Jump to content
Mr. Sandwich

CPU testing, old AMD vs Ryzen with Rx 580

Recommended Posts

Basically:

My old AMD cpu had terrible, terrible FPS in Squad (under 30 most of the time), even when overclocked from 2.6 to 3.6GHz and paired with a brand-new RX 580 8GB graphics card. Other games give me very nice 50+ FPS on decent graphic settings. 

 

With the new overclocked CPU, I still get pretty low FPS but at least it's 50+ for lows and 80+ for highs. Other games easily reach 100+ FPS on decent settings. 

 

 

Old Important stuff:

  • CPU Stock: AMD 1035t 2.6 GHz 6-core with a boost to 3.1GHz
  • CPU Overclocked: AMD 1035t 3.7 GHz
  • RAM: Corsair XMS3 DDR3 1600MHz Ram 8GB running at 1493mhz (game crashed any higher than that)
  • GPU: Powercolor Rx 580 Red Dragon 

 

  • CPUz single thread score: 140-150 max when overclocked
  • CPUz multi thread score: 1200-1300 max when overclocked

 

  • Overclocked 1035t CPU roughly equal to a 1090t or an AMD FX or Phenom ii 955/965 BE running at the same Overclocked GHz speeds (roughly)

 

New Important Stuff:

  • CPU: AMD Ryzen 7 1700 3.64GHz  1.286v
  • RAM: Corsair Vengeance LPX 3000MHz 16GB (2 x 8 GB) Running at 2926.4 Mhz 
  • GPU: Powercolor Rx 580 Red Dragon (yeah same GPU in both tests).

 

  • CPUz single thread score: 420-440 max when overclocked
  • CPUz multi thread score: 4500-4700 max when overclocked

 

 

 

OLD AMD 1035t 1920 x 1080 Low/Off settings 

  • AMD 1035t @ 2.6 GHz- 41 FPS in firing range, unknown unplayable rage FPS in multiplayer (ASSUME 19-22 area).
  • AMD 1035t @ 3.7 GHZ @ 1920 x 1080 LOW/OFF- 55ish FPS in firing range, 28-31 FPS average in multiplayer (70 players in server).

 

This is the extent of the testing with this old CPU. It was just bad. I get 50+ FPS on decent settings with the RX 580 in other games. 

 

NEW Ryzen 7  @ various resolutions and clock speeds:

 

Fraps Benchmarks:

 

@3.64 GHz 1080p Epic

Chora Valley: 

Frames, Time (ms), Min, Max, Avg
144910,   2670422,   45,  97, 54.265

 

Chora Valley again:

Frames, Time (ms), Min, Max, Avg
237548,   4492500,  39,  95, 52.877

 

Sumari Bala:

Frames, Time (ms), Min, Max, Avg
127894,   2305453,  42,  82, 55.475
 

 

 

More map info coming:

 

 

All fps numbers beyond this point are "eyeballed" just by watching the counter during gameplay. Basically with 78 players 45-55 during city combat, 60+ in nature; almost always 60+FPS with under 60 players in server. 140FPS+ on empty or near empty servers - 90-100FPS in firing range. Low vs Epic was extremely minor when it came to FPS differences. CPU usually only has 1-2 cores reaching 100% out of 8, sometimes no cores reach 100%. 

 

 

eyeballed FPS numbers:

Spoiler

 

Ryzen 7 1700 - 1920x1080 - Low/Off Settings:

 

 

  • Ryzen 7 @3.00GHz stock: TBD 
  • Ryzen 7 @3.20GHz: 50-55 Avg in multiplayer. Dip to 45 during combat. 90-94 FPS running in firing range, dip to 88 going full auto on ground.

 

  • Ryzen 7 @3.64GHz: 55-65 Avg in multiplayer. 99-101 FPS firing range, dip to 94 FPS when full auto on the ground below me.  Chora Valley map 55-78 with dips to 48FPS in heavy smoke combat. avg 50s in the city areas, 70s in nature. Fool's Road map (half cores enabled): 60-75 FPS avg with a dip to 58.  Chora Valley map: 100-134 FPS in main base spawn before everyone else spawned, full 70+ server 70s when in nature/fields, 50s in buildings, dip to 48. Sumari Bala map: 50s in city, 70s in nature. Chora Valley (half cores enabled): still about the same FPS. 50s-70s and same temp also. Sumari Bala: 45-55 in city all round. Kohat Toi River Valley:  50-55 FPS. Sumari Bala  28 players 90-120, 80FPS with 50 players, 49-65 with 72 players. 

 

  • Ryzen 7 @3.80GHz: 60-70 Avg in multiplayer. 101-102 FPS when running around in firing range, dip to 94 FPS when full auto on the ground below me. 106 max.. multiplayer Al Basrah map 55-60 fps, Kokan Valley (20 players): 110-160 FPS, (30 players 105-120) (48 players 90-100) (60 players 75-80 fps) (78 players 55-60 in city, 80s in nature) 60c max after multiplayer round 1.384v.  Fool's road (72 players) 60-70s in heavy combat, 80-90 in nature max temp 63c. Sumari Bala (40 players) 90-120 FPS, 72 players 60-70FPS max temp 53c 1.406v. OP First Light Map: 60-78 fps 63c. Kokan Valley 70-80FPS 58c 1.384v,

 

  • Ryzen 7 @3.80GHz with affinity set to only use the best 4 cores: 100-111 FPS when running around in firing range 116fps max, dip to 97 FPS when full auto on the ground below me. Chora Valley (74 players 49 ping): 55-65 FPS in city, 55c max at 1.373v , Al Basrah: 50-55 in city, 55c 1.373v

 

 

1920x1080 Epic preset settings

(motion blur off in multiplayer)

 

 

  • Ryzen 7 @3.20GHz  45-58 multiplayer. Basically 45-50 in combat, 55-60 in nature. 45-55 on average. 68-93 FPS in firing range running round.
  • Ryzen 7 @3.64Ghz55-65 on average. 49-55 multiplayer in laggiest conditions, 65-75 alone in nature.  70-98 FPS in firing range running around, Sumari Bala (60 players)75-80 FPS pretty steady, (44c evo 212) 1.243v, Fool's Road (50 players not normal server 55-65 FPS )
  • Ryzen 7 @3.80Ghz  60-70 FPS on average. 70-100 FPS in firing range running around. Multiplayer Kokan Valley 50-55 FPS in city heavy combat, 60-70 in nature 61c 1.395v,  al basrah dips to 50-55 in city heavy combat 61c 1.384v, Kohat toi river 50-66 city 70-78 nature 62c 1.384v, Sumari Bala (50 players)70-80fps

 

 

 

FPS seems to plunge as more players load into the server. I join a 78 player server with only 30 players in it and will get 120 FPS, it's amazing.. then as soon as 70+ players are in, I will get 50 FPS. Perhaps 78 player servers should consider dropping down to 64 players if this lag is a problem for everyone. Also, 74/76/78 player servers kind of suck if you are the last one to join (like from a queue), usually all squads are 9/9 and you are stuck alone or with 1-2 other stragglers.

 

 

 

Ryzen 7 @ 3840 x 2160 4k resolution

Spoiler

Low/OFF Firing Range: Dips to 33 FPS when full auto on ground below me (bipod LMG full mag). When running around it's steady at 70-80 fps, peak of 98 fps
Ryzen 7 @ 3840 x 2160 4k resolution Low/OFF Firing Range with audio quality set to LOW: still dips to the same 33 fps when I unload on the ground below me, no difference.
Ryzen 7 @ 3840 x 2160 4k resolution Low/OFF Online 70 players: 30-60 FPS, dips to low 30s during firefights

 

Ryzen 7 @ 3200 x 1800 resolution

Spoiler

, Low/OFF: 84-99 FPS in Firing Range. Dips to 45 fps when going full auto on the ground otherwise steady at 84-99 FPS when running around
Ryzen 7 @ 3200 x 1800 resolution, Low/OFF, Online 70 players: 45-65 FPS, likes to hover right around 55 FPS. 

 

Ryzen 7 @ 2560 x 1440

Spoiler

Low/Off Firing Range: Running around steady at 90-99 FPS, dips to 67 FPS when full auto into the ground
Ryzen 7 @ 2560 x 1440 Epic Preset in Firing Range: 51-68 FPS running around, 60 average. Dips to 42 FPS when going full auto on the ground below me
Ryzen 7 @ 2560 x 1440 Low/Off online 70 players: 55-75 FPS, over 60 fps most of the time
Ryzen 7 @ 2560 x 1440 Epic Preset online 70 players: 47-59 FPS, usually right around 50 with a very short dip to 34 fps during a large battle


 

Other hardware just in case you are interested:

Spoiler

userbench scores + hardware running at 3.80 GHz:

http://www.userbenchmark.com/UserRun/4715025

 

Hard drive: KingDian 480gb SSD on Sata 3 (KingDian? it's actually been awesome and gets great benchmark scores too).

Power Supply: EVGA Supernova 650 G2 Gold (love it)

OS: Windows 10 64bit

Motherboard: Asus Prime x370-A

CPU Cooler: Stock spire cooler that came with 1700. Have 212 evo I will use and test in the future.

Hooked into a Sony STR DH550 4k Receiver, hooked into a 55" Vizio 4k TV (60hz actual with fake 120 hz), Logitech G35 headphones, razer deathadder mouse 3500 dpi

 

 

More info about ryzen at different speeds (stock cooler vs hyper 212, temps, voltages, scores etc): 

Spoiler

 

@ 3.64 GHz Hyper 212 cooler Vs Stock Ryzen 7 1700 CPU cooler.

     Temps and voltages, SMT on/OFF, full cores/half cores

 

***PC has failed to post a couple times while SMT disabled. Very minor temp difference, no FPS difference, recommend "auto". 

***PC Can be just laggy in general with 4 cores disabled, not worth it unless you really have heat problems.

 

Stock Cooler installed:

  • idle temp: 30c @ .403v
  • 55c peak after playing a long round 1.253v  [52c with SMT off 1.264v jump to 3662MHz for some reason]
  • 54c peak with half cores after playing a long round  [48c SMT off 1.264v 3643MHz]
  • CPUz stress temp: 63c [52c SMT OFF]
  • CPUz stress temp half cores: 47c  [46c SMT off]

 

With evo 212 cpu cooler installed:

*not meant for am4 boards, it works if you jerryrig it. Am4 adaptor on the way from ebay.

  • 48c peak 1.253v after playing a long round.
  • idle temp: 25.5c ryzen master,  25c hwmonitor,  25c core temp  .403v
  • CPUz stress temp: 54c. Quick to 50c then slowly climbs to 53c with 53.88c max (ryzen master) 54c max hw monitor, 54c max coretemp  1.253v max
  • CPUz stress temp SMT OFF: 50c quick to 48.5c then slow to 49.5c, max of 50.00c RM 50c HM,  NOREADING from core temp, 1.243v
  • CPUz stress temp affinity half cores: 49c. Quick to 47c then slow to 48.75, max 49.00c RM, 49c HM, 49c CT, 1.243v max
  • CPUz stress temp affinity half cores SMT OFF: 45.5c. Quick to 44.88c, slow to 45.25, max 45.5c 1.232v
  • CPUz stress temp Bios half cores SMT OFF: 44.25c Quick to 42.25, slow to 43.36, max 44.25c  1.221v
  • Prime 95 stress temp: 56c max. Quick to 51c, slow to 53, really slow up to 56.00c max temp ryzen master, 56c hwmonitor, 56c core temp 1.286v max
  • Prime 95 stress temp SMT Off: 56.88c. Quick to 51c, slow to 53.5c, max of 56.88c RM, 57c HM, NOREADING Coretemp,  1.275v
  • Prime 95 stress temp affinity half cores: 54.25c max. quick to 52c, slow to 53.5c, really slow up to 54.25c max RM, 54c HM, 54c CT,  1.243v max
  • Prime 95 stress temp affinity half cores SMT Off: 51.63c. Quick to 48c, slow to 50.5c, max of 51.63c RM, 52c HM, 1.243v
  • Prime 95 stress temp Bios half cores SMT Off: 52.13c. Quick to 48c, slow to 49c max of 52.13c  1.243v

 

on stock cooler:

CPUz Single thread score: 424.0 [Tested again after a full match, ran 3 times, 423.7, 424.2, 423.2] *With SMT turned off in bios 423.7, 424.5, 420.6*

-- With affinity 0-7: 422.7 [422.7, 422.3, 422.7]

-- With affinity 0-6: 421.2 [422.7, 422.0, 423.8]

-- With affinity 0-5: 416.4

-- With SMT OFF Set affinity 0-3 (4 cores, same as using 0-7 w/ SMT on): 424.4, 422.7, 422.2

 

CPUz Multi thread score: 4530.8 *with SMT off 3342, 3349.3, 3356.1 *

-- With affinity 0-7 : 2267.5

-- With affinity 0-6: 2121.5

 

evo 212 installed:

smt on: 421.7, 421.7, 421.4

smt off: 423.5, 423.3, 421.9

smt on affinity half cores 5-8: 424.6, 424.3, 424.7

smt off affinity half cores 5-8: 425.1, 425.1, 425.0

smt off bios half cores (it chose 0, 1, 8, 9): 420.7, 418.8, 420.7

 

 

@ 3.7 GHz 

CPUz Single thread score:

-- With affinity 0-15 (all): 426.2, 426.8, 426.3 

-- With affinity 0-7: 428.4, 426.4, 427.4

-- With affinity 0-6: 425.5, 427.4, 426.0

-- With affinity 0-5: 426.3,  426.1, 425.8

-- With affinity 0-4: 427.9, 425.9, 426.4

-- With affinity 0-3: 423.0,  421.5, 421.8

-- With affinity 0-2: 419.5

-- With affinity 0-1: 303.9

-- With affinity 0-0: 418.7

 

0-0 only: 418.1

1-1 only: 420.0

2-2 only: 421.0

3-3 only: 418.9

4-4 only: 420.8

5-5 only: 420.7

6-6 only: 419.5

7-7 only: 421.2

8-8 only: 421.7

9-9 only: 421.1

10-10: 421.4 

11-11: 421.8

12-12: 420.9

13-13: 421.0

14-14: 421.2

15-15: 421.4

 

0-7 single: 426.2, 426.7, 427.3

multi: 2298.2, 2300.0, 2299.0

 

8-15 single: 430.5, 430.5, 430.9

multi: 2300.8, 2301, 2300.8

 

Top 6 from above test 7, 8, 10, 11, 14, 15:

430.4, 430.6, 430.0

 

Top 4 from above test 7, 8, 10, 11: 

429.1, 430.3, 430.3

 

Worst 6 0,1,3,4,5,6:

427.1, 425.3, 427.7

 

Worst 4 0,3,6,1

423.9, 425.2, 425.8

 

 

CPUz Multi thread score: 4576.6, 4584.7, 4581.8

-- With affinity 0-7: 2299.4, 2299.4

-- With affinity 0-6: 2150.7

 

Ok so based on the above test, it seems cores 8-15 are better than cores 0-7. This makes a difference because I've been disabling cores 8-15 when running tests on half cores -- oops, it appears that I should be disabling cores 0-7 instead.

 

Max Temp after full round: 3.7ghz 54c 1.275v, 60-75fps avg

SMT off CPUz Stress test 58c all cores 1.286v... half cores 49c 1.264v

 

 

@ 3.80 GHz *Max Stable Speed with reasonable voltages*

*pc failed to post this morning  after being turned off all night. turned SMT back on and it seems to work for now*

 

 

SMT on, 3.8GHz:

Idle: 26c-27c

CPUz single core score (all cores enabled): 438.3, 438.9,  438.1

                                          --- top 4 cores only: 442.1, 441.8, 441.9

multi core score (all cores enabled): 4698.4, 4634.9, 4706.9

                             --- top 4 cores only: 2253.8, 2363.2, 2363.3

smt on cpuz stress test (all cores): 67c @ 1.395v

smt on cpuz stress test (half cores): 51c @ 1.384v

prime 95 stress test (all cores): 75c @ 1.406v 

prime 95 stress test (half cores):64c @ 1.384v

 

3.8 GHz SMT on, set in bios to only be a 4 core cpu. 

437.1, 438.9, 437.9

2329.8, 2336.3, 

cpuz stress test 1.363v 55c

Prime 95: 1.384v 65c

Firing Range 92-99 FPs, 104 FPS max looking at sun   1.363v 45c

 

SMT off 3.8 Ghz

cpuz single thread score all cores enabled: 438.3, 440.6, 438.4

best 4 cores enabled only: 442.4, 442.5, 442.2

 

smt off cpuz stress test (all):  62c @ 1.375v,

smt off cpuz stress test (half) 55c @ 1.363v

 

@ 3.20 GHz, I just selected "eco mode" in my motherboard settings. I wanted a slower clock just for reference. Voltage varies 0.403 - 1.41, 30c idle , 50c max temp after running squad ~1hr in firing range then multiplayer. This mode had a turbo boost to 3.7ghz which seemed to do absolutely nothing except increase voltage.

@3.64 GHz, this is the "performance" preset in my  Asus prime x370a motherboard. This is my preferred setting for now because of the low voltages.

@3.80 GHz this is the max that I can raise the multiplier with stock auto voltage settings. Peaks of 1.384v isn't bad for casual gaming. 

@ 3.875 GHz: boots, 1.439v 72c CPUz stress test all cores **Crash during prime 95 stress test**

@ 3.90 GHz: boots, crashed during cpuz stress testing 1.428v

 

@ 3.925 GHz boots, completes userbench (http://www.userbenchmark.com/UserRun/4714786) 100% cpu ovr, 117pt single, 469pt quad,  1320 multi

***Crash during cpuz stress testing***

All cores:

CPUz Single: 452.3, 452.9, 452.4

Multi: 4865.2, 4863.8, 4862.8

 

Best 4 cores set affinity:

Cpuz single: 457.1, 456.9, 456.5

multi: 2442.0, 2440.4, 2441.7

 

 

@ 3.950 GHz boots, completed userbench, (no link saved. 102% cpu, 118pts single core, 199% multicore is what i remember) crashed during cpuz stress test

@ 4.00 GHz boots at 4GHz, crashes during userbench 8 core multi test

 

 

 

Side notes:


 

Spoiler

 

Windows Power Plan  is set to "Ryzen Balanced" - this is new, it keeps the speed up for less delay in gaming. works great! no worse than "high performance"

 

If you have a Ryzen 3 or Ryzen 5 running at similar GHz, I'm pretty sure you can expect similar results. Squad does not care about my extra cores (actually disabling 4 of my cores shows an increase of 2-10 fps in firing range* unsure about multiplayer, can't tell).

 

To find out what your CPUz score is. Download it from http://www.cpuid.com/softwares/cpu-z.html (official site, safe, small program) and install, click on the "Bench" tab on the top, then click "Bench CPU". Run 3 times to get a good average. 

 

setting priority to "high" in task manager does absolutely nothing for my fps. running apps are chrome, fraps, hwmonitor and squad. windows 10

 

With the 6 core AMD 1035t, I found no loss of FPS in training mode when I only had 2 cores enabled. I kept 4 cores enabled when I overclocked, turning off 2 cores allowed me to increase voltage without as much heat. Task manager shows all cores being used at a low %, however Squad doesn't seem to use 6 cores all at once. No fps loss with only 2 of 6 cores enabled,  a lot of FPS lost with only 1 core enabled.

 

While running the old system at 3.7 GHz, the game would crash unless my RAM speed was below 1500MHz even though it was proven stable at 1600MHz (the ram speed is supposed to be 1600 MHz but you have to enable XMP to get above 1333MHz). Voltages and timings were not in play here. Game insta-crash at over ~1550Mhz then 0 issues at 1493Mhz).

 

New system, no issues so far at 3000.

 

Why do I have a 1035t? Because I burned up my AMD 4Phenom 965 and my FX 8150 then found this guy on ebay for pretty cheap, full intentions of burning him up too. 

I originally ordered a Ryzen 5 1600 but it failed to post with the motherboard, so I RMAd it, while I waited I overclocked the crap out of my CPUs. Little did I know that core temp and HW Monitor showed readings roughly 20 degrees below actual. The first CPU I thought ok sure, then the 2nd... 

So then the replacement Ryzen 5 1600 and the replacement motherboard showed up, aaand still CPU error, no post. Luckily I got to just return it all. I bought the Ryzen 7 1700 simply because I've read DOZENS of horror stories about how the Ryzen 5's do not post with many motherboards unless you borrow a ryzen 7, plop it in, update the bios, then it's good to go.. After 2 failed Ryzen 5's, Give me the Ryzen 7.

 

Windows now boots so fast, it no longer says "Loading Operating System" in the upper left corner of the screen. The black screen appears, but the text doesn't have time to show, already off to the next screen.

 

 

 

 

 

Edits (keeping track of all the numbers):

Spoiler

 

@ 3.80 GHz 1920 x 1080 low/off, looking at 1 direction without moving. Number in [brackets] is the same test ran at 3.20 GHz 

Set affinity 0-15 cpu (all cores enabled): 108 FPS [95 FPS] 

Set affinity 0-14 cpu (unchecked 1 of 16):  108 [95] 

Set affinity 0-13 cpu: 108 [95] 

Set affinity 0-12 cpu: 108 [95]

Set affinity 0-11 cpu: 108 [95]

Set affinity 0-10 cpu: 108 [95]

Set affinity 0-9 cpu: 108 [95]

Set affinity 0-8 cpu: 108 [95]

Set affinity 0-7 cpu: 110 FPS [96]

Set affinity 0-6 cpu: 110 FPS [96]

Set affinity 0-5 cpu: 108  [94]

Set affinity 0-4 cpu: 106 [92]

Set affinity 0-3 cpu: 102  [89]

Set affinity 0-2 cpu: 102 [90]

Set affinity 0-1 cpu: 86 [76] (2 boxes checked)

Set affinity 0-0 cpu: 68 [54] (only 1 box checked)

 

With SMT Turned off same test 3.64GHz:

Set affinity 0-7: 106

Set affinity 0-6: 106

Set affinity 0-5: 106

Set affinity 0-4: 106

Set affinity 0-3: 110

Set affinity 0-2: 110

Set affinity 0-1:  108

Set affinity 0-0: 67

 

 

 

5120 x 2880 resolution (highest possible resolution) using virtual super resolution AMD setting: 27-65 FPS in firing range, dip to 25 when full auto on the ground below me, otherwise steady in the mid 40s. Online untested, assuming unplayable FPS (below 30)

1024 x 728 resolution LOW/OFF (lowest possible setting): 99-103 steady fps running around in Firing Range, dip to 97 FPS going full auto on the ground. Online untested, probably in the 80-90 range based off of the numbers.

 

 

@ 3.64GHz at 1920x1080 low/off looking at the same spot: 101-102 with all cores enabled, 106-107 with only 0-7 enabled. 

3.7 GZz: 105FPS, affinity 0-7 110FPS, affinity 8-15 112 FPS.

EDIT: at either 1024x728 or 1920x1080 on LOW/OFF I am able to get 108 FPS at 3.8 GHz looking at the same spot.

EDIT AGAIN: If I set affinity and disable the last 4 cores, I now get 110 FPS as a max at 3.8 GHz 1024x728 or 1920x1080 low/off looking a little below the sun on firing range off into the distance (same spot as before).

Edit Again Again: Now if I do the same test at 3.8 GHz but disable the first 4 cores (because they are worse apparently): 116 FPS

 

 

Edited by Mr. Sandwich

Share this post


Link to post
Share on other sites

Thanks for posting that extensive review. I just bought a new CPU etc combo today and after much deliberation, decided to stick to Intel this time even though I so badly want to support AMD for picking up their game with Ryzen :(. 

 

Glad it's working out for you (for the most part) though! 

 

Share this post


Link to post
Share on other sites

Squad alone is not worth skipping Ryzen. All other games runs great with Ryzen, it's a great CPU line.

Anyway - thanks for posting FPS ranges, It's nice to have it.

Share this post


Link to post
Share on other sites
15 hours ago, palanoid said:

your 1080p epic preset result is a lot better than what i have been getting.

 

are you on the latest radeon crimson driver?

Yes the latest version, 17.7.2 

My Rx 580 is very slightly overclocked from manufacturer as well

Share this post


Link to post
Share on other sites
12 hours ago, Mr. Sandwich said:

Yes the latest version, 17.7.2 

My Rx 580 is very slightly overclocked from manufacturer as well

so it'd be running at 1350mhz, right? i wanna try to match your clockspeeds with the rx480

Share this post


Link to post
Share on other sites
5 hours ago, palanoid said:

so it'd be running at 1350mhz, right? i wanna try to match your clockspeeds with the rx480

Yes 1350

 

GPU Radeon RX 580
Boost Clock 1350 MHz

Memory

Effective Memory Clock 2000 MHz (8.0 Gbps)
Memory Size 8GB
Memory Interface 256-Bit
Memory Type GDDR5

Share this post


Link to post
Share on other sites

i'm interested to know your fps ranges in maps like fool's road, yehorivka and gorodok. my fps usually drops to low 40s even high 30s, in epic preset

 

specs: ryzen 5 at 3.8ghz, ram at 3066 c16-18-18-36, rx480 at 1350

Share this post


Link to post
Share on other sites
16 hours ago, palanoid said:

i'm interested to know your fps ranges in maps like fool's road, yehorivka and gorodok. my fps usually drops to low 40s even high 30s, in epic preset

 

specs: ryzen 5 at 3.8ghz, ram at 3066 c16-18-18-36, rx480 at 1350

I've made the 1920 x 1080 results a bit easier to read. On epic or low (doesnt seem to matter) I generally stay over 60 fps and the lowest ever I've seen was 49 FPS for a brief second, most of the time in heavy combat its 50-55ish and 70+ in open fields/BFE

Share this post


Link to post
Share on other sites

Edit: ill keep posting fps numbers from fraps benchmark because it's easy and accurate. 

 

 

Feel free to ask if you want me to run the game at any other settings or clock speeds.

Edited by Mr. Sandwich

Share this post


Link to post
Share on other sites

What are you using to actually benchmark your FPS? The only program that I found that could do it reliably was Fraps (it creates CSV files with all the individual frame data, and at the end spits out a min/max/average summary). If there is any newer software I would love to hear about it!

 

Or are you just eyeballing the FPS numbers?

Share this post


Link to post
Share on other sites
16 hours ago, Psyrus said:

What are you using to actually benchmark your FPS? The only program that I found that could do it reliably was Fraps (it creates CSV files with all the individual frame data, and at the end spits out a min/max/average summary). If there is any newer software I would love to hear about it!

 

Or are you just eyeballing the FPS numbers?

I'm using fraps and eyeballing. Specifically I watch for the lowest numbers during heavy firefights, highs, and overall range where the numbers hover.

 

 It's definitely not 100%, this was quick testing to see any real noticeable differences. If I made a change that resulted in +/- 5 fps I could usually see it immediately. 

Share this post


Link to post
Share on other sites
7 hours ago, Mr. Sandwich said:

I'm using fraps and eyeballing. Specifically I watch for the lowest numbers during heavy firefights, highs, and overall range where the numbers hover.

If you already have fraps then I would highly recommend its benchmarking feature. My eyeballed low/high figures were totally different from what fraps reported after 10 minutes of play, and the true average calculation is really helpful. But you've already done so much testing that I would say that's more "for next time" advice :P 

Share this post


Link to post
Share on other sites
On 23.8.2017 at 4:52 AM, Psyrus said:

If there is any newer software I would love to hear about it!

Mirillis Action!

It can meter your average fps on real-time on your screen. If you play 10 hours, then theres yours 10 hours average fps. You can reset it anytime with key-combo, and it start meter your avg fps on that point, to the eternity.

 

I bought it about 1y ago, 29.95usd.

- Display current and average game framerate (FPS) during gameplay and save your benchmark results.

https://mirillis.com/en/products/action.html

 

Share this post


Link to post
Share on other sites
2 hours ago, Psyrus said:

If you already have fraps then I would highly recommend its benchmarking feature. My eyeballed low/high figures were totally different from what fraps reported after 10 minutes of play, and the true average calculation is really helpful. But you've already done so much testing that I would say that's more "for next time" advice :P 

I didnt know about that feature, thanks.

 

*deletes everything*

Share this post


Link to post
Share on other sites
2 hours ago, Hiikeri said:

Mirillis Action!

 

Thanks for the info! It's always nice to have alternatives. 

58 minutes ago, Mr. Sandwich said:

I didnt know about that feature, thanks.

No worries!

Share this post


Link to post
Share on other sites

You can also record FPS / frametimes with a combination of Afterburner and HWinfo + hardware stats at a given time point, bit more work to tansfere data frome the log file to a calc sheet and prep the Diagramms.

 

 

Edited by moppel

Share this post


Link to post
Share on other sites
1 hour ago, moppel said:

You can also record FPS / frametimes with a combination of Afterburner and HWinfo + hardware stats at a given time point, bit more work to tansfere data frome the log file to a calc sheet and prep the Diagramms.

 

 

I havent experiened micro studder at any resolution. I'll try some more 4k tonight.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×