Ashes of the Singularity benchmarks.

Dan_D said:
Is there a way to just download the benchmark alone? All I see on the website was crap for buying the actual game.
Click to expand...

Oh you just need to contact them and explain to them that you're a member of the press. I PM'd you a couple of weeks ago with the information I believe. ;) I'll look up the information again.
 
I probably still have that PM. Not meaning to ignore anyone, just been busy as of late. Working two jobs and all of that. One of which is bringing you guys motherboard reviews. :)
 
jwcalla said:
It's amazing that even that CPU is bottlenecking the game so much. Looks like they need to go back to the drawing board on this one.
Click to expand...

Oh these FX-9000 series processors are just old. It's amazing that they are still scaling on a motherboard from June 2011. I mentioned this earlier in the thread when asked, but my R9 290 only runs at 8x speeds due to a bug that was in the first batch of these Sabertooth 990FX mobos. Asus patched in the subsequent Sabertooth 990FX 2.0 and 3.0, but neglected the 1.0 models.

From what I read if I were to trust Asus RMA, I can get a replacement that runs at 16x speeds. The catch is that Asus RMA is terrible and I don't want my motherboard in RMA hell for months. Maybe it's better now. I know what I have right now, and a bird in hand is worth more than two in the bush. If you catch what I mean. ;)
 
jwcalla said:
It's amazing that even that CPU is bottlenecking the game so much. Looks like they need to go back to the drawing board on this one.
Click to expand...

Here is a 5.0GHz on High settings with Temporal AA Quality set to Mid. Usually I leave it off. Notice the GPU load compared to the other tests.

CwMoMWW.png
 
cageymaru said:
Having too much fun with this. Figured out that a setting called Temporal AA Quality doesn't seem to change in the benchmark when you choose Low, High, Crazy, etc. So turning it off yields the faster results that look like everyone else's on the net. I wonder what the default for it is?

Here is a fun run just for you. 800MHz FX-9370. Do you remember what Brad Wardell stated was the lowest speed that the FX-8350 could run at and yield favorable results? Guess I'll try 2.4GHz next as I think that's what my old Q6600 ran at. I doubt I'll actually play the game until 3am.
Click to expand...

Haha, I remember a video from Oxide with an AMD FX8350 CPU underclocked to 2ghz running Mantle, not sure about 800mhz but I can imagine they'd try it just for kicks to confirm the huge amount of CPU time wasted under DX11.

Hmm, Temporal AA is originally an old AA technique ATI brought out in ... +/-2004 linky here. It was subsequently taken out of the Catalyst Options years back, might still be in the driver code though. nvidia re-hashed the name with their Txaa. In any case the name looks like it is being re-used again with DX12. Curious if the option has all the coding behind it finalized though.

All fun to see this in action!
 
Yakk said:
Haha, I remember a video from Oxide with an AMD FX8350 CPU underclocked to 2ghz running Mantle, not sure about 800mhz but I can imagine they'd try it just for kicks to confirm the huge amount of CPU time wasted under DX11.

Hmm, Temporal AA is originally an old AA technique ATI brought out in ... +/-2004 linky here. It was subsequently taken out of the Catalyst Options years back, might still be in the driver code though. nvidia re-hased the name with their Txaa. In any case the name looks like it is being re-used again with DX12. Curious if the option has all the coding behind it finalized though.

All fun to see this in action!
Click to expand...

it was also used in watch dogs with a great implementation.. low to null performance hit and was the sweet spot in AA quality in that game..
 
5.0Ghz Crazy setting. It enables Temporal AA Quality and sets it to High. The WCCFTECH test had Glare Quality set to off. The benchmark can load up the GPU quite nicely in my opinion. I still have MSAA left off. Anyway I'm dead tired now. ;)


oQSyzPI.png
 
So the real questions is: Should we buy this game, if we never played the old supreme commander games? In all honesty the most complicated rts games i ever played was Sins of a solar empire and CnC series. I loved both of those and am wondering if i will be able to pick up on the learning curve enough with this game to get hooked into learning it. Were you excited to wake up and play the game or is it just a glorified dx12 benchmark tool? edit and one last thing, can you post a direct comparison of your benchmark results vs. a comparative nvidia system? without that i have no idea how good or bad your results are.
 
We're not allowed to show game footage because the game is in a Pre Beta state of development still. I just woke up at 9am after doing stuff until 6am last night. After I get my mom situated, and make her some lunch, I'll fire up the game and see how it is.
 
I bought the game and I am downloading it now so I hope to see if X58 is still fast enough for DX 12.
 
primetime said:
So the real questions is: Should we buy this game, if we never played the old supreme commander games? In all honesty the most complicated rts games i ever played was Sins of a solar empire and CnC series. I loved both of those and am wondering if i will be able to pick up on the learning curve enough with this game to get hooked into learning it. Were you excited to wake up and play the game or is it just a glorified dx12 benchmark tool? edit and one last thing, can you post a direct comparison of your benchmark results vs. a comparative nvidia system? without that i have no idea how good or bad your results are.
Click to expand...

Oxide has a scenario where you're thrust into battle and only have one base and very limited resources. The A.I. controls the all of the rest of the map, but their forces are very limited at each node. There seems to be a way to win by resources as the I lost the first game in a few minutes to the A.I. this way. So you have to quickly mobilize, take over as many resources as you can from them before they get real defenses, and then crush the A.I.

I'm grinning from ear to ear. This game is damn fun. I have never seen so many points of contention with myriads of attackers. I built some Logistics Arrays to increase my army to a paltry 400 units. I have literally just started playing so no idea what the unit cap is. Not going to read the forums as discovering what to do is a lot more fun than reading a strategy guide.

Need to contact my nephew and make him get the game. He would love this!


I need to add that the game is Pre Beta. I can't emphasize enough that there will be bugs and you're job is to find them. If you see balance issues report them. If your game crashes report it. Helping to develop the game is the main purpose of a Pre beta. Being able to play the game is the icing on the cake. Alpha developer builds are not for the faint of heart. Might want to wait for a proper release.
 
fanboy said:
I bought the game and I am downloading it now so I hope to see if X58 is still fast enough for DX 12.
Click to expand...

If I can run it, I'm sure that your x58 can also. If you flip through the last pages you see that the game scales very well from a 800MHz PC to a 5.0GHz PC.
 
Ok.. this would be on hardware 5+ plus years old and limited to 6Gb of ram and a 500Mb WD -Black hard drive with 32Mb buffer ..

X5660%20-290X%201.jpg
 
@fanboy

What is the default for Temporal AA Quality? Is it supposed to be off or on in the default settings? It makes a huge difference for me as far as GPU frame rates go.
 
Can you guys do screenshots of windows 10 task manager for cpu usage under heavy usage. Just curious just how well this game uses multiple cpu cores.If you alt tab just as the benchmark ends and grab a snip it might reveal interesting results.
 
I can't get the benchmark to run in full screen .. also I don't know about the AA ..
 
primetime said:
Can you guys do screenshots of windows 10 task manager for cpu usage under heavy usage. Just curious just how well this game uses multiple cpu cores.If you alt tab just as the benchmark ends and grab a snip it might reveal interesting results.
Click to expand...

When I run the benchmark in fullscreen, it will crash the game if I tab out. I think I can get you that info though. :)
 
Hey you wan to try out the latest drivers and see if there is any difference? Wonder if they were for DX 11 or DX 12.
 
After looking at more data .. I think the avg fps is useless without knowing the batch count itself as one video may offer more batch's .. like say Heavy fps 29.5 with 35789 batch's as different cpu's may affect this as lower avg does not mean slower if the batch count was higher.

So run a normal bench run on high and compare the one I will post to see if the batch's have a set limit or unlimited in the benchmark.
 
fanboy said:
After looking at more data .. I think the avg fps is useless without knowing the batch count itself as one video may offer more batch's .. like say Heavy fps 29.5 with 35789 batch's as different cpu's may affect this as lower avg does not mean slower if the batch count was higher.

So run a normal bench run on high and compare the one I will post to see if the batch's have a set limit or unlimited in the benchmark.
Click to expand...

Ok! I will leave the Temporal AA off as I think that is the default.
 
This what I mean .. find the same results and compare the batch count to see if it is limited..

bacths.png
 
I see what you mean. Look at my Pastebin upload and hit CTRL F. Then search for "Low Vista". You see that I got these results.

== Shot panoramic battle shot =========================================
Total Time: 9.988016 ms per frame
Avg Framerate: 37.044395 FPS (26.994638 ms)
Weighted Framerate: 36.869244 FPS (27.122879 ms)
CPU frame rate (estimated if not GPU bound): 42.211391 FPS (23.690287 ms)
Percent GPU Bound: 93.460701 %
Driver throughput (Batches per ms): 3561.177734 Batches
Average Batches per frame: 21715.246094 Batches

== Shot mountain air battle shot 2 =========================================
Total Time: 5.007316 ms per frame
Avg Framerate: 46.731621 FPS (21.398787 ms)
Weighted Framerate: 46.468781 FPS (21.519825 ms)
CPU frame rate (estimated if not GPU bound): 48.628727 FPS (20.563976 ms)
Percent GPU Bound: 27.105888 %
Driver throughput (Batches per ms): 3938.419678 Batches
Average Batches per frame: 15059.551758 Batches

== Shot low vista =========================================
Total Time: 10.005378 ms per frame
Avg Framerate: 33.481995 FPS (29.866800 ms)
Weighted Framerate: 33.381996 FPS (29.956266 ms)
CPU frame rate (estimated if not GPU bound): 39.526688 FPS (25.299362 ms)
Percent GPU Bound: 96.079399 %
Driver throughput (Batches per ms): 2948.953613 Batches
Average Batches per frame: 19750.308594 Batches
 
I was trying to understand the benchmarks that everyone was talking about 290x vs 980Ti but they don't breakdown the batch count the gpu renders at said FPS for a given scene ..

So just because one scored 50 fps and the other scored 48 fps does not mean 48 fps was slower if it rendered more batch's in the same scene as the demand was greater.
 
Capture1_zpsvofni6h7.jpg


My old card and cpu seems to do pretty well. Game definitely makes use of hexa cores.;)
 
primetime said:
Capture1_zpsvofni6h7.jpg


My old card and cpu seems to do pretty well. Game definitely makes use of hexa cores.;)
Click to expand...

man that's a great performance for that 280X... run it full screen and you will see even more performance yet..
 
Think my old hexa core xeon doesn't hurt one bit.:) Finally a game that uses all my cpu cores
Capture2_zpssh8eekda.jpg

dx12 is looking good so far:)
 
lol prime..wtf you have it cranked up to 4.8Ghz ?

but your score for your avg fps in both cpu and gpu needs the batch counts pre scene to compare with others as to see what your X5670/280x has rendered on a given scene..

I am thinking that even the cpu's avg fps is effected by the batch count rendered as the higher the batch count the greater the demand placed on the system to render.
 
You must log in or register to reply here.
Back
Top