Crysis Warhead perf

I'm having a bizarre problem - worse performance with SLI on than off.

Specs: 2x8800 GT, XP SP3, 3GB RAM, CD2 @ 3 GHZ.

With SLI off, the frames in FRAPS are generally 5-10 FPS lower, or about the same, but very smooth.

With SLI on, there's noticeable stuttering and pause as I try to pan or zoom in and out.

This is true with latest beta Nvidia drivers and WHQL.

Sidenote: I loaded up the original Crysis with the beta drivers on SLI and non-SLI. In this case, SLI clearly improved the FPS without any stuttering, but the game crashed the entire system within 5 minutes.

Hassles of SLI are just not worth it...
 
Yeah I agree.....judging about how well the new 4870's actually run....if so...with AA on to max what fps will you get then?

Don't know, but it would likely be unplayable judging by the performance of higher end cards running max settings.
 
Rig is in sig. CPU was clocked at 3.6 GHz at time of playing (its now up to 4.2 and going so I'm taking a bit of a pause from gaming ;) )

Using Cat 8.9 (BIG differnece from Cat 8.8 btw), I'm playing Warhead at 1920 x 1200 with all settings on Enthusiast + 4xAA and I can play every level smoothly, even the ice ones. Only time I've noticed slowing was in heavy smoke area (such as the purple smoke at the docks, and the initial scene) but beyond those, no problems whatsoever at all. Always 25+ fps, usually 30-40 fps no problem.

In fact, I can play 2560 x 1600 with no AA and almost all settings on enthusiats (3/4 of them or so) and its smooth. Definitely an improvement over Crysis.

Oh yeah, all DX10 also
 
Ok, here we go with the "Crysis is not optimized" garbage. This game runs like crap on most systems because the graphics are more complex than anything currently out there. End of story. This was the exact same complaint when Farcry came out, and that game has held up only because the graphics were so far ahead of the curve.

I've found Warhead runs marginally better than Crysis at the expense of some textures and detail: It is very small step backwards in terms of graphics, but the level design and fluidity of the storyline is much improved over the original. Also, the amount of enemies they throw at you is much higher in parts offering higher difficulty if you choose to fight. All in all, more enjoyable than the original and well worth the price.

I'm using DX10, Vista 64, Catalyst 8.9 at 1900x1200 with everything at Enthusiast except Post-processing, no AA. I'm getting a steady 40-45 FPS... and 4870 X2s are supposed to run like shit on this engine. I have noticed no bugs whatsoever.
 
I've been playing Crysis Warhead at near-Enthusiast 1920x1200 with an E8500 / GTX 280 and getting mostly in the 30-40 fps range. The game is smooth and just plain awesome.

I run the following settings:

DX9
Everything on "Gamer" EXCEPT:
Objects on Enthusiast
Physiscs on Enthusiast
Post Processing on Enthusiast

The following custom config:
Code:
con_restricted=0
r_UsePOM=0
r_sunshafts=1
e_water_ocean_fft=1
q_Renderer=3
r_colorgrading=1
r_useedgeaa=0
This is pretty close to Enthusiast shaders quality, with not much of a hit. I turned off the edge smoothing because I think it sucks. The game is kind of alias-y but it's still really fun, and it runs so fast, I can't complain.
 
For me Warhead runs better than Crysis, but after reinstalling Crysis again as well, I do have to say it runs smoother than I first thought, perhaps because of the drivers.


I run the game on 1280x1024 all on "Gamer" except textures "Mainstream" on a HD 4850. Strangely enough I get blurred textures in some places when setting textures on "Gamer" on my system. At the same settings, Warhead runs somewhat better than Crysis, but the latter game does well too.


I'm sorry to hear that Warhead is given performance problems for some people. On one side I can understand that the game really taxes the hardware heavy, but on the other hand I did read often in reviews that the game is very playable on "Enthusiast" settings. The difference in performance between "Gamer" and "Enthusiast" seems to be huge. The latter looks awesome, but the former is also beautiful and definitely worth the increase in performance and smoothness.
 
My card (9800GX2) runs fine on Enthusiast settings @ 1680x1050...

...until it overheats and framerates drop through the floor. I've seen heat-induced stuttering/performance-suckage on my roommate's laptop. His CoD4 got so bad that even at minimum settings he was getting a jerky slideshow. He bought a laptop cooler and now he's fine.

The fan is at maximum 24/7 through eVGA Precision.
 
Ok, here we go with the "Crysis is not optimized" garbage. This game runs like crap on most systems because the graphics are more complex than anything currently out there. End of story.

That's more or less true, although Crytek has always had a tendency to say things about their product that are totally false.

quads being better than dual, 64-bit being 15% faster per core than 32-bit, dx10-only features of Crysis, end-product graphics quality etc.

I've found Warhead runs marginally better than Crysis at the expense of some textures and detail: It is very small step backwards in terms of graphics,

The performance enhancements allowed for more overhead, which Crytek used to iterate their particle system, making it graphically richer and more complex as well as introducing properly completed Time-Of-Day settings which can also cause a Perf<->IQ trade-off.

In short: they see better performance as a chance to crank up the render quality rather than provide majorly faster framerates. The graphics improved in most areas, although the beach level huts look shockingly cheap for a graphical "sea change" game like this.

I'm using DX10, Vista 64, Catalyst 8.9 at 1900x1200 with everything at Enthusiast except Post-processing, no AA. I'm getting a steady 40-45 FPS... and 4870 X2s are supposed to run like shit on this engine. I have noticed no bugs whatsoever.

Those new drivers are putting out great numbers so far. As a recent GTX260 owner I'm hoping our "Big Bang II" drivers net a similar increase from my current 25 FPS in DX10 1680x1050 Enthusiast.
 
Dont you guys remember the DX 9 vs 10 thing H did a while ago? 10 is more memory efficent than 9, we knew this guys
 
Warhead ran terribly at first for me, vista x64 [email protected], 4870 and 4gb ram. Lag when turning around, fps dropping to unsatisfactory levels etc.

With a little http://www.blackviper.com/ tweakage, full defrag (drive was 44% fragmented o_O), killed some useless taskbar progs that had escaped my wrath far too long and installed 8.9 catalysts.

Changed the game entirely, higher minimum framerate and much more immersive. Well worth the 20 minutes work, plus the hour or so defrag.

To be fair the hdd thrashing was a major sign and most likely the real cause, but I was suprised the most basic tweaking really did make a difference. Even if the 100mb or so saved tweaking the services and startup programs is more psychological, than anything else :)
 
The original Crysis had some performance hack ins by altering some of the configuration game files....I wonder if Warhead allows the same.
 
The original Crysis had some performance hack ins by altering some of the configuration game files....I wonder if Warhead allows the same.

It wasn't a hack, it was just a custom config (using system.cfg or autoexec) - Warhead allows the same. Most of the settings are the same.
 
The real question is how the Warhead PC actually runs the game hahaha.

Anyways, any other similar setups to rampantandroid's? Let's see more benchmarks! I see people all over the place w/ the same video card even but no numbers :(
 
DX10 has been shown to be more efficient....but it seems that Crytek isn't using it right.

I guess my numbers are normal. I'm VERY curious about people playing this game at 19x12 on a GTX260 (as one reviewer claimed)...

I remember reading a good article on it, where basically it was summed up as: "performance increases are very possible with DX10, but naive or lazy ports are only going to hurt". Devs slap together a DX9 path, then DX10 gets tacked up with some chewing gum and string.

Crysis does a lot wrong in the performance department. There's good ideas, like the occlusion system for instance, but it's not done very well. When you set up a model, a simple hull of a sort is used as the occluder. Too bad this apparently wasn't done half the time, so there's often enormous overdraw that shouldn't be in lots of places. Instead of barely drawing anything while being behind something, I'd see that I would pushing 2.3 million triangles.
 
Tried one, no difference. Link me to one you know works, please.

Specs :

AMD Opteron 165 S939 @ 2.3GHZ ( It was a crummy overclocker :( )
AMD Radeon 4850 512MB
4GB Corsair XMS PC3200 DDR1 lol

Anywho, I have a custom config for regular Crysis that I modified for Warhead. IMO it looks just as good as "Enthusiast" and runs me at 1920x1080 @ Well beyond 35-40 90+% of the time.

Here is the config, just copy the following it into your "My Documents\My Games\Crysis_WARHEAD" directory as a file named "system.cfg" using notepad or what have you.This worked for me in Windows XP x64 , and should be the same in x86. In Vista the folder structure might be different.)

Another note is that you should set all advanced system options to low, before using this.

Code:
con_restricted = 0

r_VSync = 0
d3d9_TripleBuffering = 0
r_displayinfo = 0

g_useProfile = 1
e_particles_thread = 1
sys_budget_videomem = 1024
sys_budget_sysmem = 4096
sys_physics_CPU = 1

r_GeomInstancing = 1
r_MultiGPU = 1
r_ShadersAsyncCompiling = 1
e_vegetation_static_instancing = 1
e_hw_occlusion_culling_objects = 1
e_hw_occlusion_culling_water = 1
e_terrain_occlusion_culling = 1
es_OnDemandPhysics = 1

e_screenshot_width = 1920
e_screenshot_height = 1080
e_screenshot_quality = 100
e_screenshot_file_format = jpg

s_SpeakerConfig = 5

r_HDRlevel = 0.7
r_HDRBrightOffset = 3
r_HDRBrightThreshold = 2
r_HDRRendering = 1
r_EyeAdaptionClamp = 2.5

e_obj_quality = 2
e_proc_vegetation = 0
ca_useDecals = 0 
e_decals_allow_game_decals = 0
e_decals_life_time_scale = 0
e_lod_ratio = 6
e_lod_min = 2
e_view_dist_ratio_detail = 20
e_view_dist_ratio_vegetation = 40
e_vegetation_min_size = 2.0
i_rejecteffects = 0
e_vegetation_bending = 1
sys_flash_curve_tess_error = 8
e_view_dist_ratio = 50
e_max_view_dst_spec_lerp = 0.5
e_vegetation_sprites_distance_custom_ratio_min = 0.5
e_vegetation_sprites_distance_ratio = 1
e_detail_materials_view_dist_z = 64
es_DebrisLifetimeScale = 0.3
e_cbuffer_resolution = 64
e_dissolve = 1
ca_DrawFaceAttachments = 1
ca_AttachmentCullingRation = 50
e_terrain_occlusion_culling_max_dist = 100

i_lighteffects = 1
g_ragdollMinTime = 1.0
g_ragdollUnseenTime = 1
g_ragdollDistance = 3.0
g_battleDust_enable = 0

e_particles_quality = 1
e_particles_lod = 0.7
e_particles_max_emitter_draw_screen = 4
r_UseSoftParticles = 0
e_water_ocean_soft_particles = 0
e_particles_object_collisions = 0
i_particleeffects = 1

e_cull_veg_activation = 25
g_joint_breaking = 1
g_tree_cut_reuse_dist = 0.25
p_max_MC_iters = 2000
e_phys_foliage = 2
es_MaxPhysDist = 45
es_MaxPhysDistInvisible = 10
e_phys_ocean_cell = 0
e_foliage_wind_activation_dist = 5
g_breakage_particles_limit = 80
v_vehicle_quality = 1
p_max_substeps_large_group = 3
p_num_bodies_large_group = 30

r_PostProcessEffects = 1
r_MotionBlur = 1
r_Flares = 1
r_Coronas = 1
r_UseEdgeAA = 0
r_sunshafts = 1
r_colorgrading = 1
r_ColorGradingDOF = 0
r_GlowScreenMultiplier = 0.2
r_WaterGodRays = 1

q_ShaderGeneral = 3
q_ShaderMetal = 2
q_ShaderGlass = 0
q_ShaderVegetation = 2
q_ShaderIce = 0
q_ShaderTerrain = 3
q_ShaderShadow = 1
q_ShaderFX = 1
q_ShaderPostProcess = 2
q_ShaderHDR = 1
q_ShaderSky = 1
q_Renderer = 3
r_LightsSinglePass = 1

e_sky_type = 1
e_sky_update_rate = 0.1
r_DetailTextures = 1
r_DetailNumLayers = 1
r_DetailDistance = 4
r_HDRRendering = 1
r_SSAO = 1
r_SSAO_quality = 1
r_SSAO_blur = 2
r_SSAO_radius = 1
r_SSAO_darkening = 0.3
e_ram_maps = 1
sys_flash_edgeaa = 1
e_vegetation_use_terrain_color = 1
e_terrain_ao = 0
e_terrain_normal_map = 0
e_max_entity_lights = 8
r_UsePom = 1
r_EnvTexUpdateInterval = 0.01
r_TexturesFilteringQuality = 2
r_HairSortingQuality = 0
r_FillLights = 7
e_particles_lights = 0

s_FormatSampleRate = 44000
s_CacheSize = 80
s_MPEGDecoders = 32
s_Obstruction = 1
s_ObstructionAccuracy = 1
s_ObstructionUpdate = 0.1
s_SoundMoodsDSP = 1
s_VariationLimiter = 1.0
s_ReverbType = 2

sys_LowSpecPak = 0
r_TexSkyResolution = 2
r_ImposterRatio = 1.7
r_EnvCMResolution = 0
r_EnvTexResolution = 1
r_DynTexMaxSize = 50
r_TexAtlasSize = 2048
r_DynTexAtlasCloudsMaxSize = 12
r_DynTexAtlasSpritesMaxSize = 8

r_Beams = 3
r_BeamsDistFactor = 0.05
r_BeamsMaxSlices = 32
e_Clouds = 1
r_CloudsUpdateAlways = 0

r_WaterRefractions = 1
r_WaterReflections = 1
r_WaterUpdateFactor = 0.001
e_water_tesselation_amount = 10
e_water_tesselation_swath_width = 10
r_WaterUpdateDistance = 0.2
r_WaterCaustics = 1
r_WaterReflectionsQuality = 1
e_water_ocean_fft = 1
q_ShaderWater = 2
r_WaterReflectionsMinVisiblePixelsUpdate = 0.05

r_DepthOfField = 1
e_terrain_texture_lod_ratio = 1.5
e_terrain_lod_ratio = 1.5
e_detail_materials_view_dist_xy = 2048
e_terrain_normal_map = 0
g_radialBlur = 1
r_shadowblur = 3
r_ShadowJittering = 1
e_gsm_cache = 1
e_gsm_lods_num = 3
r_TexturesStreaming = 0

Issues :

* HDR levels are based on cubans, if you dont like them delete the HDR levels in the system.cfg.

* In rare instances there is a slight glow around the gun. It does not affect gameplay. If you know how to fix this please share.

If you want better FPS give it a try! Also if you know how to improve upon it further please share :)
 
Seems game is very cpu bound on my system, 2560x1600, no AA, everything on high, except textures which I have on very high.

Dual 4870x2's seem very cpu bound due to my Q6600 as I was able to keep turning up the settinsg from min to gamer(high) with no change in fps :) I haven't tried putting anything else to very high yet, and according to CPU-Z my GPU load avrages only around 58% about equally on all 4 core, I donno if that means anything in terms of headroom thou. And yes it stays steady even in the ice levels (just got inside the mine now, great game, and they actually made fighting the aliens fun, unlike the first Crysis! :)
 
My settings:

Textures: Mainstream
Objects: Gamer
Shadows: Gamer
Physics: Gamer
Shaders: Enthusiast
Volumetric: Enthusiast
Game FX: Enthusiast
Postprocessing: Enthusiast
Particles: Enthusiast
Water: Enthusiast
Sound: Gamer
Motion blur: Gamer

And my system: Vista x64, Athlon X2 @ 2.6GHz, 2GB DDR, 9800GTX overclocked by about 25%.

I run it at 1280x960, DX10, with antialiasing turned off, and vsync on. Really smooth, very playable and looks gorgeous save for some fugly textures. But with textures set to Gamer it stutters and hitches constantly, with pauses lasting several seconds whenever you turn.

I'm guessing minimal = 256MB, mainstream = 320-512MB, gamer = 640-768MB, enthusiast=896-1024MB

Or maybe I need more RAM to run it in Vista.
 
I was running a mix of mostly gamer and enthusiasst settings but in general not averaging above 30fps using a variety of 177.xx drivers. I just downloaded this performance config yesterday here:

http://www.pcgameshardware.com/aid,...clusive_Uber_quality_and_performance_configs/

(you can download it from the forum link about it without registering) and it really smokes! The visuals look great so far, gonna test it some more, but now my fps are in the 40-mid-50's and it looks really good;

I don't like the enthusiast shader settings, too much HDR> Gamer is about right but I like to enable the highest level of game fx, post processing etc. (the stuff on the right).

Try the performance config, the HQ config is very nice as well but wasn't as good for fps as my own mix of gamer/enthusiast settings.

edit: btw, I'm using driver ver. 177.84, seems to be the best ATM.
 
Specs :

Code:
con_restricted = 0

r_GeomInstancing = 1
r_MultiGPU = 1
r_ShadersAsyncCompiling = 1
e_vegetation_static_instancing = 1
e_hw_occlusion_culling_objects = 1
e_hw_occlusion_culling_water = 1
e_terrain_occlusion_culling = 1

Issues :

* HDR levels are based on cubans, if you dont like them delete the HDR levels in the system.cfg.

* In rare instances there is a slight glow around the gun. It does not affect gameplay. If you know how to fix this please share.

If you want better FPS give it a try! Also if you know how to improve upon it further please share :)

HW occlusion culling objects for me is actually SLOWER (which is somewhat surprising)

Be weary of some settings. Also the 'glow' is probably SSAO.
 
Gys just finished plying crysis not bad but if your PC isn't runnign with it I'd forget about one Blink and the games finish its miles to short, makes COD4 look like an epic!!!
 
I finished the game in less than two days, it's very short, looks outstanding, it's performance is better, but it's dependency on CPU is harder, made my experience a bit more difficult since it was harder to select the settings that would match my system. Doesn't matter if it was all in mainstream or enthusiast, the FPS wouldn''t change by much, it would run even slower with everything on mainstream than in gamer, in the snow level which is a part that many systems struggle, it ran faster than in the beach or forest level loll, that proves that it's CPU dependency is greater than Crysis, also my GPU usage is stuck at 70% average, and sometimes would dip to 40%, CPU bottleneck anyone? loll, considering that my chip is as fast as an Athlon X2 4400+ in gaming.

On Crysis with everything on medium was completely playable. Overall, was able to raise the eye candy considerably, the game played between 18 to 40 most of the time in DX10 without FSAA, at 1,024x768 with textures on gamer, physics on gamer, post processing on enthusiast, volumetric lightning on min (This last option pretty much is invisible), shaders on enthusiast, game effects and object effects on gamer. Game was pretty much as fast as Crysis on high using DX9 and some mods to enable some very high effects, not bad, beautiful and playable when running around, beautiful and barely playable slideshow when shooting at enemies or vehicles.
 
So, this is what Half Life episodic content has caused- Extremely short games for a "discount" $30. It's a sad day in gaming- Companies are going to start following this and make even more money off of gamers. :rolleyes:
 
The game is fun and entertaining, is like the other side of the story and is not repetitive and packs more action, but is short overall, a bit longer than Half Life 2 Episode series, but shorter than the original game, probably because there's less blah blah blah and more action which makes the game more fun and hence, shorter. Timeshift is long as hell because the game is boring, little blah blah blah, much repetitiveness and lack of innovation during the plot.
 
Okay, some basic questions for those that have the game, and hardware similar to mine (GTX 280 on Core 2):

Is the visual quality for Enthusiast equal to or greater than Very High in Crysis?

Is the visual quality for Gamer equal to or greater than High in Crysis?

If the above are true, does Warhead yield more performance given equal settings? In essence, is there a noticeable framerate boost if I play Crysis on High at 1920x1200, and then go play Warhead on Gamer at 1920x1200?

If yes, is this boost enough to increase playable settings? For example, if max playable is Crysis on High at 1920x1200, then in Warhead will you be able to jump that up to Enthusiast at 1920x1200?
 
The settings are not directly comparable. For instance post-processing can be kept to high or very high and is independent of motion blur, meaning you can keep light shafts but get rid of blur without messing around with custom configs.

In general, the game does not run faster, period. All that's happened is that Crytek has packed more action into the same frame rates, so technically the engine itself is faster, but not the game.
 
Like the poster on the previous page, I can also run the game at 1680x1050 on all enthusiast settings with no AA and get 30-40 fps constant with my q6600 @ 3.6, 4 gigs ram, and 9800gx2. However I suffer from the memory leak problem - that is, the game runs great until I complete the level and load the next one. As soon as it loads the game runs like shit - 0-20 fps max and only fixes itself after restarting the game, and sometimes, only after restarting the entire system.

I noticed that creating a 512 mb paging file instead of using none on vista x64 helped the memory leak problem.

But I'm wondering if my sometimes crappy performance is caused by heat issues like the previous poster said. I know the GX2 runs incredibly hot, I usually see it in the 80's or maybe even 90's during crysis play since I leave the EVGA precision of auto fan control. Do you think these high temps could cause my fps drops??
 
I thought Windows was supposed to be run with the pagefile just because some programs (and Windows itself) uses it from time to time?
 
I thought Windows was supposed to be run with the pagefile just because some programs (and Windows itself) uses it from time to time?

Windows doesnt need it unless it runs out of physical memory.
If you run a program that needs it then leave it turned on.
 
Yeah, it's pretty short. Finished it in 7.3 hours according to STEAM.
 
I finished the game in less than two days, it's very short, looks outstanding, it's performance is better, but it's dependency on CPU is harder, made my experience a bit more difficult since it was harder to select the settings that would match my system. Doesn't matter if it was all in mainstream or enthusiast, the FPS wouldn''t change by much, it would run even slower with everything on mainstream than in gamer, in the snow level which is a part that many systems struggle, it ran faster than in the beach or forest level loll, that proves that it's CPU dependency is greater than Crysis, also my GPU usage is stuck at 70% average, and sometimes would dip to 40%, CPU bottleneck anyone? loll, considering that my chip is as fast as an Athlon X2 4400+ in gaming.

On Crysis with everything on medium was completely playable. Overall, was able to raise the eye candy considerably, the game played between 18 to 40 most of the time in DX10 without FSAA, at 1,024x768 with textures on gamer, physics on gamer, post processing on enthusiast, volumetric lightning on min (This last option pretty much is invisible), shaders on enthusiast, game effects and object effects on gamer. Game was pretty much as fast as Crysis on high using DX9 and some mods to enable some very high effects, not bad, beautiful and playable when running around, beautiful and barely playable slideshow when shooting at enemies or vehicles.

Same here, except I am on a q6600 at 3.2GHz, and 2x 4870x2's and at 2560x1600
 
Okay, some basic questions for those that have the game, and hardware similar to mine (GTX 280 on Core 2):

Is the visual quality for Enthusiast equal to or greater than Very High in Crysis?

Is the visual quality for Gamer equal to or greater than High in Crysis?

If the above are true, does Warhead yield more performance given equal settings? In essence, is there a noticeable framerate boost if I play Crysis on High at 1920x1200, and then go play Warhead on Gamer at 1920x1200?

If yes, is this boost enough to increase playable settings? For example, if max playable is Crysis on High at 1920x1200, then in Warhead will you be able to jump that up to Enthusiast at 1920x1200?

i) Greater
ii) Greater
iii & iv) My average FPS in Crysis @ 1920 with an extreme quality config (everything maxed + crazy draw distances) is ~30 with combat dips down to 20 with heavy particles + action

My average FPS in Warhead @ 1920 set to enthusiast (with r_UseEdgeAA = 2 and e_shadows_water = 1) is about the same, but it doesn't break 30 as often. The triangle and particle load is much heavier in Warhead.

edit: oops, I had 16xAF and 4xMSTRAA on. the latter shouldn't have made any impact since it wouldn't have worked, but the 16xAF might have done.
 
Funny thing that I installed Crysis yesterday with the latest patch and I'm running the game at 1,024 on Dx10 mode with textures on high, physics on high, post processing on very high, volumetric on low since this option doesn't do a thing in graphics when sunshafts is on, shaders on very high, game effects on high and object effects on high. It ran much faster than running Warhead at the same settings and I was able to use 4x FSAA with almost no penalty at performance, like 1 or 2fps less, that proves that I'm more CPU bound in this game than GPU bound, but I don't know why Warhead ran considerably slower than this game. Warhead looks better overall but not by much.

But also I found an issue which would cause sound stuttering in Vista when single core CPU's are used, it's because when the FMod EX sound engine scan for audio devices and initialize, it uses the HDMI found on the HD 3X00 series of card instead of the Sound Card and it sounds horrible and lower the FPS considerably since it maxes out the CPU and the CPU cannot cope with the load from both sides, audio and graphics. My solution was to disable the HDMI device in the Sound Options in Control Panel and everything worked fine again, in Crysis the sound was almost horrible and imposible to hear. Since I found the fix after I finished Warhead, in Warhead the sound stuttering issue wasn't that bad, but sometimes the game ran for 3 seconds without any sound, so I'll use it the next time I play Warhead.
 
Why aren't you people running the WH bench tool and then you will know what you got....

On avalanche at 1680x1050 at enthusiast I get 36 fps average.....

no cfgs......
 
I have a 4870x2 with the official 8.9's , I'm reading about people getting 30+ fps on Enthusiast setting but I can't seem to get that level of performance

I have a [email protected] with 4gb ram on Vista 64.

Here are my results:


DirectX 10 ENTHUSIAST 3X @ Map: ambush @ 0 1920 x 1200 AA 0xx
==> Framerate [ Min: 16.93 Max: 36.60 Avg: 26.87 ]

DirectX 10 ENTHUSIAST 3X @ Map: avalanche @ 0 1920 x 1200 AA 0xx
==> Framerate [ Min: 7.34 Max: 21.11 Avg: 15.48 ]

DirectX 10 ENTHUSIAST 3X @ Map: frost @ 0 1920 x 1200 AA 0xx
==> Framerate [ Min: 9.97 Max: 15.38 Avg: 13.00 ]

DirectX 10 ENTHUSIAST 3X @ Map: ambush @ 0 1920 x 1200 AA 4xx
==> Framerate [ Min: 17.06 Max: 36.64 Avg: 26.73 ]

DirectX 10 ENTHUSIAST 3X @ Map: avalanche @ 0 1920 x 1200 AA 4xx
==> Framerate [ Min: 6.81 Max: 20.99 Avg: 15.48 ]

DirectX 10 ENTHUSIAST 3X @ Map: frost @ 0 1920 x 1200 AA 4xx
==> Framerate [ Min: 9.61 Max: 15.54 Avg: 13.10 ]



Does this seem low to you? Could I be bottlenecked by CPU? (even though I don't see it maxing out all 4 cores in terms of load)

Any advice?
 
19x12 + Enthusiast + 4xAA is asking too much even from a 4870x2 IMO. Turning off AA (which doesn't seem to do much in this game anyway) should net you a good jump. As far as the CPU is concerned, it seems to my Crysis only use 2 cores. I've never see my CPU usage go past 40% load.
 
Hey sorry to change topic on this thread....but is it worth buying this crysis sequel game?

I mean you guys have completed it, I never went past the start of the original Crysis because it got a bit frustrating for me to compensate the lag....is this one worth it?
 
Mine runs like crap now. There's a memory leak (that's my opinion at least) that shows up with the later, more complex levels (such as the mine). Thus, even on lowered settings, the game "pauses" every so often and the driver may time out, causing the "Display driver has stopped responding and has recovered" if you're lucky or "Crysis has stopped working" if you're not.
 
Back
Top