• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Minimum frame-rate in GPU reviews

Joined
Jul 14, 2008
Messages
438 (0.08/day)
Location
Oslo, Norway
System Name Water
Processor Q9550 Stock
Motherboard Asus P5Q Pro
Cooling Noctua big ass cooler
Memory Corsair Dominator PC8500 4 GB
Video Card(s) Sapphire HD7950
Storage Velociraptor 300 GB, Samsung 500 GB, MyBook 300 GB
Display(s) 1: Eizo 23" LCD Foris FS2331-BK 2: Samsung 226BW
Case Cooler Master ATCS 840
Audio Device(s) Asus STX
Power Supply Corsair HX520
I love TPU's graphic cards reviews, the layout and summaries are top-notch and I've not seen anyone do similar.

But what I'm missing is minimum frame rates. See this example:



Compare the 5870 to the 4890. While the average FPS increased by 50%, the minimum FPS increased by 100%. The increased average FPS does not follow increased minimum FPS (look at 4890 vs GTX260 where the GTX has lower average but higher min).

I hope this will be implemented in a foreseeable future.
 

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
27,046 (3.71/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
can't measure this reliably in all tests. no plans to implement in the forseeable future
 
Joined
Jul 14, 2008
Messages
438 (0.08/day)
Location
Oslo, Norway
System Name Water
Processor Q9550 Stock
Motherboard Asus P5Q Pro
Cooling Noctua big ass cooler
Memory Corsair Dominator PC8500 4 GB
Video Card(s) Sapphire HD7950
Storage Velociraptor 300 GB, Samsung 500 GB, MyBook 300 GB
Display(s) 1: Eizo 23" LCD Foris FS2331-BK 2: Samsung 226BW
Case Cooler Master ATCS 840
Audio Device(s) Asus STX
Power Supply Corsair HX520
Ok. Too bad really.

So if you say that it can't be measured with repetitive accuracy, does that mean that other tests which have min frame-rates should be taken with a pinch of salt?

Another approach could be to have a statistical standard deviation, to know in what range the majority of the frames rates are in. But this maybe is the same for all boards.
 

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
27,046 (3.71/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
if the benchmark mode doesn't support minimum frame rate measurements, how do you propose to get such data? one could use something like fraps to measure the frames but then how do you eliminate title and loading screens which run at super low or super high frame rates?
 

AsRock

TPU addict
Joined
Jun 23, 2007
Messages
18,874 (3.07/day)
Location
UK\USA
Processor AMD 3900X \ AMD 7700X
Motherboard ASRock AM4 X570 Pro 4 \ ASUS X670Xe TUF
Cooling D15
Memory Patriot 2x16GB PVS432G320C6K \ G.Skill Flare X5 F5-6000J3238F 2x16GB
Video Card(s) eVga GTX1060 SSC \ XFX RX 6950XT RX-695XATBD9
Storage Sammy 860, MX500, Sabrent Rocket 4 Sammy Evo 980 \ 1xSabrent Rocket 4+, Sammy 2x990 Pro
Display(s) Samsung 1080P \ LG 43UN700
Case Fractal Design Pop Air 2x140mm fans from Torrent \ Fractal Design Torrent 2 SilverStone FHP141x2
Audio Device(s) Yamaha RX-V677 \ Yamaha CX-830+Yamaha MX-630 Infinity RS4000\Paradigm P Studio 20, Blue Yeti
Power Supply Seasonic Prime TX-750 \ Corsair RM1000X Shift
Mouse Steelseries Sensei wireless \ Steelseries Sensei wireless
Keyboard Logitech K120 \ Wooting Two HE
Benchmark Scores Meh benchmarks.
if the benchmark mode doesn't support minimum frame rate measurements, how do you propose to get such data? one could use something like fraps to measure the frames but then how do you eliminate title and loading screens which run at super low or super high frame rates?

What of those that do, although not sure how many that is lol.

As when possible i would not mind seeing min rates more often as they give you a better idea.
 
Joined
Jul 14, 2008
Messages
438 (0.08/day)
Location
Oslo, Norway
System Name Water
Processor Q9550 Stock
Motherboard Asus P5Q Pro
Cooling Noctua big ass cooler
Memory Corsair Dominator PC8500 4 GB
Video Card(s) Sapphire HD7950
Storage Velociraptor 300 GB, Samsung 500 GB, MyBook 300 GB
Display(s) 1: Eizo 23" LCD Foris FS2331-BK 2: Samsung 226BW
Case Cooler Master ATCS 840
Audio Device(s) Asus STX
Power Supply Corsair HX520
All are valid points and I don't have a good suggestion. I didn't think of it that way, so I'm sorry for my noobness.
 

Solaris17

Super Dainty Moderator
Staff member
Joined
Aug 16, 2005
Messages
25,879 (3.79/day)
Location
Alabama
System Name Rocinante
Processor I9 14900KS
Motherboard EVGA z690 Dark KINGPIN (modded BIOS)
Cooling EK-AIO Elite 360 D-RGB
Memory 64GB Gskill Trident Z5 DDR5 6000 @6400
Video Card(s) MSI SUPRIM Liquid X 4090
Storage 1x 500GB 980 Pro | 1x 1TB 980 Pro | 1x 8TB Corsair MP400
Display(s) Odyssey OLED G9 G95SC
Case Lian Li o11 Evo Dynamic White
Audio Device(s) Moondrop S8's on Schiit Hel 2e
Power Supply Bequiet! Power Pro 12 1500w
Mouse Lamzu Atlantis mini (White)
Keyboard Monsgeek M3 Lavender, Akko Crystal Blues
VR HMD Quest 3
Software Windows 11
Benchmark Scores I dont have time for that.
if the benchmark mode doesn't support minimum frame rate measurements, how do you propose to get such data? one could use something like fraps to measure the frames but then how do you eliminate title and loading screens which run at super low or super high frame rates?

cant you hot key enable/disable? i mean it would probably make it mroe hands on then you want but..
 

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
27,046 (3.71/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
All are valid points and I don't have a good suggestion. I didn't think of it that way, so I'm sorry for my noobness.

dont worry, maybe you'd had an idea that i didnt think of. discussion is good
 
Top