• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD's next big Mantle Game

I dont understand why people compare star swarm results. Its not a benchmark. Its a test for their graphic engine. Each time you get different results. AI affects the route of the ship. Sometimes a laser can make your ship spin all the time and throw you out of the fight. Just compare the batches on each run. Different. You need a custom scenario for consistent results.

http://www.sendspace.com/file/eok37u

Download this(its a custom scenario) and drop it to star swarm/assets inside the steam folder. Its way more consistent than the follow
 
pgi947 said:
A good choice for mantle this, iirc CIV5 is crazy heavy on the CPU.

A shame its a game series I've never had the patience to get into :(
Click to expand...

Yeah it is pretty heavy game wise anyway.

You should give it a go. Civ 5 was a bit pants compared to Civ 4 on release but with the two expansions they've now got a game which is thoroughly enjoyable and has a worthwhile amount of depth.

I still think it could be a little more strategised but that's just me. :p
 
Sugarhell said:
I dont understand why people compare star swarm results. Its not a benchmark. Its a test for their graphic engine. Each time you get different results. AI affects the route of the ship. Sometimes a laser can make your ship spin all the time and throw you out of the fight. Just compare the batches on each run. Different. You need a custom scenario for consistent results.

http://www.sendspace.com/file/eok37u

Download this(its a custom scenario) and drop it to star swarm/assets inside the steam folder. Its way more consistent than the follow
Click to expand...

Nice one Sugar. Will try this later.
 
Didn't know about Mantle relationship, interesting. I won't be building another PC until Star Citizen is out though so I can truly get the latest and greatest tech to (hopefully) run 4k games well at whatever cost,
 
Noxia said:
Didn't know about Mantle relationship, interesting. I won't be building another PC until Star Citizen is out though so I can truly get the latest and greatest tech to (hopefully) run 4k games well at whatever cost,
Click to expand...

Mantle will work on Star Citizen as well you'll be pleased to know.
 
Another article to add to the OP if you wish Humbug.

AMD has scored another win for their Mantle API, with upcoming title Sid Meier's Civilization: Beyond Earth to support it alongside DirectX 11. Previous Civilization titles, such as Civilization V, were reasonably CPU intensive, so the reduced CPU overhead and lower-level features of Mantle should help gamers run the game on entry-level hardware including AMD's own APUs.
Click to expand...

Source
http://www.techspot.com/news/56376-...-is-sid-meiers-civilization-beyond-earth.html
 
humbug said:
They are sabotaging Nvidia?

Its just to make the CPU (including those 12 thread £450 i7's) work more for it, the bench is designed to test CPU performance, they obviously felt it wasn't working CPU's like the 3960K hard enough, (if my guess is right)

There is no conspiracy.
Click to expand...

Because when I test and see no single core of a 4930k using more than 60% and a Gpu using 40%, my first reaction is "how can I add more bottleneck to this to make everything worse"
Sugarhell said:
I dont understand why people compare star swarm results. Its not a benchmark. Its a test for their graphic engine. Each time you get different results. AI affects the route of the ship. Sometimes a laser can make your ship spin all the time and throw you out of the fight. Just compare the batches on each run. Different. You need a custom scenario for consistent results.

http://www.sendspace.com/file/eok37u

Download this(its a custom scenario) and drop it to star swarm/assets inside the steam folder. Its way more consistent than the follow
Click to expand...

We tried that in the Starswarm bench thread, but people got uppity when it didnt show mantle outright winning and demanded we change it back to the standard demo
 
Last edited:
andybird123 said:
Because when I test and see no single core of a 4930k using more than 60% and a Gpu using 40%, my first reaction is "how can I add more bottleneck to this to make everything worse"
Click to expand...

Well your 40% GPU is because it is a bottleneck

I get the same (No one thread working anywhere near max) when running anything which bottlenecks, it does not mean there is a conspiracy. Or am i to believe World of Tanks has it in for my rig?
 
humbug said:
Well your 40% GPU is because it is a bottleneck

I get the same (No one thread working anywhere near max) when running anything which bottlenecks, it does not mean there is a conspiracy. Or am i to believe World of Tanks has it in for my rig?
Click to expand...

I've just tested Star Citizen, Thief, BF3, PS2 and Crysis 3

when I force a situation where I'm alway <70% GPU, at least one core is constant >80%

with star swarm, when I have 32fps, I have GPU around 30% and CPU <40% on any one core and most cores at 20%, so it seems to have no trouble creating threads and using them (the biggest criticism of DX)
this is why I struggle with star swarm as an indication of anything, it certainly isn't CPU or GPU bottlenecked, and "because DX" doesn't actually cut it with me when all these other games show a clear CPU bottleneck on at least one core

I'm downloading WoT now
edit: tested WoT, bit tricky as it caps itself to 125fps, which'll be why you can get low usage with low CPU, but when I found some settings that got around 100fps, I had 99% GPU and constant 95-99% CPU
with WoT, the bottleneck is artificial and provably so (it caps at 125fps), so what else does star swarm have in common?
 
Last edited:
andybird123 said:
I've just tested Star Citizen, Thief, BF3, PS2 and Crysis 3

when I force a situation where I'm alway <70% GPU, at least one core is constant >80%

with star swarm, when I have 32fps, I have GPU around 30% and CPU <40% on any one core and most cores at 20%, so it seems to have no trouble creating threads and using them (the biggest criticism of DX)
this is why I struggle with star swarm as an indication of anything, it certainly isn't CPU or GPU bottlenecked, and "because DX" doesn't actually cut it with me when all these other games show a clear CPU bottleneck on at least one core

I'm downloading WoT now
Click to expand...

In starswarm the API itself limiting, can't cope with 20-30000 draw calls.
Also if you want ro compare results the best is the RTS view where the camera standing still. Not 100% consistent but much better than follow mode.
 
andybird123 said:
I've just tested Star Citizen, Thief, BF3, PS2 and Crysis 3

when I force a situation where I'm alway <70% GPU, at least one core is constant >80%

with star swarm, when I have 32fps, I have GPU around 30% and CPU <40% on any one core and most cores at 20%, so it seems to have no trouble creating threads and using them (the biggest criticism of DX)
this is why I struggle with star swarm as an indication of anything, it certainly isn't CPU or GPU bottlenecked, and "because DX" doesn't actually cut it with me when all these other games show a clear CPU bottleneck on at least one core

I'm downloading WoT now
Click to expand...

Hmmm. We have many bottlenecks. Dx use 2 threads for rendering and another one for the driver thread. Experience developers offload audio,AI,physics etc etc to the other cores. So you can have a bottleneck when the 2 rendering threads except the rendering they do other stuffs too and you get 90+% usage or you can get a driver thread bottleneck (Its all the work that dx11 does). For example AC black flag is the first bottleneck and CiV,Total war is the second one. Usage means nothing. For example frostbite is an extremely multithread engine still on bf4 mp you are limited from your cpu(dx11). Driver thread API no matter the optimizations you can fix it. Thats why RTS games have a limit on the units and a limit on the camera zoom out. On sc2 after 30 mins match even with an intel cpu i am dropping down to 30 fps with 40-50% cpu usage.
 
Last edited:
Mtom said:
In starswarm the API itself limiting, can't cope with 20-30000 draw calls.
Also if you want ro compare results the best is the RTS view where the camera standing still. Not 100% consistent but much better than follow mode.
Click to expand...

yes, but the reason why is that DX has lots of CPU overhead, meaning one core maxes itself out and can't process more "draw" calls
 
Rusty0611 said:
Yeah it is pretty heavy game wise anyway.

You should give it a go. Civ 5 was a bit pants compared to Civ 4 on release but with the two expansions they've now got a game which is thoroughly enjoyable and has a worthwhile amount of depth.

I still think it could be a little more strategised but that's just me. :p
Click to expand...

One day, maybe :p

The worst part is, at some point I've managed to buy all of the CIV games AND the DLC, scouring through it all steam tells me out of it all I've played 17mins on CIVIV thats it :(
 
You must log in or register to reply here.
Back
Top Bottom