I have always taken it as granted that my next laptop is quad core but I just realized that even the power supply must be very heavy (..?) as it has to supply up to 120W of power. In many situations (e.g. mains socket in airplane in long duration flight provide up to 60W (or was it 70W) power. Would quad core be able to load in such a situation)..?
So If you have two laptops with the exact same processor, one with quad core version and one with dual core version (e.g. Macbook Pro 13 vs. Macbook Pro 15), how much does the amount of plugins that you can use inside the daw increase practically?
So if you can have a project with lets say 30 instances of NI Massive in Macbook Pro 13, does it mean that you can have a project of 60 instances (or even near 60) in 15 inch Macbook Pro or would the increase be maybe from 30 to 40 instances or even smaller...?
I mean if the difference is very small I could probably manage with a dual core as I really don't need the lowest latency (the most important factor for me is amount of simultaneous heavy plugins).
How much does quad-core really boost DAW (MBP 13 vs 15)
-
- Banned
- 22457 posts since 5 Sep, 2001
Ok well i have gone through THIS scenario..
i had a dual core macbook pro i5 2.4ghz November 2010 and then went to a quad core 2.2ghz i7 march 2011.
It was more than double in this case..
As long as the host uses multithreading, and we are comparing same architecture, say a 2.4 sandy bridge dual vs a 2.4 sandy bridge quad, to use an example, then, yes, it really IS double the plug in power.
In my case it was more.
It's also around 6 x the power of my 2008 dual core core 2 duo which was 2.4hhz.
so a 2.2ghz quad is 6x more powerful than a 2.4 dual in this case..
my point being the cpu architecture version also has a lot to do with it.
It would be good to know what speed and model you are using now and what you plan to get or are looking at, so we can give you a rough guideline of what to expect
i had a dual core macbook pro i5 2.4ghz November 2010 and then went to a quad core 2.2ghz i7 march 2011.
It was more than double in this case..
As long as the host uses multithreading, and we are comparing same architecture, say a 2.4 sandy bridge dual vs a 2.4 sandy bridge quad, to use an example, then, yes, it really IS double the plug in power.
In my case it was more.
It's also around 6 x the power of my 2008 dual core core 2 duo which was 2.4hhz.
so a 2.2ghz quad is 6x more powerful than a 2.4 dual in this case..
my point being the cpu architecture version also has a lot to do with it.
It would be good to know what speed and model you are using now and what you plan to get or are looking at, so we can give you a rough guideline of what to expect
-
- KVRist
- Topic Starter
- 489 posts since 6 Feb, 2010
I have a Core2Duo T6670 2.2ghz (bought in 2010) Lenovo Thinkpad SL510.TheoM wrote: It would be good to know what speed and model you are using now and what you plan to get or are looking at, so we can give you a rough guideline of what to expect
I really don't know how to judge plugin performance. I have watched at passmarks but I am not sure if it is a good indicator as it takes hyperthreading etc.. into account that don't afaik boost plugin performance.
-
- Banned
- 22457 posts since 5 Sep, 2001
ok if you go for a current quad core laptop then, the performance difference will be mind blowing, no jokes.
-
- KVRist
- Topic Starter
- 489 posts since 6 Feb, 2010
I have also quad desktop (i7 3770K)TheoM wrote:ok if you go for a current quad core laptop then, the performance difference will be mind blowing, no jokes.
I made a test, I could run 15 instances of Massive with my older lenovo, with i7 3770K I could do 80 instances. The question is that how many instances could a macbook 13 or 15 inch do. My i7 clock frequency is 3.5 ghz, so when I scale it down to 2.3ghz of MBP15, it would make 52 instances. Then if I divide this by two and scale up to 2.5ghz (MBP13 frequency) it would make 28 instances, less than double from my Lenovo.
Does these numbers sound right or did my logic for calculating them was too wrong..?
-
- KVRAF
- 5948 posts since 19 Jun, 2008 from Melbourne, Australia
If you are able to access the machines you want to compare, you can use DAW Bench:
http://www.dawbench.com/benchmarks.htm
There are projects for various hosts available on the above page. Basic idea is that you load the project and compare how many instances of a specific plugin can be used. The plugin is a compressor from Cockos that was coded specifically for the benchmarking test.
Peace,
Andy.
http://www.dawbench.com/benchmarks.htm
There are projects for various hosts available on the above page. Basic idea is that you load the project and compare how many instances of a specific plugin can be used. The plugin is a compressor from Cockos that was coded specifically for the benchmarking test.
Peace,
Andy.
... space is the place ...
-
- KVRist
- Topic Starter
- 489 posts since 6 Feb, 2010
Nothing from Cockos, at least on the front page.ZenPunkHippy wrote:If you are able to access the machines you want to compare, you can use DAW Bench:
http://www.dawbench.com/benchmarks.htm
There are projects for various hosts available on the above page. Basic idea is that you load the project and compare how many instances of a specific plugin can be used. The plugin is a compressor from Cockos that was coded specifically for the benchmarking test.
Peace,
Andy.
And I have been in that web page before but I find it pretty confusing, especially the forum. I really don't understand what is a universal suite, etc... They could structure and explain a bit better the information in there, or that is at least my opinion
Well I guess I got already what I wanted, at least from my machines, Just would need to do the same thing in macbooks
-
- KVRist
- 231 posts since 17 Jan, 2010
I was given a MBPr (2.6ghz i7) a couple of months back and I was simply blown away by it... the thing is a frigging monster, Just the idea of being able to load crap loads of plugins and not having to worry about CPU usage. Don't know how I coped with my old iMac (2.4ghz Core2Duo).
My workflow and productivity has dramatically increased! It really is the best machine I have EVER used by a million miles...
k3ith
My workflow and productivity has dramatically increased! It really is the best machine I have EVER used by a million miles...
k3ith
-
- KVRAF
- 7897 posts since 12 Feb, 2006 from Helsinki, Finland
In a typical mixing scenario you will observe roughly twice the CPU power with twice the cores. This assumes that in such a "typical mixing scenario" the majority of CPU is spent on synths and per-channel insert effects (with the channels parallel rather than serial), where the large cumulative costs (rather than one or two specific instances) dominate the CPU load. It also assumes the host doesn't try to do anything stupid with regards to scheduling (unfortunately that is not always the case.. but even then you will normally see almost twice the parallel power).
If you want to build longer chains of plugins, then you will observe zero benefit from simply adding more cores. I don't think this is very typical "run out of CPU" situation in practice though.
If you want to build longer chains of plugins, then you will observe zero benefit from simply adding more cores. I don't think this is very typical "run out of CPU" situation in practice though.
-
- KVRAF
- 10540 posts since 20 Nov, 2003 from Lost and Spaced
Upgraded from a dual to quad core. Makes a BIG difference.
-
- KVRAF
- 25053 posts since 20 Oct, 2007 from gonesville
I don't know anything about the hardware as per the OP, but the way I use my machine, 8 cores is far more than 4x a dual core. It's going to depend on the software, if it's geared towards it. I was initially somewhat skeptical as to more than 8 cores (16 logical cores) for music but the better I understand it the more I'd embrace it and would probably acquire more cores if possible.
-
- KVRAF
- 4205 posts since 21 Oct, 2001 from my bolthole in the south pacific
I would wait for quad core Broadwell chips to be released or even for Skylake. Seems like a bad idea to go down to 2 cores.alexfsu wrote:What about a new Broadwell i7 dual vs. the old Haswell i7 quad. In my case it'd be a 15in 2.4 quad core 2011 MBP vs. 13in. 3.1 dual core MBP released today
Should be along next quarter.
"I got a car battery and two jumper cables that argue different."
Rust Cohle
Rust Cohle
-
- KVRAF
- 2290 posts since 18 Oct, 2010 from Japan
For productivity purposes, in most cases, more cores are better: BUT, and this is a big butt (:hihi:), there is one caviat to that.
Know your brands and what makes them different:
Intel uses hyperthreading instead of cores. Hyperthreads, in layman's terms is effectively "simulated" cores. So if you have a quad core i7 with 2 threads per core, you effectively have the power of a standard 8 core CPU. Hyperthreading is an important technology that makes Intel stand out from AMD.
AMD works strictly on raw cores. They do not have Hyperthreading technology at all.
Knowing this will help you make a choice, but one thing to keep in mind is application's ability to use cores. This is where hyperthreading excels. It's up to the programmer to determine how the software utilizes each core. So if a programmer designs a software to be able to utilize only 4 cores, have more than 4 will simply sit idly. So how do you make 4 cores more powerful? Well there is the clock speed for one, but this doesn't make as much of an impact. Hyperthreads which are threads per core can be utilized, effectively making a quad-core application run as a powerful 8 core.
In general, Intel is better for productivity, though it comes at quite a cost. However, on the bright side, most Intel chips are: More efficient, and run much cooler. However, the difference in electricity costs between Intel and AMD saved per year is fairly negligible, so I don't know if I would count that as a win on Intel's part, but this is something to keep in mind if you're running on a laptop which airflow and energy consumption is a bigger deal.
Haswell is the series you want to go with for Laptops. Haswell's key determining factor is it's efficiency and how cool it runs, roughly about 20% better than any desktop CPU on the market. But, Intel DOES NOT MAKE hyperthreaded Haswell CPUs for the i3 and i5 models. i7 and Extreme parts are the only ones to receive hyperthreading.
The only reason why you should ever use a dual core at this day is either for micro computing or gaming-on-a-budget. Dual cores is a severe bottleneck for 64x OS. Even at high clock speeds, dual cores just can't keep up with modern parts very well. If you are still using dual core, I highly suggest moving to any quad core.
If you want my opinion, my Lian-Li mini ITX lanbox can kick many large itb machines to the curb. It uses z97 architecture with an quad i5 Haswell at 3.4ghz, with 8gb RAM (probably the biggest short coming of the build), a 750w Platinum PSU, and a (irrelevant to music) EVGA GTX780 6GB edition.
Doing a Diva Stress Test:
ten 3-voiced instances of Diva on Divine hit 50%
on a kontakt stress test:
Roughly about 10 different instruments could be loaded before maxing RAM usage.
on a "Normal Project" test (this is subjective because my projects could be vastly different from yours)
the intro minimum was around 9% usage. Intro max was around 13% usage. The average was around 11%
During peak playback"
Minum was around 36%. Max was around 49%. the average was around 41%
This project uses 5 instances of Diva, 3 different kontakt instruments with a total of 14 instances of Kontakt and roughly 70 effects (plus a couple of Microtonic and Synth1 instances too, though their cost of CPU is negligible)
All tests performed on Mixcraft 6
Know your brands and what makes them different:
Intel uses hyperthreading instead of cores. Hyperthreads, in layman's terms is effectively "simulated" cores. So if you have a quad core i7 with 2 threads per core, you effectively have the power of a standard 8 core CPU. Hyperthreading is an important technology that makes Intel stand out from AMD.
AMD works strictly on raw cores. They do not have Hyperthreading technology at all.
Knowing this will help you make a choice, but one thing to keep in mind is application's ability to use cores. This is where hyperthreading excels. It's up to the programmer to determine how the software utilizes each core. So if a programmer designs a software to be able to utilize only 4 cores, have more than 4 will simply sit idly. So how do you make 4 cores more powerful? Well there is the clock speed for one, but this doesn't make as much of an impact. Hyperthreads which are threads per core can be utilized, effectively making a quad-core application run as a powerful 8 core.
In general, Intel is better for productivity, though it comes at quite a cost. However, on the bright side, most Intel chips are: More efficient, and run much cooler. However, the difference in electricity costs between Intel and AMD saved per year is fairly negligible, so I don't know if I would count that as a win on Intel's part, but this is something to keep in mind if you're running on a laptop which airflow and energy consumption is a bigger deal.
Haswell is the series you want to go with for Laptops. Haswell's key determining factor is it's efficiency and how cool it runs, roughly about 20% better than any desktop CPU on the market. But, Intel DOES NOT MAKE hyperthreaded Haswell CPUs for the i3 and i5 models. i7 and Extreme parts are the only ones to receive hyperthreading.
The only reason why you should ever use a dual core at this day is either for micro computing or gaming-on-a-budget. Dual cores is a severe bottleneck for 64x OS. Even at high clock speeds, dual cores just can't keep up with modern parts very well. If you are still using dual core, I highly suggest moving to any quad core.
If you want my opinion, my Lian-Li mini ITX lanbox can kick many large itb machines to the curb. It uses z97 architecture with an quad i5 Haswell at 3.4ghz, with 8gb RAM (probably the biggest short coming of the build), a 750w Platinum PSU, and a (irrelevant to music) EVGA GTX780 6GB edition.
Doing a Diva Stress Test:
ten 3-voiced instances of Diva on Divine hit 50%
on a kontakt stress test:
Roughly about 10 different instruments could be loaded before maxing RAM usage.
on a "Normal Project" test (this is subjective because my projects could be vastly different from yours)
the intro minimum was around 9% usage. Intro max was around 13% usage. The average was around 11%
During peak playback"
Minum was around 36%. Max was around 49%. the average was around 41%
This project uses 5 instances of Diva, 3 different kontakt instruments with a total of 14 instances of Kontakt and roughly 70 effects (plus a couple of Microtonic and Synth1 instances too, though their cost of CPU is negligible)
All tests performed on Mixcraft 6
-
- KVRAF
- 3897 posts since 28 Jan, 2011 from MEXICO
Is mixcraft cpu efficient?ntom wrote: It uses z97 architecture with an quad i5 Haswell at 3.4ghz, with 8gb RAM (probably the biggest short coming of the build), a 750w Platinum PSU, and a (irrelevant to music) EVGA GTX780 6GB edition.
Doing a Diva Stress Test:
ten 3-voiced instances of Diva on Divine hit 50%
I work on Live and an i5 760 ( I think its two generations before the one you have) and I can only get 4 Diva instances on Divine.
I wanna try to overclock my current CPU to see if I gain some performance.
dedication to flying
