Gadgetory

welcome back to hardware unboxed all right guys you've been hounding me since late last week to benchmark the Windows 10 for creators update so that's what we're gonna do I spent the last few days benchmarking and getting all the data for this video but unfortunately most of it I won't actually bother showing you guys because it's very uninteresting ok so a quick recap on what's happened so far early this year Microsoft here's with what they called the creators update for Windows 10 and part of this update included a feature called game mode which was designed to ensure the operating system was focused on delivering the best possible performance for gamers and I did this by disabling non-critical processors to free up resources given that that's all it really did those using modern hardware game mode didn't really offer anything performance remained exactly the same Brian over at tech yes City did a nice little video on this feature and found that if you're running a dirty old potato gaming rig then under certain conditions the game mode did help smooth out frame rates and improve performance so technically it did work it just wasn't as amazing as someone hoped now Microsoft has recently just released a new update the full creator update which has brought about a few tweaks to the game mode that as far as I'm aware Microsoft hasn't really made any performance claims here but still as usual that hasn't stopped the internet from creating rumors and getting all excited and making a few bold claims even here in the harbor unbox comment sections I've seen a few users claiming massive performance gains you know upwards of 10 15 even 20 FPS more in games that said though there were just as many probably more of you that said it made no difference at all if I don't have time to test myself I generally look to other trusted sources to see what they found there have been a few reports so far from those that have tested the update the most controversial of which I'd have to say is a video by Joker Productions which showed massive truly massive gains on a core i7 semi 7rk system using a veil 56 graphics card at 1440p just one of the six games tested that was Batman Arkham Knight showed no performance uplift while games such as Rainbow six siege saw an insane 26% increase in frame rate I have to admit to me this didn't seem possible and if I came across the numbers like this I'd do a serious amount of digging it just seems too good or too implausible to be true having had this video spammed to me by all you guys wanting me to do a follow-up video I decided I would take a look while I was benchmarking and investigating reinstalling Windows updating windows on a number of different systems to try and work out what was going on LHD ville long-haired steve published an article titled windows 10 full craters update rising benchmark CPU bound and in that he showed no performance gains from the update frankly this is really what I was expecting a fine but we have one source now claiming a twenty percent on average gain and another claiming zero it has to be said that both source has really only tested a limited hardware configurations but this should really tell us all we need to know still with Joker testing a core i7 77 okay with Vega 56 in gamers Nexus testing horizon 7 1700 CPU with a GTX 1082 I decided to test a range of CPUs with Vega 64 liquid-cooled since joke I found big gains with Vega and I'd just done a heap of Vega 64 liquid-cooled testing I thought this is the way I'd go that said that I should just note that I did also do some testing with the GTX 1080i and found yeah much the same as what I'm about to show you before we get to the benchmark graphs let's just check some side-by-side gameplay footage using some canned benchmarks as they're generally a bit more accurate for this kind of thing please note the data in the graph shown later in the video are based on an average of three runs obviously the side-by-side stuff is just showing framerate data from a single run so there will be a little more variance there I'm also measuring the data using different software for the graphs the previous version of Windows 10 was tested with the game mode enabled which has no impact on the results disabled you'd get the same figures the following data for the full crater update is with the game mode off but I did also test with it enabled on a per game basis and I found any the reduced performance all the results were just the same first up we have Rainbow six siege and here the frame rates look much the same before and after the update these side-by-side comparisons have been done using the core i5 8600 k and vega 64 liquid core graphics card there's a little back and forth but overall the update was about 2 FPS faster on average from this individual run but i'll compare the 3 run average soon something I noticed apart from the fact that we were at best seen just a 1 to 2 percent performance improvement is the fact that CPU utilization was constantly lower after the update after around 10% lower so that's certainly very interesting and it could point to systems with weaker CPUs seeing some kind of gain not 20% gains but gains all the same next up we have Warhammer 2 and here the frame rates are virtually identical before and after the update I was using the DirectX 12 API with the high quality preset here honestly there wasn't even a frame in it but again the CP utilization was quite a bit lower after the update so again I hope those running older systems may see some sort of performance improvement for the most part frame rates remained much the same in the division and here I was again using the DirectX 12 API CPU utilization was slightly lower began performance with the 80 600k was pretty much the same with the full crater update finally the last side-by-side was recorded with Deus Ex mankind of Ireland here performance was unchanged from the previous Windows 10 build even the CP utilization was much the same here so not much to report on alright so time for some graphs and well there really isn't much to discuss here so I'm just gonna breeze through these really quickly like I said previously I did a lot more testing than the stuff I'm just showing you here but it's all very repetitive and boring Rainbow six siege at 720p is pretty much margin of error stuff same is true at IEP and finally at 1440p we again find a very similar story total war Warhammer two was much the same at 720p pretty much exactly the same at 1080p in exactly the same at 1440p so that was a rather uneventful again something interesting I found was that with the game mode enabled the full credit actually reduced performance in the games I tested his Rainbow six siege as an example here I saw up to a 7% reduction in performance with the game mode enabled I also tried close in the game after enabling the game mode reloading the game and then checking the performance over again and also confirming that game mode was turned on and yes every single time without fail the game was slightly slower with the game mode turned on so I'm not quite sure what's going on there well they have it if your system was running correctly with the previous Windows 10 build then after installing the new full crater update you should be looking at very similar performance it's possible that those running quad-core CPUs maybe the 7600 K for example you might see some improve performance there in CPU intensive titles such as battlefield 1 but you won't say anything like 20% improved performance at least I can't imagine you would I'd say somewhere around maybe up to 10% gains could be seen due to the reduced CPU usage as I just alluded to if you are seeing massive performance gains or even a really strong performance gains there is a chance there was probably something wrong we set up previously I wouldn't rule out all performance gains but you're not going to see up to 20% from this update unless a lot of other media outlets start reporting that but I don't think it's going to be the case so you can which though I am keen to see some other media outlets do this sort of testing to see what kind of results they dig up for now though at least I have confirmed the results gamers Nexus published and I did that with different Hardware not sure if they're going to do it part two where they also do very good testing that would be kind of cool if they do might ask Steve about that directly and finally it would be cool if Brian over at tech your city dug out a few of these old bangers and looked into seeing what kind of performance gains there might be there that something could be interesting you did do this last time so go over there harass him and you'll probably make it happen anyway that's gonna do it for this one I'm your host Steve see you guys again soon