Go Back   Cockos Incorporated Forums > REAPER Forums > REAPER General Discussion Forum

Reply
 
Thread Tools Display Modes
Old 04-25-2017, 01:33 AM   #41
C-H
Human being with feelings
 
Join Date: Apr 2015
Location: Tucson, AZ
Posts: 373
Default

Why wouldn't a AMD RYZEN be good for a DAW?
I don't understand reasoning here.
  • Its single-core performance is just a hair shy of Intel.
  • Most of them overclock to almost 4.0GHz (3.8GHz should be easy).
  • A 4C/8T RYZEN is cheaper than it's Intel counterpart and even a 6C/12T RYZEN cost less.
  • An 8C/16T RYZEN is still below $500.

So what gives?
Is it driver quality or just lack of support for Audio specific hardware?

I looking into building a new rig this year but I'm holding off until mini-ITX boards become readily available for the RYZEN platform. There are 1 or 2 out now based on the B350 and X370 chipsets. The B350 is the one to go for here I think, since it supports overclocking.
The X370 platform is more for gamers and heavy graphics users since it supports dual GPU.
__________________
Stash : M-Audio Oxygen 49 KeyMap
Stash : M-Audio Oxygen 49 v2 KeyMap
C-H is offline   Reply With Quote
Old 04-25-2017, 02:59 AM   #42
EvilDragon
Human being with feelings
 
EvilDragon's Avatar
 
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
Default

Low latency performance seems to be the issue currently.

http://vi-control.net/community/thre...4#post-4079064

http://vi-control.net/community/thre...4#post-4079889

Last edited by EvilDragon; 04-25-2017 at 03:10 AM.
EvilDragon is offline   Reply With Quote
Old 04-25-2017, 03:22 AM   #43
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

I think a lot of the naysayers are basing their opinions on brand loyalty to Intel, and the 'mindshare' Intel has over much of the consumer base. AMD wasn't competitive for a long while, and the performance just wasn't there compared to the Intel chips. But Ryzen is finally here and by pretty much all accounts it competes very well, which can only be a good thing for the consumer.

It is early days though and there are some complications; the dual CCX design of the Ryzen silicon is a bit different to the way the Intel chips are put together. This split design with the two CCXs linked by AMD's "Infinity Fabric" is thought to be what has caused the oddly low performance in some games. If the software/game engine is designed in such a way to minimise cross-CCX traffic then the performance defecit can be mitigated. This is why I asked if the REAPER devs have looked into Ryzen at all.

See here for some interesting testing on the subject - https://www.pcper.com/reviews/Proces...-Silver-Bullet
There's a video there explaining it all if that may be easier to digest. This was a little while ago and it's all quite fast moving still, so there might be some even more up-to-date information now.

I think this probably doesn't affect DAW performance for most people though, the greater number of cores and threads compared to similarly priced Intel chips should make them much better for doing big productions with many tracks. Perhaps the design might cause issues at low latencies as shown here - http://www.scanproaudio.info/2017/03...ook-for-audio/ (still should be at least as good as an i5 though..)
But there hasn't been enough testing on this so it remains to be seen how much if any noticeable effect this has on real world performance in music production.
Gurt Tractor is offline   Reply With Quote
Old 04-25-2017, 03:38 AM   #44
EvilDragon
Human being with feelings
 
EvilDragon's Avatar
 
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
Default

Dawbench is easy to try out, and as linked above shortcomings at low latencies do exist with Ryzen...

No brand loyalty there, just pure numbers of how many ReaXcomp instances can be ran before CPU starts crapping out.

What's interesting to me is that Ryzen craps out the audio way below 90% CPU usage, which is very weird, and that would definitely be very noticeable in big DAW projects, no matter which plugins are used. If you can't tax the CPU all the way (as noted by Jim above, socket 2011 CPUs don't crap out till they hit 99%, which is what you'd kinda want for big productions, no?), what good is it?
EvilDragon is offline   Reply With Quote
Old 04-25-2017, 03:52 AM   #45
snooks
Banned
 
Join Date: Sep 2015
Posts: 1,650
Default

We'll see what the follow up results are though, there's a discrepancy between JR's results and Scan's at the moment. But from the current figures available we can say that even at the lowest latencies, an 1700x @ 3.8GHz at least roughly matches the performance of an i7-7700k and at best seriously outperforms it, keeping up with the i7-6900k.

Since the 1700x and the i7-7700k are the same price, I don't see what's bad about that. Especially since the worst case scenarios will undoubtedly be improved upon and at higher buffer sizes and offline rendering etc it has the raw power advantage.
snooks is offline   Reply With Quote
Old 04-25-2017, 03:59 AM   #46
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

I think that even if the usage on the Ryzen cannot be maxed without glitching it doesn't really matter if it's still able to achieve more tracks and plugins than an Intel chip for the same price (such as Ryzen 7 vs i7 7700k; both are pretty much the same price(and the 1700 comes with stock cooler good enough to get it to 3.8-33.9Ghz comfortably)).

It's a shame there is not the same level and breadth of testing for this kind of stuff across the internet as there is for gaming, one publication's findings isn't enough to properly conclude anything IMO.

Seems like it's performing well enough in Fruity Loops at least, from here - http://www.overclock.net/t/1626661/a...production-dpc

Quote:
The Intel i7 4790k craps out @ about 97% cpu usage audio glitches buffer under-runs

The AMD Ryzen 1700 still running & with only 46% / 48% usage
So i am able to increase the track count to all most double with the 1700 until audio glitches buffer under-runs appear this was with cpu usage @85% so its a bit strange that it craps out with only being at 85%... i am hoping micro code bios updates will solve this issue ...

To add

The 7700k @ 4.8Ghz with the above settings is still going strong but at 91/93% cpu usage add a few more tracks and i hit 97/98% usage then we get audio glitches buffer under runs

So you can clearly see here with this Info that the 1700 for more a less the same cost as the I7's is the clear winner and bare in mind the ryzen is at stock with stock bios lame memory settings and far worse DPC latency so worse case setup out of the box and it shines ...i am very impressed with ryzen .

Clear winner as of 2/3/17 and improving is the 1700 non x

For me its a gift from AMD..which enables me to have around double the track count which means i can do away with my 3 DAW pc setup and use 2 x 1700 setup and have even more cpu grunt..

And with using less power from the wall while doing so

Yes things all-so improve once you clock up the ryzens to 3.7/3.9 ect
Interesting for what's to come with Ryzen maturing
And this guy is very happy with the upgrade from a 3770k to the 1800x -

https://www.youtube.com/watch?v=Iq7u-agaxlM
https://www.youtube.com/watch?v=xama-Np0yKo
Gurt Tractor is offline   Reply With Quote
Old 04-25-2017, 04:24 AM   #47
EvilDragon
Human being with feelings
 
EvilDragon's Avatar
 
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
Default

Quote:
Originally Posted by Gurt Tractor View Post
Seems like it's performing well enough in Fruity Loops at least, from here - http://www.overclock.net/t/1626661/a...production-dpc
Yeah... at buffer size 512.


Also, 14 Divas all with multicore option enabled is not always the best idea, since they start fighting for cores...


3770K is a 5 year old CPU right now. Quite a comparison there, of course it would be an upgrade... of sorts!
EvilDragon is offline   Reply With Quote
Old 04-25-2017, 04:32 AM   #48
Softsynth
Human being with feelings
 
Join Date: Jun 2015
Posts: 8,696
Default

Quote:
Originally Posted by Gurt Tractor View Post
And this guy is very happy with the upgrade from a 3770k to the 1800x -

https://www.youtube.com/watch?v=Iq7u-agaxlM
https://www.youtube.com/watch?v=xama-Np0yKo

Little wonder, 2012-2017!
A 2017 £500 processor vs circa £170 (now/used) 2012 CPU with DDR4 instead of DDR3 (and to a much lesser extent new mobo tech too).
You could get similar performance to that K model Intel CPU (if not comparing overclocked CPU) from a 3770 (£130).

Compare that price bracket to anything brand new from Intel or AMD it's a different story.

Great news that AMD is finally producing competitive product to challenge Intel after so long. A monopoly is never healthy for any market.
Softsynth is offline   Reply With Quote
Old 04-25-2017, 04:43 AM   #49
somebodyelseuk
Human being with feelings
 
somebodyelseuk's Avatar
 
Join Date: Jun 2011
Location: Birmingham, UK
Posts: 1,125
Default

Brand loyalty? Not here. I've been using AMD exclusively for the last ten years, BUT this is my hobby, my 'work' doesn't require firespitting performance and I don't have deep enough pockets to justify the cost of Intel's CPUs.
However, where a person asking my advice WHO DOES THIS FOR A LIVING (if they're asking on an internet forum, they almost certainly don't), then I would ALWAYS advise they leave the new unproven gear alone until such time it's got a track record.

You can transport parcels around in a Vauxhall Zafira, but there's a good reason why most carriers use Ford Transit vans.
__________________
"As long as I stay between the sun & my shadow, I guess I'm doing well."
somebodyelseuk is offline   Reply With Quote
Old 04-25-2017, 05:41 AM   #50
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

There's actually surprisingly little performance difference between Intel's 5 year old Ivy Lake architechture and the latest Kaby Lake chips. The latest is better sure but there's not really that much reason to upgrade a 3770k that can clock to 4.4-4.7 Ghz to a 7700k that will go to 4.8-5.1 (with a really good cooler as the CPU IHS isn't soldered (unlike Ryzen)). Kaby Lake has the exact same IPC (instructions per clock) as Skylake, so there was no improvement at all from that generation to this one except in clock speeds.

The percentage increase from the 3770k to the 7700k is about 25% if you measure at the same clock speeds, so maybe 30-35% improvement at most with Kaby Lake's higher clocks. The upgrade and switch to the Ryzen platform was a much better upgrade for him, allowing far more tracks running at once.

I think it's fairly clear that if you are running a professional recording studio then low-latency performance and stability is going to be of paramount importance to you, and in this case going AMD is certainly a bit riskier, and sticking with your current Intel system or buying the 7700k (Broadwell-E is horrible value unless you have the money to burn) might be the best option for now. But for the average hobbyist of which there are far more out there that are looking for a good CPU, the AM4 platform and Ryzen CPUs seem to be a really good option. The paint is still drying on the new platform as a whole, but from all I've read the good reports have far outweighed the bad.

The Ryzen 5 1600 looks to be the absolute best 'bang for the buck' for this purpose, around £210 for a 6 core 12 thread CPU compared to a 4 core 4 thread overclockable i5 7600k. That is some incredible value right there...

Last edited by Gurt Tractor; 04-25-2017 at 06:17 AM.
Gurt Tractor is offline   Reply With Quote
Old 04-25-2017, 06:39 AM   #51
EvilDragon
Human being with feelings
 
EvilDragon's Avatar
 
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
Default

Quote:
Originally Posted by Gurt Tractor View Post
so there was no improvement at all from that generation to this one except in clock speeds.
And power consumption. That's a biggie, and where most of differences between 3770k and latest Kabies come from. This is not something that should be neglected.
EvilDragon is offline   Reply With Quote
Old 04-25-2017, 07:04 AM   #52
Softsynth
Human being with feelings
 
Join Date: Jun 2015
Posts: 8,696
Default

Quote:
Originally Posted by Gurt Tractor View Post
There's actually surprisingly little performance difference between Intel's 5 year old Ivy Lake architechture and the latest Kaby Lake chips. The latest is better sure but there's not really that much reason to upgrade a 3770k that can clock to 4.4-4.7 Ghz to a 7700k that will go to 4.8-5.1 (with a really good cooler as the CPU IHS isn't soldered (unlike Ryzen)). Kaby Lake has the exact same IPC (instructions per clock) as Skylake, so there was no improvement at all from that generation to this one except in clock speeds.

The percentage increase from the 3770k to the 7700k is about 25% if you measure at the same clock speeds, so maybe 30-35% improvement at most with Kaby Lake's higher clocks. The upgrade and switch to the Ryzen platform was a much better upgrade for him, allowing far more tracks running at once.

I think it's fairly clear that if you are running a professional recording studio then low-latency performance and stability is going to be of paramount importance to you, and in this case going AMD is certainly a bit riskier, and sticking with your current Intel system or buying the 7700k (Broadwell-E is horrible value unless you have the money to burn) might be the best option for now. But for the average hobbyist of which there are far more out there that are looking for a good CPU, the AM4 platform and Ryzen CPUs seem to be a really good option. The paint is still drying on the new platform as a whole, but from all I've read the good reports have far outweighed the bad.

The Ryzen 5 1600 looks to be the absolute best 'bang for the buck' for this purpose, around £210 for a 6 core 12 thread CPU compared to a 4 core 4 thread overclockable i5 7600k. That is some incredible value right there...
Yes it isn't a massive difference, incremental performance upgrades with better efficiency. A bit of a yawn for the desktop market.
Certainly when you compare second hand price for i5/i7 vs i3 or lower end new i5 or any AMD buying new makes less sense for the hobbyist not going all out for latest top of the range performance.

That is why two years ago I built an i5 system (up until that point I had nearly always bought new) out of a second hand i5 2400 and a new unused 2012/13 spec mobo with new memory, and I recently upgraded to used i7 3770.
Comparing floating point performance and seeing performance in multiple comparison charts highlights how stagnated CPU performance gains have become. The i5 2400 is still very respectable.

The hope is that AMD will reinvigorate the market with aggressive pricing and new tech.

Power consumption is mainly of benefit to portable laptop design. The cost of running a 3770 versus a new i7 isn't going to figure in most people's plans.
Lower power requirements does mean less heat and the potential to run fans even slower though. However those i5/i7 from 2011/12 are cool running and can be run with slow quiet fans anyway, so unless you desperately need really small form factor.........

If I was building a brand new system now instead of having upgraded two years ago I would look seriously at AMD too.

Last edited by Softsynth; 04-25-2017 at 07:10 AM.
Softsynth is offline   Reply With Quote
Old 04-25-2017, 07:06 AM   #53
EvilDragon
Human being with feelings
 
EvilDragon's Avatar
 
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
Default

I don't know about you lot, but power efficiency is kind of a biggie IMHO. It means the CPU dissipates less heat while doing the same amount of work, which means you don't need to have as beefy PSU, and an airplane turbine for a cooler... My i7-6700K is air cooled, OCd to 4.4 GHz, and it runs really quiet. Win all around, as far as I'm concerned.
EvilDragon is offline   Reply With Quote
Old 04-25-2017, 07:09 AM   #54
alanofoz
Human being with feelings
 
alanofoz's Avatar
 
Join Date: Sep 2009
Location: Oz - Blue Mountains NSW, formerly Geelong
Posts: 943
Default

Interesting to see how much more efficient these chips have become over the years. Easy to get too - you just buy them from your local efficient chip shop...
.
.
.
I can hear the groans from here...
__________________
It's "its" except when it's "it is".

alanofoz, aka Alan of Australia
alanofoz is offline   Reply With Quote
Old 04-25-2017, 07:10 AM   #55
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

I was referring to the difference between the i7 6700k which was released back in 2015 to the 7700k which was released at the start of this year.



Ryzen is actually about on par or better than Intel on power consumption -


Last edited by Gurt Tractor; 04-25-2017 at 07:14 AM. Reason: Image posting troubles :)
Gurt Tractor is offline   Reply With Quote
Old 04-25-2017, 10:05 AM   #56
airon
Human being with feelings
 
airon's Avatar
 
Join Date: Aug 2006
Location: Berlin
Posts: 11,817
Default

That's a big difference.

If and when I need more power, the Ryzen platform is on my check list.
__________________
Using Latch Preview (Video) - Faderport 16 setup for CSI 1.1 , CSI 3.10
Website
"My ego comes pre-shrunk" - Randy Thom
airon is offline   Reply With Quote
Old 04-27-2017, 06:38 PM   #57
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

Hullo again, I tried out the little Diva CPU test project that SmajjL made in the Best CPU thread over in the compatibility section, using a 1700x @3.85Ghz and 16GB RAM @2666Mhz. - https://stash.reaper.fm/30216/CPU-Test.rpp

Using the demo version of the latest Diva VST instrument, I could duplicate the tracks up to a total of 24 before I got crackling during playback. This was when my CPU was just about getting maxed out in the task manager. That was with the minimum latency setting on the Cambridge Audio Dacmagic XS that I happened to be using, this was 80 samples and 2.8m/s as shown in Reaper. I also tried using a Lexicon Alpha with Asio4All drivers and had pretty much the same results. I couldn't really go further than 24 tracks with any latency settings I tried.

Just tried the same but with 64 samples latency on the Dacmagic, managed 25 tracks, but that was at the absolute limit, 24/23 was more sensible.

Comparing this to SmajjL's 7700k performance (don't know what clock speed) he got up to 20 tracks at 64 samples. So they are both quite similar.

Don't know how good a real world test it is but it would be interesting to see some comparisons if anyone else would like to have a go with their CPU.
Gurt Tractor is offline   Reply With Quote
Old 04-27-2017, 07:02 PM   #58
Lawrence
Human being with feelings
 
Join Date: Mar 2007
Posts: 21,551
Default

Not sure about AMD "blowing" anything.

We would be remiss not to remember that people who use daws are just grain of sand on the beach of personal computing. AMD will be just fine.
Lawrence is offline   Reply With Quote
Old 04-28-2017, 12:11 AM   #59
Win Conway
Human being with feelings
 
Join Date: Dec 2010
Posts: 3,826
Default

Not sure why you are even mentioning "Blowing" the only person who said that was me, because i viewed the scan store and it seems there was 4 extra cores on the price compared systems, and this was pointed out about a week ago by Evil Dragon and others, are you not reading the entire thread or just reading my comment ?

I am still struggling to find a system that is massively cheaper than a comparable Intel right now, but i haven't priced up parts to build my own (Not building for a long while yet, need a bunch of other new gear first)
__________________
Stop posting huge images, smaller images or thumbnail, it's not rocket science!
Win Conway is offline   Reply With Quote
Old 04-28-2017, 02:00 AM   #60
snooks
Banned
 
Join Date: Sep 2015
Posts: 1,650
Default

@Gurt Tractor: could you run DAWbench at your lowest buffer size with your memory at different under/stock/over clocks to see what difference it makes for low latency audio?
snooks is offline   Reply With Quote
Old 04-28-2017, 03:30 AM   #61
Curtis
Banned
 
Join Date: Oct 2015
Posts: 708
Default

How about some one do test with reaper doing a set of tasks that show reaper performance , this would be something that reaper people could relate to.
Curtis is offline   Reply With Quote
Old 04-29-2017, 01:47 PM   #62
snooks
Banned
 
Join Date: Sep 2015
Posts: 1,650
Default

Quote:
Originally Posted by EvilDragon View Post
It looks like the rerun tests with higher memory speed (and 64 bit) have the Ryzen smoking the 6850k at 48 samples now. The remaining issues are down to BIOS/microcode improvements (inc memory compatibility). So it really does look as though the fundamentals of the chip are sound.
snooks is offline   Reply With Quote
Old 04-29-2017, 01:53 PM   #63
EvilDragon
Human being with feelings
 
EvilDragon's Avatar
 
Join Date: Jun 2009
Location: Croatia
Posts: 24,790
Default

The problem being there's only one mobo that can run RAM at 3200 MHz. And even then you cannot have more than 16 GB. Sucks...
EvilDragon is offline   Reply With Quote
Old 04-29-2017, 02:21 PM   #64
snooks
Banned
 
Join Date: Sep 2015
Posts: 1,650
Default

Yes, the supporting hardware/firmware situation needs to be addressed. I sense Intel coming out with an affordable 8 core to counter AMD. We'll see, but it's all good.
snooks is offline   Reply With Quote
Old 04-30-2017, 05:21 AM   #65
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

Quote:
Originally Posted by snooks View Post
@Gurt Tractor: could you run DAWbench at your lowest buffer size with your memory at different under/stock/over clocks to see what difference it makes for low latency audio?
Unfortunately DAWbench VI seems to need some Kontakt library stuff, including some legacy from 3/4, that I don't have. So I can't really run that. The other benchmark needs a bunch of trial versions of plugins that I don't really have the time or inclination to find and install. I can do some benchmarks for comparison but they need to be fairly easy and quick to run with not too many random trial VSTs that I will need to uninstall later.

There are definitely some issues currently with RAM compatibility, I think fixes and improvements are slowly rolling out, I read somewhere that it a software problem rather than an inherent hardware issue (hopefully). The best thing to do if you are building a new system is to check the motherboard manufacturer's site for a list of compatible RAM and configurations that have been tested.

Annoyingly DDR4 prices have gone through the roof recently, and might stay that way for a little while. So if you don't absolutely need 32Gb, then I would recommend sticking with 16 for the moment, and upgrading once prices have settled.
Gurt Tractor is offline   Reply With Quote
Old 04-30-2017, 10:15 AM   #66
snooks
Banned
 
Join Date: Sep 2015
Posts: 1,650
Default

Quote:
Originally Posted by Gurt Tractor View Post
Unfortunately DAWbench VI seems to need some Kontakt library stuff, including some legacy from 3/4, that I don't have. So I can't really run that. The other benchmark needs a bunch of trial versions of plugins that I don't really have the time or inclination to find and install. I can do some benchmarks for comparison but they need to be fairly easy and quick to run with not too many random trial VSTs that I will need to uninstall later.
The one that Scan and Studiocat have used is the ReaXcomp one - Dawbench-DSP-R4-RXC-EXT.RPP in the Reaper folder - from DAWbench 2014. It comes with ReaXcomp (32 and 64 bit VST versions) in the RXC Builds folder, so you just need to drag the 64 bit one into your VST folder. It only uses that one VST.
snooks is offline   Reply With Quote
Old 04-30-2017, 12:25 PM   #67
randall7
Human being with feelings
 
randall7's Avatar
 
Join Date: Jul 2014
Posts: 251
Default

Quote:
Originally Posted by Gurt Tractor View Post
I will be getting a Ryzen 7 1700x and Asrock X370 motherboard this week as an upgrade...
I will be interested as I too plan to build a Ryzen system to upgrade my existing AMD A10 cpu system.
__________________
Registered Reaper 6, AMD Ryzen 9 cpu, 128GB RAM, Windows 10 x64, MuseScore 3 and 4, VST libraries.
randall7 is offline   Reply With Quote
Old 05-04-2017, 07:47 AM   #68
Gurt Tractor
Human being with feelings
 
Join Date: Apr 2013
Posts: 75
Default

Quote:
Originally Posted by snooks View Post
@Gurt Tractor: could you run DAWbench at your lowest buffer size with your memory at different under/stock/over clocks to see what difference it makes for low latency audio?
Ok I just tested my 1700x at 3.85Ghz using 2133Mhz and 2666Mhz memory speeds with Dawbench.

With 2666Mhz I can comfortably hit 320-325 RXC instances running at 64 samples latency, dropping the memory speed down to 2133 seems to lead to more intermittent pops and crackles at similar numbers but not total garbling. After I bypass maybe 10-25 instances then it stabilises. So Memory speed does have an effect, though small, defintely worth trying to get it as fast as you can. Unfortunately even though my 16GB Corsair DDR4 kit is rated for 3000Mhz I cannot exceed 2666Mhz without BIOS post failure. Hopefully there will be some microcode updates from AMD soon to improve the compatibility. I'm quite happy with the performance currently though.

I am actually on a different motherboard now... I was on an Asrock X370 Fatal1ty Gaming K4, the fan speeds were controlled by the motherboard's CPU temperature sensor which seems to be based in or under the socket, which means there is a massive delay for the fans to get up to speed and control heat in the chip when it is put under full load. Pretty useless for overclocking, and is quite deceptive as I thought for ages that I was only getting around 55c with a stress test when it was probably getting towards 90c!! This could be offset by tweaking the fan curves once you gain knowledge of this, but it is really not ideal when you want a mostly silent system. I will be sending it back to the retailer soon, I absolutely would not recommend the board unless the issue is addressed in BIOS update at some point. I think the other Asrock boards in their lineup might be affected too, odd that no reviewers seemed to pick up on it.

I have instead gone with a Gigabyte AX370 Aorus Gaming K7 (I wish the naming was less cringy with these things -_-), the fan control is much much better out of the box, I haven't even felt compelled to change the default settings yet though I will soon so I can have a near passive system at idle and low loads. It seems to boot and shutdown a bit faster and has a dual BIOS which I have needed to use a few times thus far, very useful. Performance seems to be more consistent somehow too, the Cinebench scores are slighty higher on average than what I was getting with the Asrock board. It is of course more expensive, but so far I would recommend it.

I think if you are going to get an motherboard for Ryzen you might as well get a higher end X370 one with less compromises, or just a much cheaper decent B350 board. I felt like I needed the extra USB ports and other connectivity you get with X370. The middle ground doesn't really seem as worthwhile IMO.
Gurt Tractor is offline   Reply With Quote
Old 05-04-2017, 03:02 PM   #69
snooks
Banned
 
Join Date: Sep 2015
Posts: 1,650
Default

Very interesting, thanks very much for taking the time to do that. It would seem that the RXC test is not very dependant on memory speed, although there is still a measurable improvement.

I wonder if the artificial nature of the DAWbench VI test is causing over dependency on the Infinity Fabric, since the instances are sharing more memory intensive resources in the form of the same samples. Which isn't a reflection of real world usage. Not a big deal normally, but with this architecture it may be.

Setting up the DAWbench VI test without the Kontakt 3 library is a PITA though so I'm not even going to ask. If one of the DAW builders out there in the ether wants to test the effect of memory speed on that test it would be wonderful.
snooks is offline   Reply With Quote
Old 10-19-2017, 11:32 AM   #70
Symbiant
Human being with feelings
 
Join Date: Sep 2017
Posts: 6
Default

I just finished my build lastnight.

Ryzen 1600x
MSI Arctic Tomahawk B350 (because it’s on board Audio is one of the best)
TeamForce dark 16gb (8x2) 3000mhz
Phantek Eclipse P400 Case
Corsair H80i V2 with the AM4 bracket
Corsair 600x power supply
WD 250gb SSD boot
WD 1TB Samples Drive
WD 2TB Project Drive
EVGA GeForce GT 1030 SC LP 2GB


Haven’t over clocked it yet.
Symbiant is offline   Reply With Quote
Old 05-22-2018, 11:54 PM   #71
khushalbhadra
Human being with feelings
 
Join Date: Sep 2009
Posts: 86
Default

Quote:
Originally Posted by EvilDragon View Post
Intel system is using i7-7700K which is the same price as Ryzen, so there... Also, Ryzen doesn't have a GPU on board which i7-7700 has, so that's additional expenditure on AMD systems.

This would also be a part of the reason, I suppose:

http://www.digitaltrends.com/computi...-ryzen-launch/


i7-7700K has HALF the cores and threads of Ryzen 1700X (Ryzen sporting 8 cores and 16 threads, 7700K sporting 4 cores and 8 threads), so they're not even comparable. Cache is pretty different, too, Ryzen having almost 3 times as much. So, actually, Ryzen is more affordable one here, pound per core. Not sure where you saw that they have the same amount of cores, that's not true.
What are your thoughts on the new AMD Ryzen 2400G with AMD Vega graphics card. And would love to hear from people who have already build ryzen pc and performance using Reaper.
khushalbhadra is offline   Reply With Quote
Old 05-24-2018, 12:56 AM   #72
curtis_dean
Human being with feelings
 
Join Date: May 2018
Posts: 10
Default

I have been pricing Ryzen 7 - 1700 vs intel i7 - 7700, as of pricing of May 23, 2018, for a complete system (no DAW Equipment included) the intel system is 5,000 more than the AMD Ryzen system. there is no way in my mind I would be willing to pay that much extra for the intel system.
curtis_dean is offline   Reply With Quote
Old 05-25-2018, 11:28 AM   #73
lolilol1975
Human being with feelings
 
Join Date: Dec 2015
Posts: 1,739
Default

Quote:
Originally Posted by gpunk_w View Post
Aaah so the Ryzen i linked is way more powerful, double the cores for the same money uh, winder how that equates in real life.
The Ryzen is slightly above, but the single core perf of the 77OOK is 25% higher.
https://www.cpubenchmark.net/compare...00K/2969vs2874
So for Reaper I think the Intel would be better, unless you work with many tracks and few plugins per track (orchestral stuff for instance), where the 8 cores of the Ryzen can shine.
In any case, both these procs are super powerful and you'll have to do some pretty heavy stuff to bring them down to their knees.

Last edited by lolilol1975; 05-25-2018 at 11:39 AM.
lolilol1975 is offline   Reply With Quote
Old 05-26-2018, 04:48 AM   #74
khushalbhadra
Human being with feelings
 
Join Date: Sep 2009
Posts: 86
Default

Quote:
Originally Posted by Symbiant View Post
I just finished my build lastnight.

Ryzen 1600x
MSI Arctic Tomahawk B350 (because it’s on board Audio is one of the best)
TeamForce dark 16gb (8x2) 3000mhz
Phantek Eclipse P400 Case
Corsair H80i V2 with the AM4 bracket
Corsair 600x power supply
WD 250gb SSD boot
WD 1TB Samples Drive
WD 2TB Project Drive
EVGA GeForce GT 1030 SC LP 2GB


Haven’t over clocked it yet.
Hows the performance with respect to recording loading plugins in Reaper, also recording and latency? How many instances of VSTi could you load with your Ryzen Built? I too am planning to upgrade to an AMD Ryzen 2400G which includes Vega graphics chip.
khushalbhadra is offline   Reply With Quote
Old 05-26-2018, 05:34 AM   #75
Plazma
Human being with feelings
 
Join Date: May 2014
Posts: 293
Default

I am running on a Ryzen system - with no issues whatever.
I can load 5 instances of Omnisphere 2, bungloads of effects, and record live guitars, without really taxing my cpu.
Just make sure you buy enough RAM - 32gb or more would be nice.

You are totally safe with Ryzen CPU.
__________________
url]www.pennysound.co.za[/url] (Free & commercial Omnisphere patches)
Plazma is offline   Reply With Quote
Old 05-30-2018, 09:09 AM   #76
khushalbhadra
Human being with feelings
 
Join Date: Sep 2009
Posts: 86
Default

Quote:
Originally Posted by Plazma View Post
I am running on a Ryzen system - with no issues whatever.
I can load 5 instances of Omnisphere 2, bungloads of effects, and record live guitars, without really taxing my cpu.
Just make sure you buy enough RAM - 32gb or more would be nice.

You are totally safe with Ryzen CPU.
Thanks for the reply. Planning to buy the new AMD Ryzen 5 2400G with an integrated graphics card with 16GB of Ram and a 250 GB SSD. Hoping to get decent performance with respect to recordings. I use a lot of Kontakt plugins hope it can handle more instances than my old 2011 iMac. Planning to shift.
khushalbhadra is offline   Reply With Quote
Old 03-11-2019, 10:17 PM   #77
Montego
Human being with feelings
 
Join Date: Jun 2017
Posts: 11
Default

What sample block?

Quote:
Originally Posted by Plazma View Post
I am running on a Ryzen system - with no issues whatever.
I can load 5 instances of Omnisphere 2, bungloads of effects, and record live guitars, without really taxing my cpu.
Just make sure you buy enough RAM - 32gb or more would be nice.

You are totally safe with Ryzen CPU.
Montego is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 02:59 AM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.