Reason 8.1 & above CPU Stress Test (songfile included)!

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
Post Reply
User avatar
dagwood
Posts: 46
Joined: 17 Jun 2016

09 Aug 2016

devilfish wrote:My new computer is ready and it runs through the whole song :mrgreen: :lol:
Attached are 2 photos. It is an Intel Xeon 2683v3 (Engineering Sample)CPU
My boyfriend wants to build me a computer with one of these, but I'm cautious. He says he can get an "engineering sample" off of eBay for $200-$300 USD, but this is a processor that normally sells for over $2000 USD!!! I have reservations about this.

What are your thoughts on this? What made you risk the engineering sample (presumably from eBay as well)?

User avatar
devilfish
Posts: 183
Joined: 20 Jan 2015

10 Aug 2016

No issues with my Engineering CPU. Everything works great. The QFQK is an Qualitiy Sample with same Clockspeeds and Spezifications like the Original CPU.
Example:
2x 3GHz (2 Core Turbo)+ 12 x2,5GHz
or
14x 2,7GHz (all Core Turbo)

Reason works very well on Multicore Systems. And for me there is nothing wrong with that kind of CPU :D
But it is not legal to sell one.

If your friend can get one --> buy one! ;)

Attantion to the Stepping!! The QFQK is C0/C1 and not all motherboards can run that Stepping!!!!!!!!!
You really have to Check your motherboard!!!

Here is a Link to my motherboard CPU Support List:
http://www.asrock.com/mb/Intel/X99%20Ex ... sp?cat=CPU

You can see it is nearly possible to run most of the Engineering CPUs.

For example, a Gigabyte motherboard can not run the Stepping C1/C0 nor the E5-2683v3 !!!
http://www.gigabyte.us/support-download ... x?pid=5125

The MSi Raider X99 can only run the Original CPU! But can not run C0 Stepping on a E5-2683v3 (Stepping: QEY7).
https://de.msi.com/Motherboard/support/ ... upport-cpu

So, at First step: Watch your Mainboard CPU Support List !!!

Further Information:
QEY7 = very First Stepping C0 (Engineering Sample) and CPUz show the CPU just as Xeon E5 Genuie Intel
http://i.ebayimg.com/images/i/152173458 ... -l1000.jpg
Also the clockspeed is reduced (-200MHz)
And Turbo Speeds are reduced!!

QFQK = 2nd Stepping C0/C1 (Quality Engineering Sample) and CPUz Show the CPU as Xeon E5-2683v3 (ES) - Same clockspeed like Original CPU

QGN5 = Last Stepping C1 - CPU with same Stepping like SR1XH (Original E5-2683v3)

SR1XH = Original Intel Xeon E5-2683v3 (Stepping C1)

User avatar
dagwood
Posts: 46
Joined: 17 Jun 2016

14 Aug 2016

devilfish wrote:No issues with my Engineering CPU. Everything works great. The QFQK is an Qualitiy Sample with same Clockspeeds and Spezifications like the Original CPU.
Example:
2x 3GHz (2 Core Turbo)+ 12 x2,5GHz
or
14x 2,7GHz (all Core Turbo)

Reason works very well on Multicore Systems. And for me there is nothing wrong with that kind of CPU :D
But it is not legal to sell one.

If your friend can get one --> buy one! ;)

Attantion to the Stepping!! The QFQK is C0/C1 and not all motherboards can run that Stepping!!!!!!!!!
You really have to Check your motherboard!!!

Here is a Link to my motherboard CPU Support List:
http://www.asrock.com/mb/Intel/X99%20Ex ... sp?cat=CPU

You can see it is nearly possible to run most of the Engineering CPUs.

For example, a Gigabyte motherboard can not run the Stepping C1/C0 nor the E5-2683v3 !!!
http://www.gigabyte.us/support-download ... x?pid=5125

The MSi Raider X99 can only run the Original CPU! But can not run C0 Stepping on a E5-2683v3 (Stepping: QEY7).
https://de.msi.com/Motherboard/support/ ... upport-cpu

So, at First step: Watch your Mainboard CPU Support List !!!

Further Information:
QEY7 = very First Stepping C0 (Engineering Sample) and CPUz show the CPU just as Xeon E5 Genuie Intel
http://i.ebayimg.com/images/i/152173458 ... -l1000.jpg
Also the clockspeed is reduced (-200MHz)
And Turbo Speeds are reduced!!

QFQK = 2nd Stepping C0/C1 (Quality Engineering Sample) and CPUz Show the CPU as Xeon E5-2683v3 (ES) - Same clockspeed like Original CPU

QGN5 = Last Stepping C1 - CPU with same Stepping like SR1XH (Original E5-2683v3)

SR1XH = Original Intel Xeon E5-2683v3 (Stepping C1)
Good information, thanks.

Yeah - I'm still reluctant to commit to this. After reading up, I've discovered that Engineering Samples do, in fact, fail a lot faster than their retail counterparts. Dave from Techpowerup.com (regular reviewer of CPU and motherboards) has many engineering sample CPU's in his possession and finds that one's results are hit and miss.

He states:
..these chips are likely to be abused while also NOT being some of the best chips ever, contrary to popular belief. ES CPUs are typically just the same as any other chip. I have 3x 4960X, and of the three, only one was "good", and the best it can do is 4.5 GHz (pretty average). Intel sends out chips that are good and bad, and average, since those that get ES chips do need to be aware of how these chips might work in general, and only select reviewers are likely to get good chips (those that benchmark). Meanwhile, I am pretty sure that using an ES chip to post benchmarks on HWBOT puts you outside of the general user category.
And finally:
ES chips tend to degrade quickly. It is like there is some sort of timer on them, and after so many hours, they start to fail, then clock worse, then don't work at all, and I have had this happen to many ES CPUs over the years.
So, great that you got a good deal on that chip, but I don't want to have problems down the line.

Good luck!

His original post:

http://www.techpowerup.com/forums/threa ... st-3372113

User avatar
devilfish
Posts: 183
Joined: 20 Jan 2015

17 Aug 2016

State:

Xeons are not Build to do Overclocking

Finally:

A Timer would take to much efford (and costs) to Build it into a waver.. I bed there is not one Single Chip in this world with a Build in Timer *lol*

My opinion:
That dude can't say: hey buy an ES Chips, if you can get one. Because it is not legal. And Intel would not send more chips to him! You know, he has to say that, and nothing else more. And really not that those Chips are perfect working ;)

Mine is working. Cool working 26-42 Celsius with prime95. No Crashes, nothing.. It is a Perfect working CPU.

Ostermilk
Posts: 1535
Joined: 15 Jan 2015

17 Sep 2016

Budget custom Skylake build. @£450

Reason 9.1
i5 6600k @ 4.3Ghz
Asus Z170 Pro Gaming
16Mb Crucial DDR4 2,133
Onboard Intel HD 530 Video

Audio Subsystem: Roland UA55 Quad Capture through ASIOLink Pro.

44,1k, 1024 samples buffer, 95% CPU

@53 seconds avg.

With a max CPU core temp during the test at cool 42C it's easy to keep quiet, moreover it's overall about 25% better than the 5y/o i7 2600k it replaced which is still kicking mild amounts of butt and still a great machine to have in reserve

Bigsby
Posts: 61
Joined: 21 Jan 2015

29 Nov 2016

Intel Xeon E5-1650 0 @ 3.20GHz 6 Cores
32 GB
Balance
44100 / 4096 samples
playback stops @ 54 seconds/28bars

This is a $350 HP z420 from craigslist. Much better than the one bar of payback I get on my 2011 Macbook Pro!

slightlyprog
Posts: 122
Joined: 02 Jan 2016
Location: Kent coast UK
Contact:

01 Dec 2016

i5 3570K 3.4ghz (not overclocked) on Asus P8Z77 V-LX mobo, win7 pro
16GB
Focusrite Scarlette 2i2
44.1 / 884 cpu at 95%

stops at 32.774 seconds

I think this thread has answered my question as to whether it would be worth it to go up to an i7 3770K, which I think is the fastest I could get for this (socket 1155) mobo. It doesn't appear to be worth it.

Most of the time it plays the stuff I am working on without any issues and on some songs I just need to go through and clean up devices and tracks that are not being used.

User avatar
Bonkhead
Posts: 335
Joined: 18 Jan 2015

16 Dec 2016

slightlyprog wrote:i5 3570K

stops at 32.774 seconds
i7 3770 doesn't appear to be worth the upgrade.
Indeed:

i7 3770 @ 3,4 ghz (non-K version, so not overclocked)
16 GB corsair vengeance at 1333 mhz (2x8gb)
Line 6 KB 37 asio @ 44.1 khz 24 bit and 4096 sample buffersize stops at 41,703 seconds
Propellerhead Balance asio @ 44.1khz 24 bit and 4096 buffersize stops at 42,576 seconds

Edit: I just found out I was running my ram on 1333 mhz while they're 1600 mhz.
Got a 10-15% better benchmark ram score on 1600 mhz but it didn't change the Reason test song result.

Manomars
Posts: 3
Joined: 15 Dec 2016

17 Dec 2016

53 sec
i7 980x
24 gyg ram UMC 404

$ 450 Hp z800 2x xeon 5660 48 gyg ram rme 9652
Plays full song without a glitch.
123 seconds on 48 Khz 128 buffer

EdGrip
Posts: 2348
Joined: 03 Jun 2016

17 Dec 2016

2013 Dell Precision M6800 workstation laptop

Your CPU model
Intel i7 4810MQ 2.8GHz - 3.8GHz Turbo
The amount of RAM you have
32Gb
Your audiocard Brand and Model
Internal RealTek, Steinberg UR22 for audio interface stuff
Your audiocard driver settings (sample rate 44100 Hz recommended and max buffer settings)
RealTek ASIO - 44.1kHz, 384 samples*

After a few goes, with the wifi switched off, it averages around 31.5 seconds with the RealTek ASIO and 29 seconds with the Steinberg ASIO.
It's using 3 of 4 cores, and the ctrl-shift-alt-+ thing seems to have no effect.

In Reason 8, I used to be able to adjust the buffer size with the ASIO drivers. Now neither the Yamaha/Steinberg ASIO driver, nor the RealTek ASIO driver, will let me change the buffer size. What's that all about, then?

User avatar
Bonkhead
Posts: 335
Joined: 18 Jan 2015

17 Dec 2016

Just playing around with turning real cores off/on i7 3770 using
EnochLight wrote:If the processor really has two or more nearly idling physical cores, you may be able to increase/decrease the number of threads in use for audio rendering, by pressing the following, unsupported and esoteric keyboard shortcuts in Reason:

To increase: Ctrl + Shift + Alt + <numpad-plus>
To decrease: Ctrl + Shift + Alt + <numpad-minus>
1 core = 6,409 sec
2 cores = 27,028
3 cores = 42.050
4 cores = 59.350 !!!!! (it started stuttering at 52 sec)

So a huge improvement turning on all 4 cores for use with Reason, instead of the automaticly 3 cores on the i7

Edgrip> did you check the windows task manager performance monitor to see how many cores were used during the song ?

EdGrip
Posts: 2348
Joined: 03 Jun 2016

17 Dec 2016

I did. I couldn't get the shortcut to change anything, but I'm probably missing something. I'm not too bothered.

I found the buffer settings - I had not spotted the "Control Panel" button, cunningly obviously hidden right there just below the audio driver settings. Changing the buffer settings didn't make any significant difference.

I think if I were looking for a workstation laptop today, I'd be looking at MSI over Dell - as silly as it sounds, mainly because the MSI computers still have a DVD drive. I find I still use the DVD drive a surprising amount.

Manomars
Posts: 3
Joined: 15 Dec 2016

17 Dec 2016

EdGrip wrote:I did. I couldn't get the shortcut to change anything, but I'm probably missing something. I'm not too bothered.

I found the buffer settings - I had not spotted the "Control Panel" button, cunningly obviously hidden right there just below the audio driver settings. Changing the buffer settings didn't make any significant difference.

I think if I were looking for a workstation laptop today, I'd be looking at MSI over Dell - as silly as it sounds.
Would advise you HP.
But realy, i prefer desktop for audio.

EdGrip
Posts: 2348
Joined: 03 Jun 2016

17 Dec 2016

Absolutely - just not practical for me at the moment.

User avatar
Bonkhead
Posts: 335
Joined: 18 Jan 2015

24 Dec 2016

I5 3570k
8 gb ram
Win 7 64

33.680 sec

Pretty standard result for this cpu.

jae1atm
Posts: 1
Joined: 24 Dec 2016

24 Dec 2016

Played full song with 2 dsp bars.



CPU usage limit set in Reason 8 = 95%
Sample rate = 44,100 Hz
Buffer length = 1024 samples (max )

Hardware:
CPU i7 6950X @ 3.0 ghz
RAM 16 GB
SSD Kignston 120 GB
UA Apollo Twin Solo

User avatar
Bonkhead
Posts: 335
Joined: 18 Jan 2015

24 Dec 2016

I'm making a list of all the results, the one thing missing is the actual score of the intel i7 5960x because the posted result was made using the beta version of reason 8.3, not a real release version. Anybody care to test that if possible ?

I suspect it would run the whole song, the 5930x did it, but need to be sure.

Also, since the original testsong is almost 2 years old, maybe make a newer LONGER one (just copy paste the 8.3 song to 2 minutes ?) so we can see how far the top 6 (E5's, i7 x's) go which all passed the whole 8.3 song ? Ofcourse in a new thread :thumbs_up:

dana
Posts: 335
Joined: 29 Apr 2015
Contact:

02 Jan 2017

Bonkhead wrote:I'm making a list of all the results, the one thing missing is the actual score of the intel i7 5960x because the posted result was made using the beta version of reason 8.3, not a real release version. Anybody care to test that if possible ?

I suspect it would run the whole song, the 5930x did it, but need to be sure.

Also, since the original testsong is almost 2 years old, maybe make a newer LONGER one (just copy paste the 8.3 song to 2 minutes ?) so we can see how far the top 6 (E5's, i7 x's) go which all passed the whole 8.3 song ? Ofcourse in a new thread :thumbs_up:
Ok, i have just built a 5960x rig i will test it :lol:

RESULTS ARE IN:

Full song with 4 DSP bars

CPU usage limit set in Reason 9.1 = 95%
Sample rate = 44,100 Hz
Buffer length = 256 samples (512/1024 samples=same result, 128=Crackles)

Hardware:
CPU i7-5960X @3.0Ghz (Running at stock speed, not overclocked)
OS: Windows 10 Pro
RAM: 32Gb DDR4
SSD 512Gb M.2 x4
MOTU 424 PCIe w/ 2408 mk2

its_JimmyJames
Posts: 1
Joined: 03 Jan 2017

03 Jan 2017

Hi guys,

Glad I found this forum, I've been having problems with Reason running slow, crackling etc... for a while now. Been in contact with Apple and Propellerhead about it. Even with small sessions, I have crackles and glitches all the time. Constantly having to up my buffer size to the max just to stay productive on songs I'm only just really getting started on.
I bought a new MacBook Pro Retina 15" just over a year ago and maxed it out on all options as I thought it would be more than enough for what I needed, but I cant work like this any longer. I need to find a solution, wether its changing my computer again or something else.

I wonder if someone can tell me wether the results I got from playing the stress test song file look normal for a computer of this spec?

MacBook Pro Retina 15 inch Mid 2015
Processor 2.8 GHz Intel Core i7
Memory 16 GB 1600 MHz DDR3

Buffer Size: 1024
Sample Rate: 44,100

Audio starts to crackle and pop from 16 seconds.
STOPS at Bar 21.1.4.186 / Time 40.472 seconds


This was done just at home with my MacBook Pro by itself. I use a Focusrite 18i20 audio interface in the studio and will try again once Im back there.
But I know that the performance I'm getting really isn't anywhere good enough at the moment so any suggestions of what I could do would be massively appreciated.

Jimmy

User avatar
Bonkhead
Posts: 335
Joined: 18 Jan 2015

03 Jan 2017

dana wrote: Ok, i have just built a 5960x rig i will test it :lol:

RESULTS ARE IN:

Full song with 4 DSP bars
Thanks !

User avatar
Bonkhead
Posts: 335
Joined: 18 Jan 2015

03 Jan 2017

This is the list until now. I'm very curious how the upcoming AMD ZEN cpu's will perform, since they seem to focus on more cores.
Attachments
2017-01-03 17_46_26-reasoncpus.ods - OpenOffice Calc.png
2017-01-03 17_46_26-reasoncpus.ods - OpenOffice Calc.png (49.45 KiB) Viewed 3015 times

User avatar
Kategra
Posts: 327
Joined: 18 Jan 2015

04 Jan 2017

Bonkhead wrote:This is the list until now. I'm very curious how the upcoming AMD ZEN cpu's will perform, since they seem to focus on more cores.
This list is so cool! Can't wait for AMD's Ryzen price and trusty reviews :D

User avatar
Kategra
Posts: 327
Joined: 18 Jan 2015

04 Jan 2017

its_JimmyJames wrote:Hi guys,

Glad I found this forum, I've been having problems with Reason running slow, crackling etc.. for a while now. Been in contact with Apple and Propellerhead about it. Even with small sessions, I have crackles and glitches all the time. Constantly having to up my buffer size to the max just to stay productive on songs I'm only just really getting started on.
I bought a new MacBook Pro Retina 15" just over a year ago and maxed it out on all options as I thought it would be more than enough for what I needed, but I cant work like this any longer. I need to find a solution, wether its changing my computer again or something else.

I wonder if someone can tell me wether the results I got from playing the stress test song file look normal for a computer of this spec?

MacBook Pro Retina 15 inch Mid 2015
Processor 2.8 GHz Intel Core i7
Memory 16 GB 1600 MHz DDR3

Buffer Size: 1024
Sample Rate: 44,100

Audio starts to crackle and pop from 16 seconds.
STOPS at Bar 21.1.4.186 / Time 40.472 seconds


This was done just at home with my MacBook Pro by itself. I use a Focusrite 18i20 audio interface in the studio and will try again once Im back there.
But I know that the performance I'm getting really isn't anywhere good enough at the moment so any suggestions of what I could do would be massively appreciated.

Jimmy
40 seconds is good for a laptop! If you really need more power, you have to get a computer with better CPU. 16 GB RAM is enough.
You can see from all the results in this thread that PCs built with i7 CPU on socket LGA 2011 give the best performance in Reason so far. If you only want Apple computers, I think those are found in the MAC PRO only, while on PC you can either buy a full system, or build it yourself for a fraction of the cost.

Also, please keep in mind that the benchmark results shouldn't be interpreted as extremely accurate in measuring the Reason performance of a computer.

What I mean: a computer that can play 60 seconds of the benchmark is not exactly 2 times more powerful than a computer who can play only 30 seconds. This is because all the idle devices in the rack still consume a not so negligible amount of CPU power even if they are not playing notes/processing sound.

A wisely configured desktop PC ~ $2000-$2500 will give you double the power of your current laptop and no mobility :)

User avatar
Re8et
Competition Winner
Posts: 1514
Joined: 14 Nov 2016

14 Jan 2017

Easy as a cake!!

New build pc:

Win 10 Xeon 2650v4 @ 2.2Ghz NO OC
32gb Vengeance 2400 DDR4
Random video card from AMD

Audio Motu ASIO MkII Ultralite Firewire 256 samples *smallest, cpu set at 95%

Songs goes all the way through the end as a piece of cake. CPU never above 60%
11 out of 12 real cores crunching on!

I wouldn't add a lot of effects to this song anyway.

CPU bench Math Floating Point 13999 with Passmark, 18600 Multi and 2160 single with geekbench.

User avatar
NaviRetlav
Posts: 448
Joined: 26 Jan 2015

18 Jan 2017

CPU: 2x Xeon E5-2670
Soundcard: Tascam US20x20

CPU limit 95% -------- playback doesn't stop or glitch at all. It went through this demo song like through butter ;)
Reason benchmark TEST.jpg
Proof
Reason benchmark TEST.jpg (563.52 KiB) Viewed 2904 times

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 24 guests