Jump to content

new mini itx build... or m2 mac mini?

BigBluester

Budget (including currency):   $500 - 700

Country: Canada

Games, programs or workloads that it will be used for: general programming... database work, mathy stuff, 2D and minor 3D graphics, some machine learning stuff

Other details (existing parts lists, whether any peripherals are needed, what you're upgrading from, when you're going to buy, what resolution and refresh rate you want to play at, etc): updating old mini-itx system. its old so the only thing that is ok is the case, everything else will have to be repurchased

 

So I want to make a new mini-itx build for myself.  It is going to be a linux based system for my programming projects.  The highest priority is being as silent as possible but I would still like some power.  I have been searching for various combinations components.  Looked at various amd and intel stuff.  Also noticed this arm based motherboard:

https://en.t-firefly.com/product/industry/itx3588j

 

For a little more than my budget, I can get a new m2 mac mini and have something quiet, fast, and works out of the box.  Does this make any sense?  Seems the research involved to pull together the correct combo of parts while having a silent machine might not be worth it, right now.

 

Anyways...

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

4 minutes ago, BigBluester said:

For a little more than my budget, I can get a new m2 mac mini and have something quiet, fast, and works out of the box.  Does this make any sense?  Seems the research involved to pull together the correct combo of parts while having a silent machine might not be worth it, right now.

Depends if you can tolerate MacOS.  Asahi Linux is coming along nicely, but I've not heard if they are working on the AI cores, they are working on OpenCL but its not ready yet.

Router:  Intel N100 (pfSense) WiFi6: Zyxel NWA210AX (1.7Gbit peak at 160Mhz)
WiFi5: Ubiquiti NanoHD OpenWRT (~500Mbit at 80Mhz) Switches: Netgear MS510TXUP, MS510TXPP, GS110EMX
ISPs: Zen Full Fibre 900 (~930Mbit down, 115Mbit up) + Three 5G (~800Mbit down, 115Mbit up)
Upgrading Laptop/Desktop CNVIo WiFi 5 cards to PCIe WiFi6e/7

Link to comment
Share on other sites

Link to post
Share on other sites

That ITX board looks great, but I don't think it will have as much power as you desire. I honestly think the M2 Mac Mini would be awesome.

 

Of course, a lot of things depend on what you are willling to sacrifice, since both options you mention will have some drawbacks. Personally, I own an M1 MacBook Air and I love it, but it has given me a few headaches. For instance, if I want to work with C++, creating headers and compiling is a bit tricky. Also even though Apple Silicon has a lot of compatibility, there are still the occasional niche programs you can't run on ARM. All those issues can be relatively easy to solve with a free VM software like UTM, but it takes a lot of space, something you lack on the base model.

 

I can only mention the drawbacks I've found on my machine, but I can also tell you it can be a great experience, I hope you find this helpful.

Link to comment
Share on other sites

Link to post
Share on other sites

8 hours ago, nieblas008 said:

That ITX board looks great, but I don't think it will have as much power as you desire. I honestly think the M2 Mac Mini would be awesome.

 

Of course, a lot of things depend on what you are willling to sacrifice, since both options you mention will have some drawbacks. Personally, I own an M1 MacBook Air and I love it, but it has given me a few headaches. For instance, if I want to work with C++, creating headers and compiling is a bit tricky. Also even though Apple Silicon has a lot of compatibility, there are still the occasional niche programs you can't run on ARM. All those issues can be relatively easy to solve with a free VM software like UTM, but it takes a lot of space, something you lack on the base model.

 

I can only mention the drawbacks I've found on my machine, but I can also tell you it can be a great experience, I hope you find this helpful.

I want to program in both python and c++.  Im not the best but I write my own code.  So why is it tricky?  

Link to comment
Share on other sites

Link to post
Share on other sites

9 hours ago, BigBluester said:

Also noticed this arm based motherboard:

https://en.t-firefly.com/product/industry/itx3588j

The RK3588 is slower than even an entry level i3 CPU, so I don't think it'd be that great for what you want.

 

How serious are you about ML? It'd be nice to know if you'll require/need a GPU or not.

 

If you're ok with MacOS, then the Mac Mini would indeed be a nice option.

FX6300 @ 4.2GHz | Gigabyte GA-78LMT-USB3 R2 | Hyper 212x | 3x 8GB + 1x 4GB @ 1600MHz | Gigabyte 2060 Super | Corsair CX650M | LG 43UK6520PSA
ASUS X550LN | i5 4210u | 12GB
Lenovo N23 Yoga

Link to comment
Share on other sites

Link to post
Share on other sites

1 hour ago, BigBluester said:

I want to program in both python and c++.  Im not the best but I write my own code.  So why is it tricky?  

Python is very easy, C++ is a bit more tricky because you have to install compilers and set everything up, it's just not as simple as on Windows. I recall that when you make headers, I was using some specific features of a C++ version that did not install automatically. Again it's not the end of the world, I did get it working in the end. If you ever have any issues you could always hit me up.

Link to comment
Share on other sites

Link to post
Share on other sites

2 minutes ago, nieblas008 said:

Python is very easy, C++ is a bit more tricky because you have to install compilers and set everything up, it's just not as simple as on Windows. I recall that when you make headers, I was using some specific features of a C++ version that did not install automatically. Again it's not the end of the world, I did get it working in the end. If you ever have any issues you could always hit me up.

Doesn't Xcode ship with Clang?

FX6300 @ 4.2GHz | Gigabyte GA-78LMT-USB3 R2 | Hyper 212x | 3x 8GB + 1x 4GB @ 1600MHz | Gigabyte 2060 Super | Corsair CX650M | LG 43UK6520PSA
ASUS X550LN | i5 4210u | 12GB
Lenovo N23 Yoga

Link to comment
Share on other sites

Link to post
Share on other sites

43 minutes ago, nieblas008 said:

Python is very easy, C++ is a bit more tricky because you have to install compilers and set everything up, it's just not as simple as on Windows. I recall that when you make headers, I was using some specific features of a C++ version that did not install automatically. Again it's not the end of the world, I did get it working in the end. If you ever have any issues you could always hit me up.

This doesn't seem to be a problem.  But from some old apple experience, I thought everything is installed when you install xcode?  I don't know if things changed that much.

Link to comment
Share on other sites

Link to post
Share on other sites

1 hour ago, igormp said:

The RK3588 is slower than even an entry level i3 CPU, so I don't think it'd be that great for what you want.

 

How serious are you about ML? It'd be nice to know if you'll require/need a GPU or not.

 

If you're ok with MacOS, then the Mac Mini would indeed be a nice option.

The RK3588 doesn't have a fan and is quiet and quietness is what I want.  I just want to think about stuff... maybe I should get a good pair of headphones instead to cover up fan noise.

 

I'm serious into ML.  But its all about the size of different datasets and the types of data.  There are two areas that I like working on... sports data and engineering/physics data.  I've been trying to use sports data to make predictions in the world of fantasy sports.  In the engineering world, I am trying to find ways to reduce the complexity of certain calculations using data science techniques.  I believe that both involve less power than dealing with some of the more intense ML problems.  

 

Link to comment
Share on other sites

Link to post
Share on other sites

19 minutes ago, BigBluester said:

The RK3588 doesn't have a fan and is quiet and quietness is what I want.  I just want to think about stuff... maybe I should get a good pair of headphones instead to cover up fan noise.

I mean, as long as that level of performance is ok for you, and you are sure you won't have issues with not having x86, then it's a good idea. You could even look into other SBCs like the NanoPi R6S and the likes.

20 minutes ago, BigBluester said:

I'm serious into ML.  But its all about the size of different datasets and the types of data.  There are two areas that I like working on... sports data and engineering/physics data.  I've been trying to use sports data to make predictions in the world of fantasy sports.  In the engineering world, I am trying to find ways to reduce the complexity of certain calculations using data science techniques.  I believe that both involve less power than dealing with some of the more intense ML problems.  

Yeah, it's as you said, depends on the volume of your data and algorithms that you are using. You may even be able to just get by using colab/colab pro for your stuff.

FX6300 @ 4.2GHz | Gigabyte GA-78LMT-USB3 R2 | Hyper 212x | 3x 8GB + 1x 4GB @ 1600MHz | Gigabyte 2060 Super | Corsair CX650M | LG 43UK6520PSA
ASUS X550LN | i5 4210u | 12GB
Lenovo N23 Yoga

Link to comment
Share on other sites

Link to post
Share on other sites

1 hour ago, igormp said:

Doesn't Xcode ship with Clang?

It does. If you install Xcode you do get clang, but it's like an old version of C++. During my last lessons on Object Oriented Programming on university, I had a few issues on virtual functions, operator overload, and things like that. It was just a matter of downloading and setting up a newer version, but it was not an obvious thing. The errors were also referencing that I was on an ARM processor, and something was not able to get compiled.

Link to comment
Share on other sites

Link to post
Share on other sites

4 minutes ago, nieblas008 said:

If you install Xcode you do get clang, but it's like an old version of C++.

Oh, that's just the default cpp standard used, you can change that by passing the -std=c++YY flag to most compilers, be it clang or gcc, so the same would happen if you were to use an older install of Ubuntu, as an example.

See: https://clang.llvm.org/cxx_status.html

 

However, I can understand how this can be annoying to a student just starting, it's always better to be in the same environment as the rest of your class so you can focus in learning about programming itself, not about the ecosystem around it.

 

Anyhow, OP seems to already know how to program and get around their stuff, so such minor issues shouldn't be a problem for them.

FX6300 @ 4.2GHz | Gigabyte GA-78LMT-USB3 R2 | Hyper 212x | 3x 8GB + 1x 4GB @ 1600MHz | Gigabyte 2060 Super | Corsair CX650M | LG 43UK6520PSA
ASUS X550LN | i5 4210u | 12GB
Lenovo N23 Yoga

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×