Post Your Framerate Benchmark Here

[Cilraaz]

Benchmark Index:
My i5 750 @ 3.8GHz vs RS25 (in this post)
Brainni's Q9550 @ 3.4GHz vs LK25HC
My i5 750 @ 3.8GHz vs ICC25 (first half)
My i5 750 @ 2.66GHz vs RS25 (stock test)
My i5 750 @ 3.8GHz vs ICC25 (first half)
None's Phenom II x4 955 @ 3.4GHz vs flight timetest
vesseblah's Phenom II x3 720 BE @ 2.8GHz vs Toravon
None's Phenom II x4 955 @ 4.0GHz vs flight timetest
vesseblah's Phenom II x3 720 BE @ 2.8GHz vs timetest, plus beta!

-----------------------------------------------------------------------

There are a lot of people asking "how would this setup do in WoW?". I thought to help alleviate some of these questions, we could have some people with different system specs post framerate benchmarks. Personally, I use FRAPS for the benchmark framerate capture, but any program that will keep track of framerate over a period of time will do. For instance, FRAPS keeps track of framerate in one second increments, and it then has the option to output framerate for each second interval or a simple output with min fps, max fps, avg fps, and total number of frames.

With your benchmark, please include the following:

• System Specs including:
  
  • CPU Brand/Model/Clock Speed
  • GPU Brand/Model/Clock Speed
  • RAM Brand/Model/Type (DDR2/DDR3)/Amount/Speed/CAS Latency
  • Anything else you'd like to include
• WoW Settings including:
  
  • Resolution
  • Color Depth
  • Multisampling Rate (AA)
  • Graphic Settings (effects tab)
  • V-Sync Setting
  • Triple Buffering Setting (if V-Sync is enabled)
  • Full Screen/Windowed Setting
  • Extra Effect Settings (Full Screen Glow, Specular Lighting, Projected Textures, etc)
• Benchmark Setup including:
  
  • Zone or Instance (including number of players)
  • Full Instance, Single Boss, or Only Some Bosses (boss name(s) included, if not a full instance)
  • Extenuating Circumstances (spent the fight staring at the wall/floor, etc)
• Benchmark Results including:
  
  • Minimum Framerate
  • Maximum Framerate
  • Average Framerate (fractions rounded down)

Here is mine from last night:

• System Specs:
  
  • Intel Core i5 750 @ 3.8GHz (200MHz base clock, 19x multiplier)
  • ATi Radeon HD 5770 @ stock speeds (850MHz core, 1200MHz VRAM)
  • G.Skill DDR3 4GB CL9 @ 1600MHz (200MHz base clock, 2:8 multiplier)
• WoW Settings:
  
  • 1920x1080
  • 24-bit Color
  • 8x Multisampling (Edge Detect forced from drivers, 16x AF forced in drivers)
  • Ultra, minus half shadows
  • V-Sync Disabled
  • N/A
  • Maximized Windowed Mode
  • All extra effects enabled
• Benchmark Setup:
  
  • Ruby Sanctum - 25-man
  • Full Instance
  • No extenuating circumstances
• Benchmark Results:
  
  • Min: 29 fps
  • Max: 109 fps
  • Avg: 51 fps

This test consisted of 2840 usable timeslices, or 47 minutes and 20 seconds.

I'm hoping to benchmark a full ICC25 next week. I'm also hoping to benchmark Ruby Sanctum again with my CPU at stock speeds to show what kind of improvements overclocking can give. I would do it for ICC also, but I'm not sure I can handle it for that long.

[edit]
Temperatures, as requested by Pretorian:

• Ambient: 24*C
• CPU:
  
  • Idle: 35/30/34/30*C (SpeedStep disabled)
  • Max Temp During WoW: 54/53/53/50*C
  • Stress Test: 68/67/64/64*C (OCCT's Linpack Test)
• GPU (default fan settings)
  
  • Idle: 41*C (set to not auto-downclock)
  • Max Temp During WoW: Untested
  • Stress Test: FurMark, 8xAA - 85*C (1280x1024, Xtreme Burning Mode), 82*C (1280x1024, normal mode), 81*C (1920x1080, Xtreme Burning Mode)

Hmm, I guess my TIM finally cured. That's about 6*C cooler on the CPU than the last stress test I ran a month or two ago.

Since WoW hardly stresses my GPU, I'd probably guess it sits somewhere around 60-65*C. That's what it was last time I checked the GPU temp while playing, and it fits in well with the idle and stress results. Maybe I'll remember to check next Tuesday during ICC.

To be thorough, I re-ran the FurMark test at 1920x1080 with Xtreme Burning Mode enabled. My GPU peaked at 81*C and refused to go higher. Over all three tests, the maximum GPU fan speed was 60% of its capability.

[Pretorian]

Cilraaz, please post your temperatures also, if you're running on air.

A little suggestion if I may: enable V-Sync.

I'm running an i7 920 @4.0GHz (191x21, daily use) and a GTX260 (default speed). If I play WoW with V-Sync Disabled, the GPU temp goes as high as 84 degrees (Celsius) and I have over 100 fps on average. If I play with V-Sync Enabled, the GPU temp goes up to 75 degrees (max) and I get to play with 60 fps (obviously). This is in 1920x1200 (24") with Ultra settings minus half shadows, like you.
My point is, V-Sync Disabled is pretty pointless on good PCs as it only "stresses" the GPU more than it's actually needed. The difference between 60 fps and 150fps is not visible to the human eye.

[Cilraaz]

Umm, actually anything over 60 fps is useless because it's not displayed by my monitor, not because my eye can't detect it. I understand V-Sync. I also understand that benchmarking with V-Sync enabled (unless suffering tearing otherwise) is pretty dumb.

I have no heat issues with either my CPU or GPU. The only time temps would really come into play during a benchmark would be if you're getting a lot lower performance than expected. But, for the record, I will edit my original post with my temps. Luckily I rebooted to clear some lag last night, so I actually have max temps during WoW in Core Temp.

[Pretorian]

Umm, actually anything over 60 fps is useless because it's not displayed by my monitor, not because my eye can't detect it.

Well, that's not exactly true... if you had a 120Hz monitor...

I also understand that benchmarking with V-Sync enabled (unless suffering tearing otherwise) is pretty dumb.

I have never thought of this, it's actually funny. Don't know what would be the point of that though.

Very nice temps for 3.8GHz. What cooler do you have on the CPU ?

[Cilraaz]

Well, that's not exactly true... if you had a 120Hz monitor...

Then I'd have more money than brains.

I have never thought of this, it's actually funny. Don't know what would be the point of that though.

The entire point of a benchmark is to see what kind of performance the system gives. Artificially limiting the system's performance invalidates the test.

Very nice temps for 3.8GHz. What cooler do you have on the CPU ?

I'm using a Xigmatek Dark Knight (using the fan that came with it) and Arctic Silver 5 TIM.

[pansertjald]

Cilraaz, please post your temperatures also, if you're running on air.

A little suggestion if I may: enable V-Sync.

I'm running an i7 920 @4.0GHz (191x21, daily use) and a GTX260 (default speed). If I play WoW with V-Sync Disabled, the GPU temp goes as high as 84 degrees (Celsius) and I have over 100 fps on average. If I play with V-Sync Enabled, the GPU temp goes up to 75 degrees (max) and I get to play with 60 fps (obviously). This is in 1920x1200 (24") with Ultra settings minus half shadows, like you.
My point is, V-Sync Disabled is pretty pointless on good PCs as it only "stresses" the GPU more than it's actually needed. The difference between 60 fps and 150fps is not visible to the human eye.

holy shit thats hot for a i7 920..... if i was you i would either check my cooler and see if it is mounted right or i would get a new one.
im never above 55 degrese in wow with my 4,0 Ghz spec in raids. every thing over 70 is WAY to hot.

and as you can see cilraaz is allso sitting on 50-55 degrese in wow with he`s OC

[Pretorian]

holy shit thats hot for a i7 920.....

GPU mate, not CPU.
CPU goes up to 65*C. But yes, I agree, the temps could be lower but the tower is in a relatively closed spot and I can't be bothered to move it as long as it runs flawlessly.

[pansertjald]

GPU mate, not CPU.
CPU goes up to 65*C. But yes, I agree, the temps could be lower but the tower is in a relatively closed spot and I can't be bothered to move it as long as it runs flawlessly.

arh sorry miss read

[Erik765]

Cilraaz, with those temps why are you only running your multiplier at 19?? Are your temps really too high if you bump up to 20? Always wondered every time I've seen your specs running an i5 at 3.8

[Cilraaz]

Cilraaz, with those temps why are you only running your multiplier at 19?? Are your temps really too high if you bump up to 20? Always wondered every time I've seen your specs running an i5 at 3.8

Odd multipliers are typically more stable. I've debated bumping the base clock up a bit more, but then I'd have to lower my RAM clocks. I'll have to see if I can get lucky and unlock 21x in the latest BIOS update.

[edit: Have it set at 191MHz at a 21x multiplier for an even 4.0GHz. So far, I've had to bump my voltages up too much to really be worth the extra 200MHz. I'm seeing temps 5*C hotter than 3.8GHz and it won't pass the OCCT Linpack test (no catastrophic failures, but it's still failing). I think I'm just going to revert to 3.8GHz and be happy with it.]

[edit 2: Went back to 3.8GHz. Temps dropped back down and stability returned. I'll have to reapply my TIM to see if I can steal back a few degrees. If I can, maybe I can pump my voltages up for some stability without having my temps grow too far.]

[shaddrak]

come on people post, i wanna see these different gpus and cpus at work here

[Deleted]

System Speccs:
Core2Quad Q9550@3.4GHz
8GB DDR2-1066 CL5
GF 260 // 216 @ Default Clock
OS: Win7 Ultimate 64bit
OS is running from a 80GB Intel G2 SSD, WoW from 2TB WD Caviar Black

WoW Settings:
1920x1200
All Sliders Max, Shadows disabled
Specular Lightning/Death Effect/Full-Screen Glow Effect Disabled
Projected Textures Enabled

Benchmark Setup:
Multiple LK25HC Wipes, about 25 minutes -> Includes infight and recovery times

Benchmark Results:
0/181/50 Min/Max/Avg

Note on the Min value: Im not sure what caused this, the FPS value was recorded in 1 second intervals, and they where 2x 0. If I remove those 2 values from a set of 1540 values, the min is 16.


//E

This shows what i allways think when people brag about there enormous framerates: >60fps while resting, 20-30fps infight.

[Cilraaz]

System Speccs:
Core2Quad Q9550@3.4GHz
8GB DDR2-1066 CL5
GF 260 // 216 @ Default Clock
OS: Win7 Ultimate 64bit
OS is running from a 80GB Intel G2 SSD, WoW from 2TB WD Caviar Black

WoW Settings:
1920x1200
All Sliders Max, Shadows disabled
Specular Lightning/Death Effect/Full-Screen Glow Effect Disabled
Projected Textures Enabled

Benchmark Setup:
Multiple LK25HC Wipes, about 25 minutes -> Includes infight and recovery times

Benchmark Results:
0/181/50 Min/Max/Avg

Note on the Min value: Im not sure what caused this, the FPS value was recorded in 1 second intervals, and they where 2x 0. If I remove those 2 values from a set of 1540 values, the min is 16.


//E
<img snip>

This shows what i allways think when people brag about there enormous framerates: >60fps while resting, 20-30fps infight.

Ooo, now I want to graph mine. Good thing I kept the raw output.

I believe the zero fps moments were from zoning back into the instance. I notice a momentary drop to zero fps which immediately goes back to normal (or above normal) fps when the loading screen appears.

Also, I fully agree with you. There are a lot of people who don't take into account minimum or average fps when talking about their system. Too many are focused on the highest number when that's actually the least important.

[Cilraaz]

Here's tonight's ICC25 (first half). I trimmed any outliers, both high and low. Essentially, if it was out of the expected range for less than 3 seconds, I assumed it was an outlier due to zoning. This included 1 second bursts up to 400fps+ and 1 or 2 second dips to 0-2fps.

• System Specs:
  
  • Intel Core i5 750 @ 3.8GHz (200MHz base clock, 19x multiplier)
  • ATi Radeon HD 5770 @ stock speeds (850MHz core, 1200MHz VRAM)
  • G.Skill DDR3 4GB CL9 @ 1600MHz (200MHz base clock, 2:8 multiplier)
• WoW Settings:
  
  • 1920x1080
  • 24-bit Color
  • 8x Multisampling (Edge Detect forced from drivers, 16x AF forced in drivers)
  • Ultra, minus half shadows
  • V-Sync Disabled
  • N/A
  • Maximized Windowed Mode
  • All extra effects enabled
• Benchmark Setup:
  
  • ICC - 25-man
  • Partial Instance - Marrowgar through Saurfang, Plague Wing, and Dreamwalker.
  • No extenuating circumstances
• Benchmark Results:
  
  • Min: 21 fps
  • Max: 111 fps
  • Avg: 57 fps

The hardest hit spot, as you can pretty much tell from the image, was Marrowgar's room. The night consisted of 7300 usable slices, or 2 hours, 1 minute, and 40 seconds.

[Deleted]

first half means you did not do sindragosa on that night ? im mostly interested in the last trash pull before her, you know the 2 frostwarden handlers that come with the massive group of dragon whelps. its a short fight, but it seems to be the most intense one (until the whelps are dead)
also if you can record that pull, can you also do an addon cpu usage profiling on that one ?

[Cilraaz]

first half means you did not do sindragosa on that night ? im mostly interested in the last trash pull before her, you know the 2 frostwarden handlers that come with the massive group of dragon whelps. its a short fight, but it seems to be the most intense one (until the whelps are dead)
also if you can record that pull, can you also do an addon cpu usage profiling on that one ?

We did not do Sindragosa. That will be Thursday. Last week I tried eyeing my framerate during the whelp fight and the lowest I saw was 18fps, but I might have missed a lower dip.

I will be skipping CPU profiling, as it destroys framerate (or at least makes it jittery).

[Deleted]

well yes, doing both fps recording and addon profiling would make the fps records useless

and if you do another graph, try reducing the width of the graph line, makes it a bit more readable

[Cilraaz]

well yes, doing both fps recording and addon profiling would make the fps records useless

and if you do another graph, try reducing the width of the graph line, makes it a bit more readable

I changed the line width to minimum (0.01") and blew up the full-size image to 1600px wide. That should help with the readability over a large amount of timeslices.

[ThrashMetalFtw]

The hardest hit spot, as you can pretty much tell from the image, was Marrowgar's room. The night consisted of 7300 usable slices, or 2 hours, 1 minute, and 40 seconds.

Hmm, this is a bit strange. If I remember correct, I used to get normal boss fps on Marrowgar (30-45), but always dropping to like 20-25 on Saurfang and especially Blood Queen.

Can't post any benchmarks and stuff like that, as I'm not raiding atm. But my specc is:
E8600 (OVerclocked to 3.6GHZ with stock CPU fan)
GTX 295 (no speeds changed)
4GB DDR3 RAM, CL9 1600MHz I believe it was (dont have that computer atm, so cant check)

Everything on Ultra except for shadows which was one step down.

[Cilraaz]

Well, I had the second half of ICC benchmarked... then I crashed at the end of Lich King (during the transition to the stupid video), and I forgot to stop the benchmark. As such, the results never saved.

I checked during whelps in front of Sindragosa. The lowest I noticed was 20fps. I believe I've seen it dip to 18fps in past runs, though.

Anyway, after ICC25, I downclocked my CPU to stock speeds and ran RS25... well, Halion at least. We somehow had an instance that had the minis cleared already. So, I present my RS25, stock i5 750 parse.

• System Specs:
  
  • Intel Core i5 750 @ 2.66GHz (133MHz base clock, 20x multiplier)
  • ATi Radeon HD 5770 @ stock speeds (850MHz core, 1200MHz VRAM)
  • G.Skill DDR3 4GB CL9 @ 1333MHz (133MHz base clock, 2:10 multiplier)
• WoW Settings:
  
  • 1920x1080
  • 24-bit Color
  • 8x Multisampling (Edge Detect forced from drivers, 16x AF forced in drivers)
  • Ultra, minus half shadows
  • V-Sync Disabled
  • N/A
  • Maximized Windowed Mode
  • All extra effects enabled
• Benchmark Setup:
  
  • RS - 25-man
  • Partial Instance - Halion only
  • No extenuating circumstances
• Benchmark Results:
  
  • Min: 18 fps
  • Max: 93 fps
  • Avg: 44 fps

Comparing this to the previous RS25 benchmark, the minimum framerate was 38% lower, the maximum framerate was 15% lower, and the average framerate was 14% lower. This "proves" one of the commonly held beliefs on this forum: CPU speed affects your minimum framerate most.

Well, time to go clock my CPU back up. It's odd seeing it only 2*C over ambient.