Has anyone done a MkII vs Turbo comparison yet?

[MackieFX]

Seen a lot of people say they've bought one, but no comparison threads videos. Interested to see what the performance gain translate to in the real world

 

[Feflicker]

My bass kitchen sink preset went from 82% to 58%.

 

[TravisG406]

Here's my current kitchen sink - Ultra-High reverb and Full Res IR's in the IR player block. Out 1 is FOH, 2 is IEM, 3 is Xitone on stage

 

[HereToday]

This puppy only runs on a Turbo. For this shot, the Full Res IR's are loaded in the cab, but are off.

 

[vrozzi]

Isn't there some misunderstanding? 25% more speed does not equal to 25% more processing power. Turbo model the way it's described means 25% faster, more responsive unit while processing power is still the same 2 CPUs but just overclocked 25%. If lets say your MKII boots in 2.75 seconds the Turbo model would boot in 2.5 sec. If understood correctly it doesn't mean it can do more, it just does the same a litle faster. Correct me if I wrong but that's impression I am getting from description of Turbo model.

 

[jasonmauer]

Isn't there some misunderstanding? 25% more speed does not equal to 25% more processing power. Turbo model the way it's described means 25% faster, more responsive unit while processing power is still the same 2 CPUs but just overclocked 25%. If lets say your MKII boots in 2.75 seconds the Turbo model would boot in 2.5 sec. If understood correctly it doesn't mean it can do more, it just does the same a litle faster. Correct me if I wrong but that's impression I am getting from description of Turbo model.

First, the Turbo CPUs are not "overclocked" -- it's a new & improved model of the CPU that's clocked faster by the manufacturer, not by Fractal.

Since this is live audio processing, when we're talking about CPU load, we're also talking about speed -- it's the amount of work the CPU can accomplish within a certain time window. A faster CPU processes audio faster, which means it can do more processing in a specific amount of time compared to a slower CPU.

Now 25% faster speed doesn't necessarily translate to exactly 25% more processing power -- it could be more or less depending on the processing you're doing. Your mileage may vary depending on the blocks you're using, as it's always been. (Reverb takes more CPU than a filter block, for example.) But it does mean you're able to do more processing, since the overall processing time -- the time from input to output -- is the same.

 

[HereToday]

^^ This. As an additional point of reference, my kitchen sink preset with Full Res went down about 13% from the Mk II to the Mk II Turbo. I know others have been getting larger increases, so I think some of it has to do with what effects you are running.

 

[vrozzi]

^^ This. As an additional point of reference, my kitchen sink preset with Full Res went down about 13% from the Mk II to the Mk II Turbo. I know others have been getting larger increases, so I think some of it has to do with what effects you are running.

so your "kitchen sink preset" went down just 13% which also can be contributed to firmware updates and many other factors as many of them mention "optimizing and reducing CPU usage". It does not matter if CPU is overclocked or it is different CPU. As long as description clearly states just 25% more speed, not 25% more processing power. Based on your example you mostly paying extra for something you barely can measure or feel which in your case (per your words) translates to barely 13%. The statistical measure would fall into measuring exactly the same patch on exactly the same firmware with different units. Do you have those benchmarks to back up your claim before you state "you feel" it's faster?

 

[sprint]

It does not matter if CPU is overclocked or it is different CPU.

This matters a lot - an overclocked cpu is like a car running in overdrive - not intended for extended periods - it was clearly stated some time back by Fractal that there is no overclocking going on - it's a different chip. Not sure why some want to read "overclocked" into it - maybe the name - dunno

More CPU = more capacity for Amps/Cabs/Fx - how much more capacity for Amps/Cabs/Fx depends on the combination of Amps/Cabs/Fx used in a preset - that's it.

 

[vrozzi]

This matters a lot - an overclocked cpu is like a car running in overdrive - not intended for extended periods - it was clearly stated some time back by Fractal that there is no overclocking going on - it's a different chip. Not sure why some want to read "overclocked" into it - maybe the name - dunno

More CPU = more capacity for Amps/Cabs/Fx - how much more capacity for Amps/Cabs/Fx depends on the combination of Amps/Cabs/Fx used in a preset - that's it.

this was NOT the point of my post. The point was it DOES NOT really matter in this case. We need statistically correct measures to make conclusions. What you are doing at the moment is called ASSUMPTIONS. And assumption as it well know is "Mother of All F****Ups".

 

[HereToday]

so your "kitchen sink preset" went down just 13% which also can be contributed to firmware updates and many other factors as many of them mention "optimizing and reducing CPU usage". It does not matter if CPU is overclocked or it is different CPU. As long as description clearly states just 25% more speed, not 25% more processing power. Based on your example you mostly paying extra for something you barely can measure or feel which in your case (per your words) translates to barely 13%. The statistical measure would fall into measuring exactly the same patch on exactly the same firmware with different units. Do you have those benchmarks to back up your claim before you state "you feel" it's faster?

Are you trying to be a jerk, or does it just come naturally for you? I didn't do anything by "feel" or make any "claims". Those are your words. While I was merely sharing a casual observation, I do know a thing or two about testing equipment, so, yes, single preset, two different units, same firmware version. Hopefully, that will satisfy your ego for now.

 

[jasonmauer]

so your "kitchen sink preset" went down just 13% which also can be contributed to firmware updates and many other factors as many of them mention "optimizing and reducing CPU usage". It does not matter if CPU is overclocked or it is different CPU. As long as description clearly states just 25% more speed, not 25% more processing power. Based on your example you mostly paying extra for something you barely can measure or feel which in your case (per your words) translates to barely 13%. The statistical measure would fall into measuring exactly the same patch on exactly the same firmware with different units. Do you have those benchmarks to back up your claim before you state "you feel" it's faster?

He said the preset went down by 13% -- but if he's referring to the CPU utilization displayed, that percentage is not the same as the overall performance gain. If the preset originally used up 50% of the CPU and was reduced to 37%, that 13% drop is actually a 26% performance improvement.

Optimizations from firmware updates benefit both the original AxeFX III CPUs as well as the Turbo; they're the same architecture. I do have both Turbo and non-Turbo AxeFx III units, and have verified the performance boost of the Turbo. Why the skepticism?

 

[sprint]

this was NOT the point of my post. The point was it DOES NOT really matter in this case. We need statistically correct measures to make conclusions. What you are doing at the moment is called ASSUMPTIONS. And assumption as it well know is "Mother of All F****Ups".

lol! alrighty then - Merry Christmas!

 

[HereToday]

He said the preset went down by 13% -- but if he's referring to the CPU utilization displayed, that percentage is not the same as the overall performance gain. If the preset originally used up 50% of the CPU and was reduced to 37%, that 13% drop is actually a 26% performance improvement.

Optimizations from firmware updates benefit both the original AxeFX III CPUs as well as the Turbo; they're the same architecture. I do have both Turbo and non-Turbo AxeFx III units, and have verified the performance boost of the Turbo. Why the skepticism?

To your point, I'll add a couple more clarifications. The preset I shared above was not the one used for comparison, I was just showing all the blocks that could be loaded on the turbo. In fact, the one above would never have even loaded onto a standard Mk II.

My previous kitchen sink, which I used for comparison, went from 87ish % cpu utilization (as shown in the top right hand corner) to about 74ish %. While I was hoping for a bit more, I was happy as I was then able to load up a bunch more effects to achieve the above preset.

Had I known I was going to be called upon by some to provide a rigorous scientific defense of my test results, I would have asked for more budget from accounting: to prepare a journal submission, hire a sexy brunette assistant and procure another digital oscilloscope.

 

[H3O2]

lol! alrighty then - Merry Christmas!

I'm with you. Just remember, "An empty barrel makes the most noise", Jimmy Cagney

 

[tysonlt]

Think of the CPU(s) like a syrup factory. Raw sugar goes in, some stuff happens to it, and delicious syrup come out. Our guitar signal is the raw material. Delicious fractal sound is the packaged syrup. Oh and it’s raining outside, so the sugar has to be processed immediately or it’s ruined.

The turbo is the same factory, but has had extension so they can fit more workers in. No NEW work is being done. The recipe hasn’t changed. It can just process more syrup at once because there are more little robots inside. So processing 1 block worth of sugar will still create one block worth of syrup. It will taste (sound) exactly the same. If you process 20 blocks at the same time, the original factory can keep up but is at capacity, and will warn you- don’t send more sugar! The turbo factory can handle another 5 blocks before the little robot workers tell you they can’t keep up. For only 15 blocks of sugar, 15/20 robots will be working and 5 will be standing around telling each other robot jokes. On the turbo, 15/25 will be working, 10 will be standing around.

You can run more blocks at once. That’s all. So unless you are hitting CPU limits on a normal Axe, you will hear absolutely no difference running the same preset on a turbo Axe.