so we did not use the built-in benchmark we used our own which at some point has played during this video and you can see it if you wanted to replicate it we tested in a bot match but also validated with gameplay and found performance to be largely comparable when in and out of combat as the maps are complex to begin with and are what's causing most of the processing drivers used we're 17.1 dot one for AMD and 378 for nine for nvidia again for full test methodology check the link the description below first here's a brief sample of the games built-in bench that I was just talking about versus our own this is with a gtx 1066 gigabyte card just for a baseline and we'll look at the full chart with all the other GPUs in a moment but for now we're seeing a huge performance hit in the 1.1 percent and 1 percent low values and averages are also down overall because the built-in benchmark doesn't represent real gameplay at all we're going to stick with our own test course for the rest 4k is the start of that 4k is a lot more abusive than lower resolutions with this game more so than normally with the GT X 1080 FTW at Full Tilt's we're seeing an average FPS of 52 with lows close by at 48 and 45 fps the msi gtx 1070 gaming X operated at 42 FPS average also with accepted ly time to lows and the our X 480 isn't really meant to be a 4k gaming device but we threw it on here anyway with frame rates lower than 30 fps as we're seeing here with the 480 you actually risk being kicked by multiplayer lobbies because of Ubisoft's p2p design for networking so that's kind of an interesting problem you don't normally face moving on to the more reasonable 1440p we immediately see nearly every card on the bench is capable of an average FPS exceeding 50 aside from the rx 470 and GTX 1050 ti4 gigabyte cards the pack has expectedly by the GTX 1080 FTW at over 100 FPS average again with our frame times tightly packed enough that there's no noticeable stuttering the same remains true as we run through the ranks though for the most part the GTX 1070 is running at 82 FPS average or about 20% slower than the 1080 the 1060 gaming x6 gigabyte is at just under 60 FPS average and the stepping is pretty easy to follow thus far as about 20 fps between each of these three devices we next took the r9 390x at 57 FPS average and then the RX for 80 gaming acts with roughly the same performance and these lows are hurting here but we'll reserve judgment until the game's promised day zero quote-unquote release drivers for now let's highlight all the AMD one percent and point one percent low frame time performance issues as possibly related to unfinished drivers and we'll revisit when they shift the driver package the GTX 1063 gigabyte is next on the bench producing fps noticeably slower than it's six gigabyte counterpart note that this is partly because of the increased clock speed of the gaming X but the low performance dips are primarily resultant of the have vram and one removed SM from the GPU and these new rx 470 post performance above the 1050 Ti as our initial 1050 Ti review would lead you to expect and we're also entering territory where 1440p is not really necessarily likely on this hardware anyway so let's move on to 1080p with 1080p and the same extreme settings that we've been testing with the EVGA gtx 1080 FTW is now capable of running the game with max settings while sustaining a 144 hertz refresh compatibility the 1070 gaming X isn't too far behind again about 17 to 20 percent with a 1060 gaming x6 gigabyte just ahead of the RX 488 gigabyte card again the RX 480 and other devices from AMD are posting generally lower frame times that we didn't visually detect by human means anyway any serious and jarring stutters until running on the RX 462 gigabyte card drivers potentially play a role on this the r9 390x and our X 480 are effectively identical on performance both leading the GTX 1063 gigabyte SC card this is where we're starting to see the three vs. six gigabytes pitch popping up more noticeably the RX 478 runs more than playable and the lows aren't bad enough to ruin the experience even the lower and gtx 1050 TI is able to operate at about 60 FPS average with a 1050 SC not too far behind this final chart looks at FPS scaling when using a mid-range card so that be aggregate data between the 10 60 and 480 with this data we're using a baseline of high and showing the performance percentage offset from that you lose about 25% of the performance by going from high to extreme with these cards but the gains from high to medium are basically zero this might change as more texture resolutions are implemented into the final version because again we're stuck on high for now but low is showing a bigger impact than medium and high we're seeing dramatically improved performance but the game does actually look noticeably it worse so entering the conclusion let's start by reminding everyone that this is again an incomplete title there's going to be changes for sure to the game itself and that would include texture resolution that's the most obvious example and AMD will launch some drivers they tell us and Vidya tells us they're improving their drivers further so that means things all shift but this provides a good baseline it won't shift that much at most we can probably expect some frame time improvement from AMD we can expect some differences in the scallion from low to medium to high because of the texture resolution but extreme which is what we tested that should remain the same and you'll just have some offset from there by whatever percentage Andy and NVIDIA improved with their drivers so we've got a baseline from the baseline we can see that the game can be handled pretty well on really almost anything at 1080p from the 1050 Ti and up from Nvidia or the RX 470 and up from AMD the RX 460 as we stated with our review of the card is still not the best value for its competition the 1050 Ti and 1050 are pretty competitive there but the RX 470 does still make more sense than the 1050 Ti given the price point of the average aiv partner models the 1050 struggles a little bit but it's really not bad it just kind of depends on the workload of the particular scene that you're rendering and everything else 1440p is fine on 10 60s and four 80s and up and four K does pretty well on the GTX 1080 but again that's a gtx 1080 so you would hope so we're getting about 60fps there that's with extreme settings basically completely maxed out other than super sampling and we're hitting sixty so that's what you can expect for performance see maybe some improvements derived from the drivers as mentioned but we'll revisit those as necessary once the game approaches launch as always thank you for watching patreon link in the post roll video to help us out directly link in the description below for the full test methodology and for the article about this topic and to subscribe for more thanks for watching I'll see you all next time you know it's really annoying is games with a prepositional phrase at the beginning of the title so now I want to say for honor I have to say for 400