'Assassin’s Creed: Syndicate' Benchmarks: A Stressing Experience

[heirdeux]

"Assassin's Creed: Syndicate" is finally out on the PC, and if you want the best frame rate for your system, you'll need a high-end system and also spend some time in the game's graphical settings.

'Assassin’s Creed: Syndicate' Benchmarks: A Stressing Experience : Read more

 

[FITCamaro]

This is terrible testing. The only common denominator in the tables is the resolution. You have multiple configurations and settings all in one table which makes the results misleading and hard to read.

Tom's if this is what you have for reviews today you have problems.

 

[namtrooper81]

I skimmed the article, that is some poor performance from a 980TI. No idea however if the graphics quality warrants such low fps.

 

[cknobman]

I skimmed the article, that is some poor performance from a 980TI. No idea however if the graphics quality warrants such low fps.

Easy explanation: Its a sh!t port done by UbiSoft.

 

[Daniel Ladishew]

While it may be of value to readers to know what settings the game automatically applied based on each systems specs, as a comparison it makes for very difficult reading. Putting all that into a graph and calling it a 'benchmark' is a stretch. I applaud that your using on hand hardware to test new games (gives readers a more realistic set of examples), but find the testing procedure itself lacking for any kind of meaningful comparisons on the hardware itself. You can do better.

 

[Kwuarter]

Can the text be any more blurry? I'm sorry but it's kind of silly that I have to click on the graph in order to see it properly.

 

[tstng]

Pretty awful review, there is no consistency, you are testing different cards with different cpu's at different quality settings. It's all a mishmash, not to mention really small and blurry images.

Also I do have to mention that currently (even with the latest 359.00 drivers) Nvidia's HBAO+ Ultra and PCSS Ultra are murder or any system and MFAA is incompatible with the game (you are forced to turn it off in ncp otherwise you get missing textures and artifacts in the game), so both the drivers and the game optimization are shoddy at best. (Game Ready Driver my posterior) On my 970/i5-2500k I can play 50-60 fps if I use High Shadows (not PCSS) and only HBAO+ (not the Ultra one).

 

[spentshells]

System 1 should be at least a fury if not fury x for AMD

 

[jimmysmitty]

Pretty awful review, there is no consistency, you are testing different cards with different cpu's at different quality settings. It's all a mishmash, not to mention really small and blurry images.

Also I do have to mention that currently (even with the latest 359.00 drivers) Nvidia's HBAO+ Ultra and PCSS Ultra are murder or any system and MFAA is incompatible with the game (you are forced to turn it off in ncp otherwise you get missing textures and artifacts in the game), so both the drivers and the game optimization are shoddy at best. (Game Ready Driver my posterior) On my 970/i5-2500k I can play 50-60 fps if I use High Shadows (not PCSS) and only HBAO+ (not the Ultra one).

I agree the image quality is pretty low but this is not a review. It is a sample benchmark. While I wish they would do a normal full review this is being looked at the wrong way. Think of this as the equivalent of a couple of people on forums poting their specs, settngs and results. It may not give a full idea but it gives somewhat of an idea.

To me this shows that the game is very poorly optimized. I don't see how a top end i7 and 980Ti should have any issues maxing the game out at 1080P. And I don;t think it is the HBAO or PCSS as I have run games maxed at 1080P with maxed out PCSS/CCSS and HBAO+. Considering their track record, Ubisoft has had a lot of performance issues with their game releases.

 

[hannibal]

This is not a graphic card comparison, it tells what setting you can use with your hardware in this game.
And also that this game is seriously power hungry game as it has been said above.

 

[schiwing]

This is terrible testing.

Hopefully this doesn't become the new trend for testing.

 

[GPUEnthusiast]

Sorry, I don't want to jump on the bandwagom, because I think this a solid review with a bunch of great setups. Next time though, I would suggest using all the GPUs with a single CPU, removing the laptop from the graph, but keep the A8.

Secondly, I would say you should stick to a single graphics preset for every setup at every resolution. So let's say "High" preset at 720p, 1080p, and 2160p. I like your graphs in that they show min, max, and average. And I enjoy the variety of video cards from different generations.

 

[jimmysmitty]

Sorry, I don't want to jump on the bandwagom, because I think this a solid review with a bunch of great setups. Next time though, I would suggest using all the GPUs with a single CPU, removing the laptop from the graph, but keep the A8.

Secondly, I would say you should stick to a single graphics preset for every setup at every resolution. So let's say "High" preset at 720p, 1080p, and 2160p. I like your graphs in that they show min, max, and average. And I enjoy the variety of video cards from different generations.

As I said, people are not looking at this the way it should. It is not a normal review. These are not systems all next to each other, these are the new teams personal systems. And while people might think they all live and work in the same area, they all do not. Some contributors like in other parts of the US and even the world.

That's why it is setup the way it is.

If it was a normal performance review it would be like their normal reviews with set hardware and tests.

 

[JasonSmasHeR]

I'm going to buy it during the Black Friday sale. Will see how it works for me.

 

[Jonathan_21]

Pretty awful review, there is no consistency, you are testing different cards with different cpu's at different quality settings. It's all a mishmash, not to mention really small and blurry images.

Also I do have to mention that currently (even with the latest 359.00 drivers) Nvidia's HBAO+ Ultra and PCSS Ultra are murder or any system

I can vouch for this, this game super ultra setting murdered my setup, GTX 980 ti with 4670k @ 4.5 ghz. I don't think it's not optimized tho, best looking game i ever saw on my pc. I play @ 1440p.

 

[MagusALL]

I am not an expert on what constitutes a 'benchmark' but I think this was a fair approach giving the reader various framerates with completely different rigs. If, I assume, it was a true benchmark everyone would expect an Intel i7-5960k with 16Gb of RAM and a GPU that is current. However many people want to play this game on THEIR PC's and they may not have this same benchmarking rig to the one that is standard. With that being said I highly doubt most gamers are using Intel's top of the line 8-core CPU. In fact I think it has been shown more than once that anything more than a four core Intel CPU yields little benefit in gaming fps. Personally I own this game, along with most of the new AAA games to come out for PC/console. My rig consists of a i7-3770k OC'd to 4.2Ghz, 16Gb of RAM at XMP 1600, and two GTX 780's with 3Gb of VRAM and I game with mostly maxed settings although I did drop a few things to get the performance where I thought it looked great on my 1200p Samsung SyncMaster T260HD. I assume most PC gamers know that games ported from consoles usually do not give solid 60fps and up especially from some developers. However this article or review or benchmark or whatever you call it gave a range (a wide one at that) of the performance of this game at different resolutions considering their respective PC specs. So it is what it is. One thing that surprised me was when I was in the graphics settings it showed the VRAM required and VRAM available at the top left corner. I have two MSI GTX 780's with 3Gb each but it showed me as having the equivalent of having 6Gb of VRAM which I found confusing because as we have been told over and over VRAM does not stack in SLI. I also currently own two MSI GTX 970's that I planned on SLI'ing but my rig just would not show a display when the GPU was changed out (one at a time.) So I can add nothing to how my rig would perform with this particular game with those two in SLI. I will be selling my GTX 970's as the 780's work fine for me and I expected the 970's, in SLI to perform better but am unable to even try seeing the results. So I guess it goes like this; try the game out on your own rig and see what happens. You may need to go below Ultra as I did but the game looks great in my eyes. Numbers are just numbers and I turned on V-Sync. My personal plan is to wait until a 34" curved 1440p 21:9 monitor comes out with G-Sync which to me gives me enough power to play the games at max and allows my physical eyes to not notice any framerate issues. To each his own, but at least give TomsHardware a little credit as I am sure it took a pretty long time to not only test out each rig with this game but to then also write a review for you to read and gain information. Its not like we pay for TomsHardware to exist so complaining about a free website seems to me a bit greedy and ungrateful. Thank you for your article and showing how this game is not well optimized unless you have a high-spec PC and want the best settings applied. Anything short of what you want and what you have is up to you. If you can afford a benchmarking rig I am sure you can SLI two GTX Titan X's and h2o cool everything, overclock those and the top-of-the-line CPU and I am positive you can get your 60fps at Ultra settings. Or you can go out and buy a PS4 or XBox One and play it that way. This is just information. Toms is not saying expect something when their various testing rigs gave those subsequent results. They are giving you an impression of how those systems will run and you have to adapt them to your own rig and expectations. Give them a break already. That or post your own benchmarks. Either way stop blaming the messenger. Toms neither made the game nor the CPU's nor GPU's, they merely gave you plenty of information on various rigs. So as my ol' Drill Instructor in the Marines would say, "suck it up and deal with it, you either lead, follow or get the hell out of the way." If you don't like the article than don't read it. Give it up already. Thanks Toms for your time and efforts, but you can't make everyone happy all the time, no matter what rig they are sitting at complaining on. Cry to Ubisoft instead.

 

[Anders235]

Pretty awful review, there is no consistency, you are testing different cards with different cpu's at different quality settings. It's all a mishmash, not to mention really small and blurry images.

Also I do have to mention that currently (even with the latest 359.00 drivers) Nvidia's HBAO+ Ultra and PCSS Ultra are murder or any system

I can vouch for this, this game super ultra setting murdered my setup, GTX 980 ti with 4670k @ 4.5 ghz. I don't think it's not optimized tho, best looking game i ever saw on my pc. I play @ 1440p.

I still suspect its terribly optimized. Ass Unity was for sure. Ubisoft has made it clear for a while now that polishing PC game performance is not something they're interested in doing.

 

[hunterjsn90]

"Pretty awful review, there is no consistency, you are testing different cards with different cpu's at different quality settings. It's all a mishmash."

I agree, these benchmarks do not make any sense whatsoever

the gtx 980 and r9 390x are competitors in their relative price range. But the benchmark goes and does this

GTX 980 w/ intel i3 4160 vs the 390x w/Intel Core i7-5930k

it does not make any sense at all

 

[turkey3_scratch]

Looking at this, I should be able to get lowmedium settings on my 390 on 1440p. The game it probably one of the most visually appealing which is why it is so intense. It looks better than what Unity started out as last year.

I don't know about you guys, but most people who bash AC games as being unoptimized probably never played them and saw how gorgeous they really are.

I don't know why you have two resolutions listed for one graph though. It says 1080p/1200p?

 

[fudoka711]

This is terrible testing. The only common denominator in the tables is the resolution. You have multiple configurations and settings all in one table which makes the results misleading and hard to read.

Tom's if this is what you have for reviews today you have problems.

Look, I don't know if it was added in after this comment or was already in there (which means you didn't read), but the second paragraph clearly states that this is only a sampling of configurations and is not to be taken as a well-done review.

Why is it bad, in a quickie review, to have configurations bundled together like they are? They show performance at 720p, 1080p, and 4k. I assume they skip 1440p for lack of time and because the 1440p/4k market overall is still extremely small and in the interest of time, they could only do one or the other. They also probably wanted to show people the highest settings they could achieve at a decent framerate (40+ avg fps at 1080p), hence combining low/high settings in one graph

Also, in a proper review, they would use a single high end processor along with standardized other components and just compare the graphics cards (or just compare cpu's), but this isn't a proper review and they admitted to it before showing you the results. Doesn't make it "okay", but also doesn't make it okay for you to bash the review for not being comprehensive when it was never intended to be.

I'm sure a comprehensive review will come out in due time. For now, at least you have some information about how you can expect your personal system to perform just by looking at how much better/worse your cpu/gpu are compared to the tested configurations.

What is very clear about this preview is that this is a very demanding game.

Looking at this, I should be able to get lowmedium settings on my 390 on 1440p. The game it probably one of the most visually appealing which is why it is so intense. It looks better than what Unity started out as last year.

I don't know about you guys, but most people who bash AC games as being unoptimized probably never played them and saw how gorgeous they really are.

I don't know why you have two resolutions listed for one graph though. It says 1080p/1200p?

1080p/1200p are essentially the same resolution with a difference of 230k pixels (or ~10%). 1440p/1600p are in the same boat with a roughly 10% difference in pixels.

And yea, AC games are notorious for being horribly optimized for PC and this preview shows it.

 

[ts1506]

On my system (i5 4440, 16gb ddr3, gtx960), I hit anywhere between 25 to 45fps depending on area on 1080p, Ultra High Env, High Textures, PCSS, HBAO+ Ultra and FXAA.

However, bump up the shadows to PCSS Ultra, and the fps drops to under 5 fps. Totally murders the frame rate.
Syndicate is better at launch than Unity was, but as of today, Unity both looks better and runs better. Hopefully will be fixed by updates

 

[turkey3_scratch]

@TS: if you have 2Gb VRAM, the shadows could cause it to go past the 2GB.

 

[stevenrix]

I played the game in 2560x1980 with a GTX 980 TI on a skylake 6700K with 16 gigz of DDR-4 and the game works flawlessly with ultra settings. On the other end the black ops 3 looks like shit with ultra settings and drops frames like crazy.

 

[Georgivs]

I tried to play this game on GTX 670 with maxsetings, w/o AA, FHD reso. In the first mission while I was under the roof of factory it was almost good (30±5 fps like on a console), but when character came out the walls on open range the framerate dropped below 3 fps. Strong lack of memory obviously. When I set lower shadows(not PCSS) and lights (SSAO) then it became acceptable (around 30). On a cutscenes with face zooming the framerate drops observed, but not so critically. If we compare CPU-depending with Unity, this game looks better, but I think it is related with decrising of npcs on the streets. IMHO.
PS Sorry for my grammatical correctness. Have a nice day.

 

[Alex Atkin UK]

I tried to play this game on GTX 670 with maxsetings, w/o AA, FHD reso. In the first mission while I was under the roof of factory it was almost good (30±5 fps like on a console), but when character came out the walls on open range the framerate dropped below 3 fps. Strong lack of memory obviously. When I set lower shadows(not PCSS) and lights (SSAO) then it became acceptable (around 30). On a cutscenes with face zooming the framerate drops observed, but not so critically. If we compare CPU-depending with Unity, this game looks better, but I think it is related with decrising of npcs on the streets. IMHO.
PS Sorry for my grammatical correctness. Have a nice day.

They didn't just decrease NPCs from what I have read, they got rid of the AI entirely. It becomes quite obvious the longer you play and is a real shame as it kinda kills the immersion.