V2 - Benchmarking Jitter's CPU/GPU performance on your computer - V2
This is a continuation of the thread "Benchmarking Jitter's CPU/GPU performance on your computer". It is a thread with postings of the results of a benchmarch Pedro Santos made. Rob Ramirez gave a suggestion to make the patch better. The results for NVIDIA cards improvd a lot.
That's why this follow up thread: here is the new patch and the relevant results. I post to results we got so for the the V2 patch with some relevant discussion underneath.
It is a great way to see the performance of your system in Jitter. When you post remember to include your systems components, system OS and Max version. Here is the patch:
Here are the results and relevant remarks with V2 of the patch:
(ROB RAMIREZ - about the old patch:
the nvidia card is showing poor results in the two gpu-geometry tests because those patches are using fixed function shading. most likely the newer cards are not optimized to perform fixed-function shading. simply connecting a gl.material object to the gl.multiple, and the 3 gl.gridshape objects in the gpu-geometry sub-patches gives similar results with the nvidia card.)
==================================================================================================
PHIOL
mbp 15"
OSX 10.9.4
Processor 2.8 GHz Intel Core i7
Memory 16 GB 1600 MHz DDR3
Graphics NVIDIA GeForce GT 750M 2048 MB
CPU: 170.4
GPU Geometry 1: 51.6
GPU Geometry 2: 198.1
GPU Pixel Shaders: 90.8
=======================================================================
DTR
i5 4690k 3.5GHz (slight OC)
GA-Z87X-OC mobo
EVGA (Nvidia) GTX670 2GB FTW gfx card
8GB RAM 1600Mhz
Win 7 pro 64 bit
max 6.1.9 first run
CPU: 126.0
GPU Geometry 1: 242.1
GPU Geometry 2: 236.2
GPU Pixel Shaders: 245.7
max 6.1.9 subsequent runs
CPU: 164.6
GPU Geometry 1: 175.3
GPU Geometry 2: 247.1
GPU Pixel Shaders: 246.4
max 7 first run
CPU: 197.2
GPU Geometry 1: 277.9
GPU Geometry 2: 307.8
GPU Pixel Shaders: 331.3
max 7 subsequent runs
CPU: 204.7
GPU Geometry 1: 243.1
GPU Geometry 2: 301.7
GPU Pixel Shaders: 331.5
I stated first and subsequent run results because there are weird discrepancies. Getting faster on 1 element and slower on the other…
Max 7 clearly improves performance over 6. Both tested in 32bit mode.
===============================================================================
PEDRO SANTOS
It’s funny, I was just going to post something related to what Rob Ramirez wrote (welcome to the thread, Rob!).
When I wrote the original patch, I noticed discrepancies between using the fixed-function OpenGL pipeline and the current programmable pipeline (using shaders). For instance, in the GPU Geometry Test 1 (using jit.gl.multiple), with my AMD Radeon HD 4870, the results were something like:
fixed-function pipeline (plain OpenGL material): 100 fps
programmable pipeline (using shaders – jit.gl.material): 68 fps
So I thought that the standard OpenGL implementation would always be quicker.
As the tendency has been for a while to "deprecate the fixed-function pipeline (here’s a nice article about the subject), probably more recent graphics cards’ drivers don’t optimise for it.
Thinking about this, I was going to suggest to build a complex OpenGL scene with objects using jit.gl.material, render-to-texture, depth buffers and post-processing effects (maybe Max7 shadows and other render passes).
Anyway, here’s another nice lesson from this thread: USE JIT.GL.MATERIAL! It’s probably faster! Who would have thought?
By the way, if that is the case, shouldn’t Jitter’s default behaviour be NOT to use the fixed-function pipeline? Or, at least, there could be a simple option similar to the OpenGL Readback mode present in the OpenGL Status menu? Just a thought…
DTR: nice numbers! Finally!
Thanks to MRMAARTEN’s findings and visits to the Apple Store ;-) , thanks to SPA’s results’ compilation and thanks to everyone’s contributions and for making this thread useful.
===============================================================================
DTR
Btw, for direct comparison we’re gonna need some of the AMD graphics owners to run the new version too. The outcome may be that the one shading method is better for Nvidia gfx and the other for AMD. Will be useful to be aware of this. Same for Intel integrated gfx.
===============================================================================
DADDYMAX
Rob Ramirez – thanks for explaining that to us – i’d been a bit worried id just bought a new computer that was going to bottleneck due to an inexplicably crap graphics card – greatly appreciated.
===============================================================================
ZIPB
Hackintosh OSX.9.5 i7 2700k AMD 6870 1 GB
Jitter Benchmark 1.012
Max 7.0.0
CPU: 174.6
GPU Geometry 1: 31.1
GPU Geometry 2: 185.0
GPU Pixel Shaders: 175.7
Max Runtime 6.0.8
CPU: 134.0
GPU Geometry 1: 30.4
GPU Geometry 2: 105.5
GPU Pixel Shaders: 172.7
===============================================================================
JESSE
Mac Pro 6-core 3.33 Ghz 12GB RAM OSX 10.8.5 GeForce GTX680 2048MB
Benchmark v. 1.012
Max 7.0.0
CPU: 147.2
GPU Geometry 1: 54.2
GPU Geometry 2: 193.9
GPU Pixel Shaders: 182.9
Max 6.1.9
CPU: 109.2
GPU Geometry 1: 51.4
GPU Geometry 2: 143.1
GPU Pixel Shaders: 127.7
===============================================================================
DTR
@Jesse: Interesting, your Mac system is kinda in the same league as my Win PC. Your GTX680 gfx card should be a bit faster than my GTX670 yet my system is faster by a large margin. This might point out OSX vs Windows performance differences. Would you happen to have a Bootcamp/Windows install on it for direct comparison?
Though I wonder how meaningful the results are when they’re in the 200+fps range. I’ll try to make a new/adapted test soon, with multi-display rendering etc to stress those high end systems more and see what comes out then. Got a bunch of gigs coming up again so no guaranteed ETA though.
===============================================================================
JESSE
No bootcamp install here but would be glad to run multi-screen tests. I agree the results are troubling regarding Mac/PC performance differences, it’s likely a driver optimization issue.
===============================================================================
MRMAARTEN
i7 920 2.7Ghz
18Gb RAM
NVIDIA GeForce GTX 660 2048 MB
Hackintosh 10.8.5
MAX7
CPU: 99.2
GPU Geometry 1: 34.5
GPU Geometry 2: 150.7
GPU Pixel Shaders: 131.3
MAX6.1
CPU: 86.0
GPU Geometry 1: 30.7
GPU Geometry 2: 101.4
GPU Pixel Shaders: 109.6
-------------------------------------------------------------------------------
Windows 7
Max7
CPU: 152.0
GPU Geometry 1: 175.1
GPU Geometry 2: 196.9
GPU Pixel Shaders: 199.3
Max6.1
CPU: 120.4
GPU Geometry 1: 161.9
GPU Geometry 2: 158.9
GPU Pixel Shaders: 165.8
===============================================================================
MRMAARTEN
MacBook Pro (Retina, 15-inch, Late 2013)
2,3 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 750M 2048 MB
MAX7 10.10.1
CPU: 168.5
GPU Geometry 1: 50.7
GPU Geometry 2: 176.0
GPU Pixel Shaders: 96.0
—–
MAX6.1.9 10.10.1
CPU: 109.9
GPU Geometry 1: 44.1
GPU Geometry 2: 140.8
GPU Pixel Shaders: 94.4
Did also a test with Windows via Bootcamp
MacBook Pro (Retina, 15-inch, Late 2013)
2,3 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 750M 2048 MB
WINDOWS 7
max7 32
CPU: 163.8
GPU Geometry 1: 54.1
GPU Geometry 2: 215.2
GPU Pixel Shaders: 110.1
max7 64
CPU: 216.6
GPU Geometry 1: 54.3
GPU Geometry 2: 213.1
GPU Pixel Shaders: 109.2
Hi MrMaarten
thanks, I got the right new benchmark patch now, it seems much smoother on my screen, never lower than 50fps, but the results seems worse?
MBP 2014 Retina
OS 10.10.1
2.5 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 750M 2048 MB
Max7
CPU: 179.6
GPU Geometry 1: 51.4
GPU Geometry 2: 193.1
GPU Pixel Shaders: 99.1
MacBook Pro (Retina, 15-inch, Late 2013)
2 GHz Intel Core i7
16 GB 1600 MHz DDR32 GHz
Intel Iris Pro 1536
osX 10.10.1
Max7
CPU: 60.0
GPU Geometry 1: 34.9
GPU Geometry 2: 59.0
GPU Pixel Shaders: 59.5
CPU: 161.8
GPU Geometry 1: 34.4
GPU Geometry 2: 187.2
GPU Pixel Shaders: 70.0
Here are my results. Please note that i am actually using the integrated graphics of my CPU for the test.
Max 6
Windows 7
i7-4770k CPU @ 3.50Ghz, 3501 Mhz, 4 cores
8Gb Ram
CPU: 164.5
GPU Geometry 1: 6.1
GPU Geometry 2: 138.6
GPU Pixel Shaders: 47.5
What's changed between v1.012 and v1.0121?
@DTR Ah by re-downloading that must have happened. Nothing changed :)
i7-2600K
16GB Ram
AsusHD7970
Win7 64bit
Max6, 32bit
CPU: 143.9
GPU Geometry 1: 124.9
GPU Geometry 2: 164.0
GPU Pixel Shaders: 166.9
Max7, 32bit
CPU: 165.7
GPU Geometry 1: 163.9
GPU Geometry 2: 170.4
GPU Pixel Shaders: 247.5
Max7, 64bit
CPU: 221.4
GPU Geometry 1: 163.8
GPU Geometry 2: 171.3
GPU Pixel Shaders: 249.3
i7-3820 3.6GHz 4 Core
Asus R.O.G. Rampage IV Formula
8 GB Ram DDR3
NVidia GeForce Gainward GTX 680 2GB GDDR5
Win 7 64 bit
Jitter Benchmark 1.012:
Max7 32 Bit
CPU: 179.6
GPU Geometry 1: 191.4
GPU Geometry 2: 195.2
GPU Pixel Shaders: 249.5
Max6 32 Bit
CPU: 121.2
GPU Geometry 1: 162.0
GPU Geometry 2: 160.6
GPU Pixel Shaders: 248.7
Chart for JITTER_BENCHMARK-V1.0121 - Max7 -32bit
Some results...
Same Desktop PC by MrMaarten i7-920-2,7 GTX-660
Win7 is 1,5x faster than Hackintosh osX8.5
Same MacBookPro by MrMaarten i7-2,3 GT-750m-2Go
Bootcamp-Win7 is similar to osX10.1
MacPro(new) 2014 is really expensive for poor results...
Want Performances? well logically,get a recent Desktop GPU in Windows:
GTX-680, R9-285, HD-7970
Spa: great chart! I'm going to do a Apple Store run again with MAX7 when I have time, and possibly a store with some nice Win laptops...
I am very curious to the lappies with a 850m - 870m, 970m or 980m
Mac Pro 6.1 (nMP)
6core Xeon E5
Firepro D700 (x2)
OS X (10.10.1)
----------------------------
Max 7 32 bit
CPU: 191.5
GPU Geometry 1: 40.5
GPU Geometry 2: 183.9
GPU Pixel Shaders: 217.5
----------------------------
Max 7 64 bit
CPU: 244.6
GPU Geometry 1: 40.7
GPU Geometry 2: 181.3
GPU Pixel Shaders: 216.7
GPU Geometry values are embarrassing.
yup, how about when you run the first version of the test, with the fixed function shading?
@ LIGHTSPEED.JOHNNY: What DTR means is that the previous version of this bench mark could work better on AMD graphics.
Maybe you could try?
On top of the thread https://cycling74.com/forums/benchmarking-jitters-cpugpu-performance-on-your-computer/ you can download the patch, please also post your results in that thread, as it keeps the data nice and clean...
i7-3820 3.6GHz 4 Core
Asus R.O.G. Rampage IV Formula
8 GB Ram DDR3
NVidia GeForce Gainward GTX 680 2GB GDDR5
Win 7 64 bit
Jitter Benchmark 1.012:
Max7 64 Bit
CPU: 233.3
GPU Geometry 1: 237.0
GPU Geometry 2: 199.7
GPU Pixel Shaders: 249.6
Remarks:
It seems to make a difference if I run the test several times. on the first run the numbers were closer to Max 7 32 bit.
GPU2 and PixelShader seem to have a cap on my machine on 200 rsp. 250.
--note - new mac with AMD card, doesn't seem any better than nvidea
iMac (Retina 5K, 27-inch, Late 2014)
processor: 4 GHz Intel Core i7
Memory: 8 GB 1600 MHz DDR3
graphics: AMD Radeon R9 M290X 2048 MB
v1.012
Max 7
CPU: 178.8
GPU Geometry 1: 45.8
GPU Geometry 2: 176.8
GPU Pixel Shaders: 214.5
CPU: 180.0
GPU Geometry 1: 47.4
GPU Geometry 2: 190.1
GPU Pixel Shaders: 214.9
I finally decided to buy a GTX 970 after this version of the benchmark. There are clear improvements by using Max 7 64bits. Thank you guys.
Hackintosh OSX 10.10.1
Processor: i7 4770k 3.9Ghz
Motherboard: Gigabyte z87x-ud5h
Ram: 16Gb
GPU: Gigabyte / Nvidia Geforce GTX 970 4Gb DDR5
Max 7.0.1 64bit
CPU: 257.6
GPU Geometry 1: 97.7
GPU Geometry 2: 271.9
GPU Pixel Shaders: 265.8
Max 7.0.1 32bit
CPU: 171.9
GPU Geometry 1: 84.1
GPU Geometry 2: 264.9
GPU Pixel Shaders: 267.5
Max 6.1.9 64bit
CPU: 212.4
GPU Geometry 1: 74.4
GPU Geometry 2: 214.4
GPU Pixel Shaders: 215.2
Max 6.1.9 32bit
CPU: 114.1
GPU Geometry 1: 72.6
GPU Geometry 2: 207.5
GPU Pixel Shaders: 215.8
It's indeed becoming quite clear that Jitter openGL performs better on Windows than OSX. Especially in GPU geometry 1. That being said, 200fps is in most cases much more than needed.
@Juandaco Do you have a Windows install on that machine too?
Sorry @DTR, I currently don't have a Windows partition on my system, but I'm curious about those benchmarks as well. I can probably do an installation just to compare, but I'm kind of busy this week. I might have some time available next week.
MacBook Pro (Retina, 15-inch, Late 2013)
OS 10.10.1
2.3 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 750M 2048 MB
CPU: 234.7
GPU Geometry 1: 51.8
GPU Geometry 2: 190.0
GPU Pixel Shaders: 98.1
@ DTR
Here is the comparison of v1 vs v2 test. I noticed a extreme drop (326189) in the Geometry1 benchmark (cross checked both versions running on 7.01)!?
Self build PC
XEON E3 1241v3 (3.5/3.9GHz Quadcore)
AMD R9 285
8GByte RAM
WIN 64 Pro, mATX board
MAX 6.1.9 Standalone (32bit)
CPU: 165.2
GPU Geometry 1: 248.4
GPU Geometry 2: 240.3
GPU Pixel Shaders: 249.5
MAX 7.0 (32bit, couldn´t build standalone because "cg.dll missing"?)
CPU: 193.4
GPU Geometry 1: 326.7
GPU Geometry 2: 244.5
GPU Pixel Shaders: 247.9
MAX 7.01 (32bit)
v2 Test
CPU: 196.1
GPU Geometry 1: 189.0
GPU Geometry 2: 248.9
GPU Pixel Shaders: 247.8
Bene
Addendum: VERY bad performance still in 7.0/7.1 compared to 6.19 with 4 contexts on 4 screens (less than half fps)
MacBook Pro 15" Retina Mid 2014
OSX 10.9.5
2.5GHz Intel Core i7
16GB 1600MHz DDR3
NVIDIA GeForce GT 750M 2048MB
Max7.0.1 32bit
CPU: 166.7
GPU Geometry 1: 50.4
GPU Geometry 2: 182.2
GPU Pixel Shaders: 86.2
Max7.0.1 64bit
CPU: 227.4
GPU Geometry 1: 51.5
GPU Geometry 2: 180.0
GPU Pixel Shaders: 85.4
Max6.1.8 32bit
CPU: 108.2
GPU Geometry 1: 49.7
GPU Geometry 2: 132.1
GPU Pixel Shaders: 83.7
Max6.1.8 64bit
CPU: 143.7
GPU Geometry 1: 50.0
GPU Geometry 2: 173.6
GPU Pixel Shaders: 85.6
MacBook Pro Retina 15"
2.8 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 650M-1024 MB
OSX V 10.9.5
MAX 7.0.1 64 bit
CPU: 140.8
GPU Geometry 1: 47.1
GPU Geometry 2: 169.8
GPU Pixel Shaders: 87.2
Macbook pro. 15 inch. 2000.
2.2 Ghz. Intel Core i7.
8 GB 1333 Mhz DDR3
OSX 10.9.5
"Copy these values:
CPU: 125.8
GPU Geometry 1: 19.8
GPU Geometry 2: 116.1
GPU Pixel Shaders: 48.8
Anyone understand the variance between benchmark v1 and v2 here:
HP z230
Xeon E3 3.4 gHz
16.0 GB RAM
Nvidia Quadro k2000
Windows 8.1
Max 7.0.1 64bit
v2 Test:
CPU: 198.7
GPU Geometry 1: 49.3
GPU Geometry 2: 322.6
GPU Pixel Shaders: 112.6
v1 Test:
CPU: 188.4
GPU Geometry 1: 88.2
GPU Geometry 2: 446.5
GPU Pixel Shaders: 113.1
The expectation is that geometry 1 would be better on v2, correct?
Same machine as above, Max 6.1.9
HP z230
Xeon E3 3.4 gHz
16.0 GB RAM
Nvidia Quadro k2000
Windows 8.1
Max 6.1.9 64bit
v2 test:
CPU: 166.2
GPU Geometry 1: 55.1
GPU Geometry 2: 286.5
GPU Pixel Shaders: 112.7
v1 test:
CPU: 165.6
GPU Geometry 1: 91.1
GPU Geometry 2: 377.5
GPU Pixel Shaders: 112.9
I pulled out the malfunctioning gfx card from my hackintosh and ran the test to see if the CPU's integrated Intel HD3000 is working correctly:
i7-2600k
intel HD3000 (512MB shared RAM allocated)
8GB RAM
OSX 10.8.5
Max702
CPU: 110.7
GPU Geometry 1: 15.3
GPU Geometry 2: 48.0
GPU Pixel Shaders: 20.2
Windows Laptop - Graphic Workstation for Jitter an others...
Finally a portable (4Kg!!) machine for everything... no more 2 systems maintenance: MBP + Win desktop
and 3 video out: 2xDisplay-port+ hdmi + laptop 17"
1,5x faster then nMP w. D700 ... half the price ... and portable. :)
2x faster than MBP i7
Schencker W705 - 17"
Win 8.1 Pro 64
16 Go Ram
2x SSD 512 Go
i7-4790S (xeon)
GTX980M-8Go - 2x DP - 1x HDMI
Driver Nvidia 354.22
Benchmark works better with Overdrive off
Max 7.0.3 32bit _bench V2
CPU: 190
GPU Geometry 1: 240
GPU Geometry 2: 250
GPU Pixel Shaders: 300
Max 7.0.3 64bit _bench V2
CPU: 240
GPU Geometry 1: 310
GPU Geometry 2: 260
GPU Pixel Shaders: 300
The correct chart is the 1st one:
HP Pavilion laptop (don t remember type number right now)
CPU i5-4210U 1.7-2.4GHz
8GB RAM
Geforce 840M graphics
Win8.1
Max 703 32bit
CPU: 79.6
GPU Geometry 1: 59.7
GPU Geometry 2: 64.1
GPU Pixel Shaders: 68.6
MacBook Pro Retina (may 2015)
2.8 GHz Intel Core i7
AMD Radeon R9 M370X 2048 MB
OSX 10.10.4
Max 7.0.4
Max32_v1.0121
CPU: 263.9
GPU Geometry 1: 37.0
GPU Geometry 2: 154.5
GPU Pixel Shaders: 93.7
The new Apple flagship MacBook Pro's GPU seems to be slower than the older models NVidia 750m. Why?
CPU: 214.3
GPU Geometry 1: 226.1
GPU Geometry 2: 222.0
GPU Pixel Shaders: 250.4
=======================================
MAX 7 x64
WIN 7 x64
MSI GT60
Core i7-4700MQ 3.2 Ghz
Nvidia Geforce GTX 780M/4Gb GDDR5
Nvidia driver 353.62
RAM DDR III(L) 32Gb
SSD 128 OCZ
SSHD 2x750Gb
This is an update from a previous post, this time with comparison to windows 8.1:
MacBook Pro Retina (may 2015)
2.8 GHz Intel Core i7
AMD Radeon R9 M370X 2048 MB
OSX 10.10.4 - Max 32bit 7.0.4 /// Win 8.1 - Max 64bit 7.0.5
v1.0121
CPU: 263.9 /// 234.6
GPU Geometry 1: 37.0 /// 72.3
GPU Geometry 2: 154.5 /// 238.2
GPU Pixel Shaders: 93.7 /// 127.5
conclusion: so far nothing new - OSX OpenGL implementation sucks.
I upgraded my dual boot hackintosh with a GTX960 and Yosemite. OSX vs Win tests:
Max705 32bit
i7 2600k 3.4GHz
MSI GTX960 2GB
8GB 1600MHz RAM
Gigabyte Z68MA-D2H-B3 mobo
OSX 10.10.5 // Win 7 pro
CPU: 197.1 // 146.6
GPU Geometry 1: 66.3 // 158.3
GPU Geometry 2: 251.8 // 244.2
GPU Pixel Shaders: 202.5 // 246.5
2008 macpro 8x2,8 ghz ati radeon hd 5870
CPU: 119.7
GPU Geometry 1: 22.8
GPU Geometry 2: 101.4
GPU Pixel Shaders: 181.4
iMac, Windows 7 vs Yosemite test
iMac retina 5K, 27-inch, Late2014
4Ghz Intel Core i7
32GB 1600Mhz DDR3
AMD Radeon R9 M295X 4096 MB
Yosemite
CPU: 258.7
GPU geometry 1 (uses jit.gl.multiple): 37.4
GPU geometry 2 (3 high poly-count objects): 188.5
GPU pixel shaders: 205.7
Windows 7
CPU: 286.8
GPU geometry 1 (uses jit.gl.multiple): 158.5
GPU geometry 2 (3 high poly-count objects): 261
GPU pixel shaders: 238.4
As you can see Windows are way faster especially in geometry 1 test.
and another Windows vs. OSX test, both 64 bit, Max 7.1
machine:
Mac Pro (Mid 2012)
2 x 3.06 GHz 6-Core Intel Xeon
28Gb 1333Mhz DDR3 ECC
ATI Radeon HD 5770 1024 MB
OSX 10.10.5
CPU: 233.5
GPU geometry 1 (uses jit.gl.multiple) : 22.7
GPU geometry 2 (3 high poly-count objects): 130.4
GPU pixel shaders: 111.2
Windows 7:
CPU: 247.6
GPU geometry 1 (uses jit.gl.multiple) : 76.5
GPU geometry 2 (3 high poly-count objects): 124.4
GPU pixel shaders: 132.0
Does anybody know why Windows perform better?
Max 7.1
iMac late 2014, Retina 5k, 4GHz i7
24Gb ram
AMD Radeon R9 M295X, 4 Gb
OS 10.11.2
CPU: 252.0
GPU Geometry 1: 37.1
GPU Geometry 2: 185.2
GPU Pixel Shaders: 201.2
I updated my HP Pavilion laptop to Windows 10 Home and Max 7.2.4 32bit
CPU i5-4210U 1.7-2.4GHz
8GB RAM
Geforce 840M graphics (btw, on these dual graphics laptops it's important to tell the Nvidia drivers to use the Geforce card with Max)
CPU: 83.8
GPU Geometry 1: 61.5
GPU Geometry 2: 83.7
GPU Pixel Shaders: 58.6
And the production PC as well:
Win 10 pro 64 bit
Max 7.2.4 32bit
i5 4690k 3.5GHz
GA-Z87X-OC mobo
EVGA (Nvidia) GTX670 2GB FTW gfx card
8GB RAM 1600Mhz
CPU: 193.9
GPU Geometry 1: 242.9
GPU Geometry 2: 237.3
GPU Pixel Shaders: 309.6
MacBook Pro (early 2011)
2,2 GHz Intel Core i7
16 GB 1333 MHz DDR3
AMD Radeon HD 6750M 1024 MB
OSX 10.11
----------
CPU: 71.7
GPU Geometry 1: 7.2
GPU Geometry 2: 43.6
GPU Pixel Shaders: 28.0
----------
seems pretty shitty to me...
@Herr Markant: sounds like Jitter is using your integrated graphics, not the Radeon card. On Windows in the Nvidia drivers settings I can select which apps will use my laptops Nvidia card instead of the Intel integrated graphics. Not sure how that works in OSX.
Hi everyone,
My results:
Intel Core I7-3770K 3.5GHz (Turbo Boost up to 4.1GHz)
16Gb DDR3-1600
NVIDIA GeForce GTX 1060
SSD
Windows 10 x64
Max 7.2.5
CPU: 228.7
GPU Geometry 1: 221.4
GPU Geometry 2: 259.7
GPU Pixel Shaders: 489.3 (black window during this test!)
Benchmark seems to be incompatible with Max 7.2.4 and later.
There are some errors in Max Console on load:
viddll.engine version: 1.0.6
newobj: jit.gl.slab.gauss6x: No such object
newobj: jit.gl.slab.gauss6x: No such object
newobj: jit.gl.slab.gauss6x: No such object
newobj: jit.gl.slab.gauss6x: No such object
Any thoughts?
MSI GS63VR Laptop
Skylake i7-6700HQ 2.6GHz
32GB RAM
Dual SSD
nVidia GTX1060, driver 369.09
Win10 Pro x64
Max 7.2.5 32-bit
CPU: 166.7
GPU Geometry 1: 194.6
GPU Geometry 2: 243.5
GPU Pixel Shaders: 273.0
Max 7.2.5 64-bit
CPU: 200.0
GPU Geometry 1: 183.6
GPU Geometry 2: 195.9
GPU Pixel Shaders: 224
I found out that disabling Hyper-threading in my bios settings gives me a sensible better framerate in my patches.
Here are the results with this test patch, with hyper-threading enabled (default bios settings), and with HT disabled :
windows 10 64 bits
intel Core i7-6700k 4 Ghz
ram : 16 Go
nVidia GTX1070, driver 373.06
Max 7.2.5 32-bit
hyper-threading enabled :
CPU: 169.8
GPU Geometry 1: 175.9
GPU Geometry 2: 228.3
GPU Pixel Shaders: 251.9
hyper-threading disabled :
CPU: 196.9
GPU Geometry 1: 201.0
GPU Geometry 2: 255.8
GPU Pixel Shaders: 264.4
Max 7.2.5 64-bit
hyper-threading enabled :
CPU: 210.2
GPU Geometry 1: 219.5
GPU Geometry 2: 232.4
GPU Pixel Shaders: 247.4
hyper-threading disabled :
CPU: 214.6
GPU Geometry 1: 248.7
GPU Geometry 2: 251.6
GPU Pixel Shaders: 252.0
CPU: 104.2
GPU Geometry 1: 35.1
GPU Geometry 2: 233.3
GPU Pixel Shaders: 369.2
Max 7.0.1 32 bit win 10 - 64bit
Amd 1090T 3.2 ghz
Nvidia Quadro 2000
8 gb ram
after upgrade to max 7.3 64 bit
all is going worst
why this ?
Max 7.3 64 bit win 10 – 64bit
Amd 1090T 3.2 ghz
Nvidia Quadro 2000 2gb ram
8 gb ram
CPU: 110.4
GPU Geometry 1: 42.7
GPU Geometry 2: 231.9
GPU Pixel Shaders: 52.5
MaxForLive 7.3.0
Intel Core I7-3770K 3.5GHz (Turbo Boost up to 4.1GHz)
16Gb DDR3-1600
NVIDIA GeForce GTX 1060
SSD
Windows 10 x64
CPU: 213.5
GPU Geometry 1: 184.5
GPU Geometry 2: 212.8
GPU Pixel Shaders: 223.2
Found a workaround for "newobj: jit.gl.slab.gauss6x: No such object" problem in my previous post:
Copyig jitter_benchmark-v1.0121.maxpat to "C:\Program Files\Cycling '74\Max 7\examples\jitter-examples\render\slab\" folder fixed the GPU Pixel Shaders test.
I't me again, same config as above, just closed Google Chrome windows :)
CPU: 223.1
GPU Geometry 1: 261.1
GPU Geometry 2: 246.8
GPU Pixel Shaders: 262.3
hey, in case people are interested I'm at the Apple Store trying the new MacBook Pro 2016, here is what I get:
MacBook Pro 15' late 2016 OS 10.12.1 - Max 7.3.1
Processor Name: Intel Core i7
Processor Speed: 2.6 GHz
Number of Processors: 1
Total Number of Cores: 4
L2 Cache (per Core): 256 KB
L3 Cache: 6 MB
Memory: 16 GB
Radeon Pro 450:
Chipset Model: AMD Radeon Pro 450
Type: GPU
Bus: PCIe
PCIe Lane Width: x8
VRAM (Total): 2048 MB
Vendor: ATI (0x1002)
CPU: 203.6
GPU Geometry 1: 35.7
GPU Geometry 2: 142.2
GPU Pixel Shaders: 99.5
i want to buy a pc which connects permanently to a videowall with 9 full-hd monitors. right now i tested it with one with one nvidia gtx 1080 - i have 3 matrox tripleheads to go dp editions connected to its 3 dp outs and 9 dp cables to the screens. That works fine, but my question is if i could improve that setup if i use a pc with 2 gtx 1080 inbuilt (nvidia sli) - can one instance of max 64bit make use of both at the same time? If yes, could i also get rid of the matrox devices, since every graphic card has 5 outputs (3 x dp, 1 x dvi, 1 x hdmi) - so can i connect the 9 screens + 1 control monitor to the resulting 10 outputs directly? Anybody has experience with such a setup? Or what do you think would be the best graphics solution on pc at this time in general?
Another question:
Max 32bit had some limitations of RAM use - is the same true for Max 64 bit?
+ For videoplayback i guess it is still best to have two ssds - one with the system, one with the video files - right?
hello Tobias, you should probably start a new topic about this.
I had very good results using two of those on one MacBook Pro recently https://www.amazon.com/gp/product/B011BQ6X2U/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1
Also, using several laptops communicating via OSC works great.
Hope this helps!
Razer Blade Pro 17"
Windows 10 Home
i7 6700HQ Quad-core
32GB RAM
GTX 1080 8GB, driver 369.56
GSYNC: disabled
Max 7.3.1 32-bit
CPU: 162.9
GPU Geometry 1: 181.4
GPU Geometry 2: 242.8
GPU Pixel Shaders: 259.4
Max 7.3.1 64-bit
CPU: 214.2
GPU Geometry 1: 204.8
GPU Geometry 2: 243.8
GPU Pixel Shaders: 249.0
Just a little update to my findings in Dec. 2014. I replaced the R285 with a Zotac GTX 1050 TI, affordable and impressive.
XEON E3 1241v3 (3.5/3.9GHz Quadcore)
Zotac GTX 1050TI (latest drivers)
8GByte RAM
WIN7 64 Pro, mATX board
MAX 7.3.1 (32bit)
v2 Test
CPU: 196.4
GPU Geometry 1: 214.9
GPU Geometry 2: 246.8
GPU Pixel Shaders: 315,4
Bene
Hi all!
CPU: 114.5
GPU Geometry 1: 15.6
GPU Geometry 2: 74.5
GPU Pixel Shaders: 31.5
Mac Mini 2.5ghz intel core i5 late 2012
8 gig ram 1600 mhz ddr3
Os 10.8.5
Max 7.3.1
Intel HD Graphics 4000 768 MB
cheers
jd
A couple more cases:
MacBook Pro 2015 (11,5)
2.8GHz Quad Core i7, 16GB RAM
AMD Radeon R9 M370X
Max 7.3.1
macOS 10.12.1
CPU: 263.2
GPU Geometry 1: 35.5
GPU Geometry 2: 195.3
GPU Pixel Shaders: 96.9
It's interesting that these are better numbers than for the 2016 MacBook Pro above!
Enabling/Disabling 'Automatic Graphics Switching' made no difference to benchmarks.
__________________
Asus UX303LN
2.0 (2.6GHz) i7 Dual Core i7, 12GB RAM
NVIDIA GeForce 840M
Max 7.3.1
Windows 10 (1607)
CPU: 130.6
GPU Geometry 1: 57.5
GPU Geometry 2: 87.2
GPU Pixel Shaders: 60.1
New test victim:
Asus GL702VM laptop
i7-6700HQ quad-core
GTX1060 6GB (G-sync disabled)
16GB RAM
Win 10 home
Max 731 64bit
Interestingly CPU rendering can differ quite a bit between runs. Who knows what Winbugs is doing in the background. GPU performance is quite steady. Also noteable is that my desktop GTX670 (results earlier in the thread) registers faster while it's slower in gaming benchmarks. It also beats Jesse's GTX1080 here, which definitely doesn't make sense. I think this shows that we're not pushing these cards to the fullest with Jitter patches like this. Or perhaps their OpenGL support isn't optimized. GPU-Z shows max GPU core load of 55% and other loads under 25% during the Jitter benchmark.
(NB: obvisouly it doesn't surprise me that high tech commercial games utilize video cards better than our Jitter patches. )
CPU: 188.0
GPU Geometry 1: 205.3
GPU Geometry 2: 199.3
GPU Pixel Shaders: 253.1
CPU: 197.4
GPU Geometry 1: 207.8
GPU Geometry 2: 200.2
GPU Pixel Shaders: 248.9
CPU: 190.4
GPU Geometry 1: 200.4
GPU Geometry 2: 201.2
GPU Pixel Shaders: 252.1
I wasn't all to satisfied with the Asus GL702VM tested above and got an MSI GS63VR to compare. Here are the results with these tweaks applied: http://forum.notebookreview.com/threads/msi-gs63vr-stealth-pro-owners-lounge.795498/page-52#post-10425582
MSI GS63VR laptop
i7-6700HQ quad-core
GTX1060 6GB
16GB RAM
Win 10 home
Max 731 64bit
CPU: 214.4
GPU Geometry 1: 219.7
GPU Geometry 2: 218.0
GPU Pixel Shaders: 232.3
CPU: 198.1
GPU Geometry 1: 197.8
GPU Geometry 2: 222.5
GPU Pixel Shaders: 233.8
Asus GL702VM laptop
i7-6700HQ quad-core
GTX1060 6GB (G-sync disabled)
8GB RAM
Win 10 home
Max 7.3.1 64bit
Result from first trial:
CPU: 185.5
GPU Geometry 1: 204.1
GPU Geometry 2: 186.0
GPU Pixel Shaders: 245.6
However, two days later this is the result:
CPU: 135.6
GPU Geometry 1: 181.0
GPU Geometry 2: 164.5
GPU Pixel Shaders: 226.4
Actually I did several runs both days with similar results: Good the first day, bad/worse two days later. I have no idea why the values dropped. It`s a brand new computer and the only change to the system between the runs was installing Max for Live.
Anyone have an idea of what might cause the drop of values?
Maybe some Automatic Windows update installed a background application or service that is consuming resources (OneDrive, anti-virus, etc). Open Windows' Resource Monitor application and sort the services by CPU usage, maybe you see something there...
Mac Book Pro Late 2016
2.6 GHz Intel Core i7
16 GB 2133 MHz LPDDR3
Radeon Pro 460 4096 MB
Intel HD Graphics 530 1536 MB
CPU: 205.2
GPU Geometry 1: 40.2
GPU Geometry 2: 154.5
GPU Pixel Shaders: 132.9
@JBAILEY: are those scores using Intel HD or the Radeon GPU?
should be Radeon because I turned off automatic graphics switching - but these results can't be right, just tested my 2012 MacBook Pro and got these results? I feel like something weird is up with max that's not taking advantage of the card.
Pro (15-inch, Mid 2012)
2.6 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 650M 1024 MB
Intel HD Graphics 4000 1536 MB
CPU: 202.6
GPU Geometry 1: 44.7
GPU Geometry 2: 191.8
GPU Pixel Shaders: 91.5
You can check what jitter is using under Options> OpenGL status.
Doesn't make sense, my geekbench OpenCL score is 11473 on the 2012 and 47846 on the late 2016. Something's up with the Max Benchmarking patch or with Max.
OpenGL status reads AMD Radeon Pro 460 OpenGL Engine
OpenCL is an entirely different thing than OpenGL. But that score is definitely bad (although I don't expect much good from Apple's these days but not this bad ;). Either the drivers are no good or the rendering methods used in Jitter and/or the benchmark are not well supported by this card/driver.
ya, this can't be right. I'm going to try benchmarking in windows bootcamp to test the drivers
Tested the new MBP late 2016 on windows. As I suspected something's up with either Max or Mac OS in relationship with the radeon Pro 460. Windows is nearly 4x faster on geometry than the same test on the same machine in MacOS.
Mac Book Pro Late 2016
2.6 GHz Intel Core i7
16 GB 2133 MHz LPDDR3
Radeon Pro 460 4096 MB
Intel HD Graphics 530 1536 MB
Mac OS:
CPU: 205.2
GPU Geometry 1: 40.2
GPU Geometry 2: 154.5
GPU Pixel Shaders: 132.9
Windows 10:
CPU: 233.9
GPU Geometry 1: 141.2
GPU Geometry 2: 229.6
GPU Pixel Shaders: 166.3
Why PC is so much faster? is it the Retina or the OS?
MacBook Pro (Retina, 15-inch, Mid 2014)
2.5 GHz Intel Core i7
16 GB 1600 MHz DDR3
NVIDIA GeForce GT 750M 2048 MB
Intel Iris Pro 1536 MB
CPU: 225.0
GPU Geometry 1: 53.5
GPU Geometry 2: 220.6
GPU Pixel Shaders: 106.0
@abao Windows takes advantage of the retina resolution same as Mac OS, it's either a problem with Sierra's drivers or with Max. I'm leaning toward drivers though because it got low scores in GFX benchmark tests too, just 28fps on the Manhattan test onscreen.
Just to follow up GFXbench scores are more than double on Windows what I'm getting on Mac. Manhattan test get's 60fps to just 28 on Mac. Same exact late 2016 MBP with AMD Radeon Pro 460. Why am I not hearing this anywhere else? :(
Perhaps because folks into graphics have moved (back) to windows?
@jbailey: yes its just embarrassing the difference between OS X and Windows in the same exact Apple machine. Recently i been benchmarking in the same way you did but with a nMP 6.1 and with Windows and a modified AMD drivers the benchmarks increased almost x4.
Results of jitter_benchmark-v1.0121 with a MacPro late 2013 3Ghz 32ram & x2 D700
Mac OSX El Capitan
CPU: 59.9
GPU Geometry 1: 31.6
GPU Geometry 2: 58.9
GPU Pixel Shaders: 59.3
Windows 10 Home with original AMD Bootcamp drivers:
CPU: 138.9
GPU Geometry 1: 37.0
GPU Geometry 2: 120.2
GPU Pixel Shaders: 136.5
Windows 10 Home with “hacked" AMD Bootcamp drivers:
CPU: 278.5
GPU Geometry 1: 161.4
GPU Geometry 2: 203.2
GPU Pixel Shaders: 171.1
The original Windows bootcamp AMD driver are downgraded, im actually using a non-official ones since one month ago without a single problem. >>> http://www.bootcampdrivers.com As mentions the video, just monitor your GPU temp and use Macs Fan Control to keep it cool.
Apple strategy is just starting to be more than ridiculous and as mentioned DTR, moving back to Windows can make sense for many of us...
fascinating! I wonder if anyone has benchmarked an External GPU with Max yet. https://bizon-tech.com/us/bizonbox3-egpu.html/
There's a bunch of these coming out soon. Razer and Alienware already released theirs.
for the sake of completeness: 13-inch
MacBook Pro (13-inch, Mid 2012, non-retina)
2,9 GHz Intel Core i7
8GB 1600 MHz DDR3
Intel HD Graphics 4000 1024 MB
MacOS Yosemite
max731
32 bit:
CPU: 137.9
GPU Geometry 1: 13.6
GPU Geometry 2: 83.4
GPU Pixel Shaders: 32.7
64 bit:
CPU: 140.8
GPU Geometry 1: 17.7
GPU Geometry 2: 87.5
GPU Pixel Shaders: 35.8
z270 ud5 mobo
i7700k oc'd to 4.8Ghz
ecvga gtx 1060 sc oc'd to 2 ghz
32gb ram ddr4 2400 mhz
CPU: 241.0
GPU Geometry 1: 256.5
GPU Geometry 2: 265.6
GPU Pixel Shaders: 299.7
Just wondering if there'd been any updates re: the Radeon Pro 460 4096 MB? My late 2013 MBP with NVIDIA GeForce GT 750M 2048 MB needs either costly repair or replacement. I could certainly benefit from faster jit.gl framerates so was looking into new MBP with the 460, but it looks like that might not be a given (unless I install Bootcamp?) Or has anything been updated on the OSX side in recent months?
Zotac ZBOX Magnus-1080BE
(4core Intel Core i7-7700, NVIDIA GeForce GTX 1080)
Windows 10, Max 734 (64 bit)
CPU: 328.3
GPU Geometry 1: 330.8
GPU Geometry 2: 330.7
GPU Pixel Shaders: 405.7
-----------------------------------------------------------------
Mac Pro 6.1 (the "Trashcan" Mac)
(6core Xeon E5, 2x Firepro D700)
OSX 10.12.5, Max 734 (64 bit)
CPU: 256.0
GPU Geometry 1: 39.8
GPU Geometry 2: 175.9
GPU Pixel Shaders: 214.0
-----------------------------------------------------------------
This, in a nutshell, is why I finally switched to PC and Windows. Also, the PC is less than half the price and equally portable.
iMac 27" Late 2012
3.4GHz i7
32GB 1600MHz DDR3
NVIDIA GeForce GTX 680MX 2048MB
OSX 10.12.4
Max 7.3.4 (64 bit)
CPU: 192.8
GPU Geometry 1: 68.0
GPU Geometry 2: 216.0
GPU Pixel Shaders: 239.7
has anyone tried one of those eGPU for Mac?
MacBook Pro Retina, Mid 2015
2.8 GHz Intel Core i7
16 GB 1600 MHz DDR3
Startup 1TB Apple SSD
AMD Radeon R9 M370X 2048 MB
OS X El Capitan 10.11.6
Max/MSP 7.3.4
32 BIT:
CPU: 254.0
GPU Geometry 1: 36.3
GPU Geometry 2: 151.2
GPU Pixel Shaders: 91.6
64 BIT:
CPU: 265.1
GPU Geometry 1: 36.5
GPU Geometry 2: 149.4
GPU Pixel Shaders: 90.8
Hi all.
Barebone Notebook Clevo P751
i7 7700K
16GB 2400 DDR4
-gtx-1060
startup SSD 128 GB
Data 1TB SSD hybrid
windows 10
max latest 64bit
CPU: 280.7
GPU Geom 1: 272
GPU Geom 2: 324.7
GPU pixel shader: 301.8
another non-jitter-use computer:
Mac mini Server (Late 2012)
2,3 GHz Intel Core i7 (quadcore)
16 GB 1600 MHz DDR3
Intel HD Graphics 4000 1024 MB
OS X 10.10.5
max 734 64-bit
CPU: 188.1
GPU Geometry 1: 17.1
GPU Geometry 2: 85.1
GPU Pixel Shaders: 35.2
Hackintosh OS10.13/i7 6700k/32 GB RAM/AMD RX480/8G
CPU: 323.6
GPU Geometry 1: 60.7
GPU Geometry 2: 262.6
GPU Pixel Shaders: 313.4
Max 7.3.4 32-bit, for the above test
Hackintosh OS10.13/i7 6700k/32 GB RAM/AMD RX480/8G CPU:
CPU: 309.3
GPU Geometry 1: 60.5
GPU Geometry 2: 278.4
GPU Pixel Shaders: 315.1
Max 7.3.4 64-bit
All tests with Max 7.3.4 64 bits
Desktop : Core i7 4770, 16 Go RAM, Nvidia GTX 660 – 2 Go VRAM
Mac 10,11,6 (Hackintosh)
CPU: 189.0
GPU Geometry 1: 79.8
GPU Geometry 2: 189.7
GPU Pixel Shaders: 180.7
Windows 10 64bits Pro
CPU: 162.6
GPU Geometry 1: 179.0
GPU Geometry 2: 226.2
GPU Pixel Shaders: 234.1
Laptop : HP Zbook G3 Studio, Core i7 6700 HQ, 16 Go RAM, Windows 10 64bits Pro
Intel HD graphics 530 :
CPU: 183.9
GPU Geometry 1: 7.3
GPU Geometry 2: 143.4
GPU Pixel Shaders: 57.8
Nvidia Quadro M1000 – 4 Go VRAM
CPU: 182.0
GPU Geometry 1: 103.9
GPU Geometry 2: 456.9
GPU Pixel Shaders: 146.1
I would be very interested in a comparison between Sierra (Metal1) and High Sierra (Metal2).
@Herr There probably is no difference, as Jitter runs on OpenGL
@MrMaarten Thought this is kind of automatic and could even improve OpenGL software.
@ HERR MARKANT Metal is Apple's own 3D api (like DirectX is of Microsoft). It is a substitute for OpenGL, it does not interface with it...
Some more info:
https://www.youtube.com/watch?v=GZaQ--Xti-k
https://developer.apple.com/documentation/metal/a_brief_tour_of_metal
Some thoughts on this:
Microsoft DirectX (Direct3D) up to version 11 is comparable to the various versions of OpenGL, being that the former is exclusive to Microsoft products (Windows, XBox), and the latter is universally implemented (Windows, Mac, Linux, Android, etc).
Then came AMD's Mantle, a lower-level, lighter and potentially faster implementation of graphics drivers for its graphics cards, including multithreading, that paved the way to the "current generation" trend, letting the developer have more control over the hardware, but also having to program more of the graphics engine.
DirectX 12 followed this trend, and the Khronos Group, responsible for the OpenGL implementation, used most of the Mantle specification (now discontinued) and built Vulkan with cross-vendor (AMD, NVidia) and cross-platform support. Unfortunately, to this day, Apple still hasn't implemented Vulkan in Mac OS X and iOS, choosing instead to build Metal from scratch. So, right now, there is not a single universally usable modern graphics API:
Windows: DirectX12, Vulkan
Linux: Vulkan
Mac: Metal
iOS: Metal
Android: Vulkan
Which brings me to the question: what can a company with a cross-platform product like Cycling74 do? I would guess these facts are a strong reason why we're still stuck with OpenGL and not Metal/Vulkan:
I'm certain that it is not a trivial task to port Jitter to something other than OpenGL, seeing as it relies so much on its built-in features.
Cycling74 would have to develop much more of the graphics pipeline, instead of relying so much on the OpenGL API features, because newer APIs lack it (leave its implementation to the developers).
On top of this additional work, Cycling74 would have to implement it twice, Metal for Mac and Vulkan or DirectX for Windows!
I think there's not an easy answer to this. The only solution would be to do it and abstract the specific implementation of each platform to the user, sort of like in a game engine such as Unreal. And that is not a simple task at all...
So... OpenGL is dead! Long live OpenGL!
Thanks for all the infos, slowly i understand it (i really thought OpenGL is a part of Metal). And maybe there is still hope for Vulkan (German for vulcano) in OS X, because Apple is a member of the Khronos Group.
Late 2011 Macbook Pro
2.8ghz intel core i7
500 GB SSD - 840 evo SATA
16GB RAM
Intel HD Graphics 3000 512 MB
CPU: 96.5
GPU Geometry 1: 8.2
GPU Geometry 2: 29.9
GPU Pixel Shaders: 15.0
Blackmagic eGPU test results using Jitter_Benchmark-v1.0121.
I was expecting better then this?
Max 7.3.5 64 bit
MacBook Pro (15-inch, 2017)
OSX 10.13.6
3.1 GHz Intel Core i7
16 GB 2133 MHz LPDDR3
Blackmagic eGPU @ Radeon Pro 580 8192 MB GPU
CPU: 240.4
GPU Geometry 1: 38.8
GPU Geometry 2: 59.0
GPU Pixel Shaders: 206.6
Built-in Radeon Pro 560 4096 MB GPU
CPU: 162.5
GPU Geometry 1: 37.1
GPU Geometry 2: 115.8
GPU Pixel Shaders: 125.7