what is the command for defaults?

-default in launch options

Athlon 64 3200+ @ 2.41ghz 2gb ram 7900 GTX 512mb

1440*900 -default dx98 fov 90

2179 frames 113.608 seconds 19.18 fps (52.14 ms/f) 0.888 fps variability

Moved from 3570K to 6600K and was interested to see clock for clock difference. I also added in a max oc ram score as well. Very pleased with improved fps as far as the 6600K goes. A bit surprised that the increased ram clock added little to no fps. Same timings were used at both speeds too. Oh well. Maybe Valve will optimize once competitive rollout completes?

EDIT: Added 2133MHz @ CL11 results - added ~10-15 fps for each test! Thanks Setsul :)

3570K @ 4.5 GHz
16GB DDR3 @1600MHz CL11 (4x4)

6600K @ 4.5 GHz
16GB DDR4 @3866MHz CL19 (2x8)

6600K @ 4.5 GHz
16GB DDR4 @2133MHz CL15 (2x8)

6600K @ 4.5 GHz
16GB DDR4 @2133MHz CL11 (2x8)

Static settings
Driver version: 364.51
Resolution: 1920x1080 @ 120Hz full screen

Additional notes: W10 64-bit / Aero on

dxlevel: 95
FPS configs enabled: none / max settings
Shadows enabled/disabled: enabled

dxlevel: 81
FPS configs enabled: none / max settings
Shadows enabled/disabled: enabled

dxlevel: 95
FPS configs enabled: none / min settings
Shadows enabled/disabled: disabled

dxlevel: 81
FPS configs enabled: none / min settings
Shadows enabled/disabled: disabled

dxlevel: 95
FPS configs enabled: Comanglia's HQ PC
Shadows enabled/disabled: enabled

dxlevel: 95
FPS configs enabled: Comanglia's HQ PC
Shadows enabled/disabled: disabled

dxlevel: 81
FPS configs enabled: Comanglia's HQ PC
Shadows enabled/disabled: disabled

DDR4 RAM timings? They affected fps significantly in my testing. I'm also pretty sure you're just seeing diminishing returns, going from 1333 to 2400MHz made quite a difference. 2133MHz RAM just isn't really holding the CPU back in the first place.

I used the stock settings of 19-21-21-41 2t, which are trash even by DDR4 standards. The sticks are rated at 4000mhz, which was overkill in retrospect. Now that I've seen speed isn't really the factor, and after reviewing your post, I will try to tighten the timings and report back with the results. I think my DDR3s from my previous build were at 11-11-11-18, IIRC.

EDIT: Tightest stable timings at 2133MHz for the DDR4 was 11-11-11-28 1T. This yielded an additional 10-15 frames! I later realized that my original timings were 2133 CL15 and 3866 CL19. I'm guessing this yielded about the same true latency in the end, hence the lack of improved performance. Now my goal is to find the overall best clock/timing. I'm thinking 3200 is going to be the sweetspot.

CPU: i5-6600k @ 4.6 GHz
Graphics Card: Nvidia GTX 660
8GB DRR4 @ 2400Mhz
Dxlevel: 95
Resolution: 1920x1080 / 1280x960
FPS configs enabled: Edited Comanglia stability config
Shadows enabled/disabled: Enabled
Additional notes: Windows 10 Aero

@1080p
2639 frames 15.506 seconds 170.19 fps ( 5.88 ms/f) 10.571 fps variability
@1280x960
2639 frames 14.792 seconds 178.40 fps ( 5.61 ms/f) 10.737 fps variability

#309
The latency wasn't really the same.
3866MHz: 19 = 9.83ns, 21 = 10.86ns
2133MHz: 11 = 10.31ns, 15 = 14.06ns

So the latency of 2133MHz CL11 is actually about the same as 3866MHz CL19.
The exception is the command rate.
Ah yes, TF2 doesn't disappoint. Not only do RAM timings have way more of an impact than the should on the low end, the one timing that usually matters the least has the most impact on the high end.

2639 frames 32.817 seconds 80.41 fps (12.44 ms/f) 5.479 fps variability

CPU and overclock: AMD FX-6300 @ Stock 3.5Ghz
Graphics Card: MSI Radeon R7 250 (GPU @1150Mhz, Memory @ 1000 Mhz)

Driver version: 15.30.1025.1001
dxlevel (default is 90): 90
Resolution: 1920x1080
Full-screen or windowed: Full
FPS configs enabled: None
Shadows enabled/disabled: Enabled
Additional notes: All high settings on an IMO shitty rig, why is this half decent ;-;

2639 frames 19.174 seconds 137.63 fps ( 7.27 ms/f) 11.027 fps variability

People still doing this? Okay, here goes.

CPU and overclock: i7-5820K @ 4.5GHz
Graphics Card: GeForce GTX 760 x2 SLI

Driver version: 364.51
dxlevel: 90
Resolution: 1920x1080
Full-screen or windowed: Windowed (Noborder)
FPS configs enabled: Max everything [From in the menu]
Shadows enabled/disabled: Enabled

2639 frames 32.817 seconds 80.41 fps (12.44 ms/f) 5.479 fps variability

Could you overclock the CPU to 4GHz+ and then set dxlevel to 95 and redo the benchmark, I am quite interested to know the result.

CPU: AMD FX-8350 @ 4.2GHz
GPU: Gigabyte 650 2GB OC @ core:1245MHz / mem:2980MHz

Driver: 361.43
dxlevel: 9.5
Resolution: 2560x1440
Full-screen or windowed: Full
FPS configs enabled: comanglia
Shadows enabled/disabled: disabled

Additional notes: mat_picmip 1, mat_specular 1

2639 frames 74.172 seconds 35.58 fps (28.11 ms/f) 2.694 fps variability

CPU: Intel(R) Core(TM) i5-4200U CPU @ 1.60GHz 2.30 GHz
Graphic Card: ACPI x64

Driver: 6.3.9600.16384
dxlevel 9
Resolution 1600x900
Windowed or fullscreen: Full
FPS configs enabled: Raphsody's performance config
Shadows enabled/disabled: Disabled

2639 frames 32.817 seconds 80.41 fps (12.44 ms/f) 5.479 fps variability

Using the stock cooler, can't get a stable OC over like 3.7Ghz

2639 frames 15.680 seconds 168.30 fps ( 5.94 ms/f) 10.289 fps variability

CPU and overclock:Intel Xeon E5-2670 v3 ES 12 core @ 2.6ghz HT disabled
Graphics Card: Gigabyte Radeon HD 7970 GHz Edition (GPU @1200Mhz, Memory @1500 Mhz)
Ram: 16gb DDR4 2400
OS: Windows 10 Enterprise 64 bit
Driver version: Crimson 16.3.1
dxlevel: 98
Resolution: 1920x1080 144hz
Full-screen or windowed: Full
FPS configs enabled: Comanglia' frames config
Shadows enabled/disabled: disabled

I still have no idea what to do with the benchmark thing, I click on it opens sfm help please

type in console 'timedemo benchmark1'

Can anyone shed some light on the CAS latency situation?
Is it being weird in as much as CAS Latency > Access time/Actual latency?
Am I going to lose potential performance at 3200 CL14? My access time should be nutty (8.75ns).

(On a side note, the demo tracks the engineer now when I sware it used to be a static camera)

2639 frames 19.552 seconds 134.97 fps ( 7.41 ms/f) 9.570 fps variability

CPU: Intel Core i7-4790k @ 4.4ghz
GPU: NVIDIA GeForce GTX 970
RAM: 16GB DDR3
OS: Windows 7 Pro 64 bit
Driver version: 359.00
Dxlevel: 98
Res: 1920x1080 fullscreen
Fullscreen
FPS configs: No ragdolls or gibs, every graphical setting on max
Shadows enabled

Additional notes: -high in launch options

#321
I don't think I understand the question, what do you mean with "CAS Latency > Access time/Actual latency"?

Yes, the demo used to be static, thank Valve for breaking demos.

2639 frames 84.678 seconds 31.16 fps (32.09 ms/f) 6.301 fps variability

CPU: 2.7 GHz Intel Core i5
Graphics Card: AMD Radeon HD 6770M 512 MB
¸¸¸¸¸
opengl
Resolution: 1920x1080
Full-screen or windowed: windowed
FPS configs enabled: none
Shadows enabled/disabled: enabled

Models: Medium
Textures: Medium
Shaders: Low
Shadows: Low

is this good or bad?

2639 frames 8.348 seconds 316.11 fps ( 3.16 ms/f) 21.147 fps variability

CPU: Intel Core i7-6700k @ 4.8ghz
GPU: R9 390
RAM: DDR4 CL14 3200mhz 16GB
OS: Windows 10
Driver version: 359.00
Dxlevel: 81
Res: 1920x1080 Fullscreen
FPS configs: Comanglia, few texture setting changes, surfaceproperties

I doubt anyone will believe this tbh, I barely do. I defy anyone to beat this score without mods/cheats :^)

Jesus, I take it that OC is on liquid right?

Nope, Cryorig R1 Universal Air Cooler, 70c on x264 benchmark. AIOs are generally about the same as high end air cooling but noisier and with more inherent risks and No I haven't de-lided my CPU either incase you're wondering.

-snip-

2639 frames 7.924 seconds 333.04 fps ( 3.00 ms/f) 24.029 fps variability
+ No Hats

You won't get 4.9 without a dangerously high/unsustainable voltage and/or a top 3% chip.

-threads 8 does nothing

one must presume you have something else in your favour because my ram (CL13) and OC (4.8Ghz) are somewhat faster.

Atleast someone else with a 6700k is here now lol

2639 frames 7.638 seconds 345.52 fps ( 2.89 ms/f) 22.865 fps variability

turns out NZXT CAM was stealing FPS, overclocked ram to 3333mhz

2639 frames 7.546 seconds 349.72 fps ( 2.86 ms/f) 22.958 fps variability

+4 fps closing discord/skype

-snip-

2639 frames 7.440 seconds 354.73 fps ( 2.82 ms/f) 23.106 fps variability
+1% from setting all drivers to MSI mode

My PC already defaults to 0.5ms on TF2 apparently.

I'll set the bar at 355 fps then lol

First of all I would like to thank Setsul for his contributions to the PC Build thread. My budget build would not have been nearly as cost effective without his advice. It's still a bit imbalanced but that's my fault as a novice.

Hardware:
ASRock Z170 Pro4S
i3-6100 Skylake
Integrated Graphics
Team Dark Pro Red DDR4-3200
Samsung 850 EVO 250 Gb

Config:
Windows 8.1 (no updates), high performance power mode, best performance visual setting
Comanglia Stable FPS (dx 81)
No explosion script
No cosmetics/effects script
No bullet dust (surfaceproperties.txt)

First test with dual-channel 2133, the memory controllers default (not OC)
2639 frames 16.571 seconds 159.26 fps ( 6.28 ms/f) 9.034 fps variability

The next test at dual-channel 2800. I fucked up when buying and didn't check that 3200 is only supported with a single module. I set it to 3200 initially but speccy returns 2133 so it must have been dialing it back. Setting to ~3066 (not sure) would BSoD, setting to 2933 would boot but tf2 would not start. So 2800 is the highest it would go (in dual channel).
2639 frames 15.335 seconds 172.09 fps ( 5.81 ms/f) 10.381 fps variability

That's not a terribly significant difference, 8.5% gain,but it still matters. I have yet to test the difference in Dwarf Fortress. I think it will be very substantial in that game.

2639 frames 15.604 seconds 169.13 fps ( 5.91 ms/f) 10.312 fps variability

CPU and overclock:Intel Xeon E5-2670 v3 ES 12 core @ 2.6ghz HT disabled
Graphics Card: ASUS GTX 780 Ti DirectCU II
Ram:16gb DDR4 2400 dual chennel
OS: Windows 10 Enterprise 64 bit
Driver version: 364.72
dxlevel: 98
Resolution: 1920x1080 144hz
Full-screen or windowed: Full
FPS configs enabled: Comanglia' frames config medium models verry high textures
Shadows enabled/disabled: disabled