[SOLVED] Building a custom liquid-cooled PC to flip. How much to invest in parts?

Sep 11, 2019
7
0
10
Hi all - first-time poster here. I have aspirations to build an insane liquid-cooled mini-ITX gaming PC with rigid tubing, the works. Before I get there, since I've never worked with liquid cooling parts, I'd like to build and sell a couple easier systems with soft tubing and a little more space to work with. I'm planning to buy parts for the first such system this weekend. My goal is just to break even - I'm mostly looking for experience and fun. I'm planning to sell locally in the San Francisco area. Parts-wise, here's what I'm thinking:

https://pcpartpicker.com/list/rKtybX
Intel i5-9600K
MSI MPG Z390I GAMING EDGE AC Mini ITX LGA1151 Motherboard
G.Skill Trident Z RGB 16 GB (2 x 8 GB) DDR4-3600 Memory (might need to sub in a non-RGB kit since it's a tight fit on this mobo)
Intel 660p Series 1.02 TB M.2-2280 NVME Solid State Drive
EVGA GeForce RTX 2070 SUPER 8 GB XC GAMING Video Card
EVGA SuperNOVA G3 650 W 80+ Gold Certified Fully Modular ATX Power Supply
Windows 10 Home

Additional components:
Case: NZXT Manta - acquired for $50 (the only thing I've bought so far, and the only used component I intend to include)
Cooling system: EK Fluid Gaming A240 and added EK-AC GeForce RTX D-RGB GPU block

All together, this should add up to ~$1800. So here's my question: for purposes of reselling, do you think it's worth it to step up to higher versions of any of these components? My concerns are:
  • Even if I'm using a cheaper aluminum-based loop, it seems silly to put all that work and extra money into custom-cooling a mid-range system, as opposed to a higher-end system.
  • If I list this thing on Craigslist, is someone going to see "i5" instead of "i7" and wonder why they'd want to spend $1800 on it? Should I spend another $130 to step up to an i7-9700K?
  • Similarly, would spending another $200 to step up to an RTX 2080 Super add $200 to the resale value? (I'm loath to do this since I'm already nervous enough about taking apart a $550 GPU, even if EVGA's warranty allows it.)
  • If this were my build, I'd add another 240mm radiator to the loop. Will a local buyer care if I do this?
Thanks in advance for your opinions/expertise.
 
Solution
My advice is to not try and resell built PCs as there is a huge risk. If you are going to do it anyway I would get the i7 and not custom cool it because few people can see the worth in such a system as compared to its cost. The 2070S is the sweet spot for price-to-performance IMO so keep it there.
My advice is to not try and resell built PCs as there is a huge risk. If you are going to do it anyway I would get the i7 and not custom cool it because few people can see the worth in such a system as compared to its cost. The 2070S is the sweet spot for price-to-performance IMO so keep it there.
 
Solution
Sep 11, 2019
7
0
10
Yeah...I mean, you're right that this generally doesn't make sense purely from a reselling standpoint. Even if I were hell-bent on using custom liquid cooling, it would probably be wiser to leave the GPU alone and just liquid-cool the CPU. However, for me, the entire point is to get some experience, especially with installing a GPU block and putting that in the loop. So, I'm kind of my own worst enemy here. Which is why I'm hoping I'll get at least a couple people here willing to tell me I haven't completely lost my mind. :)

Also - thanks for your thoughts re: hardware. If Tom's Hardware says the RTX 2070S is ideal for VR, that's a good selling point.
 
Last edited:
Sep 11, 2019
7
0
10
Because it looks cool? :p This is why I'm building and listing a finished product, as opposed to taking commissions. You're right, no one will commission me for a custom PC until I have a few builds under my belt.

And let me reiterate - I'm not doing this for profit. I'd even be fine taking a modest loss on this project. The fact that I'm willing to sell this thing wholesale is my one advantage over letsbld.
 

g-unit1111

Titan
Moderator
My advice is to not try and resell built PCs as there is a huge risk. If you are going to do it anyway I would get the i7 and not custom cool it because few people can see the worth in such a system as compared to its cost. The 2070S is the sweet spot for price-to-performance IMO so keep it there.

Yeah I agree with this, especially if you have no experience. Something goes wrong, you could potentially lose double your initial investment. Get some experience building with loops first before you even think about reselling them.
 
Sep 11, 2019
7
0
10
Something goes wrong, you could potentially lose double your initial investment.

This is true regardless of whether or not I'm reselling. If I spend $3000 on the system I really want, that one's theoretically just as liable to break as the $1800 system I'm trying to flip.

Get some experience building with loops first before you even think about reselling them.

Getting experience is the entire point of this whole idea. If there's a better way to do it, I'm all ears. I believe you when you say there isn't a great market for what I'm offering. I'm just trying not to throw away gratuitous amounts of money building systems that no one actually wants, not to mention wasting the parts themselves. Do I build the system, test it out, then resell the individual parts? That seems like an even worse idea.

I guess I could just put out a more networking-style ad, asking if anyone who is already planning to build their own rig wants a build buddy. It's not like I have zero experience building PCs. I just don't seem to personally know anyone right now who needs a new PC and is interested in the kind of system I want to build.
 

g-unit1111

Titan
Moderator
This is true regardless of whether or not I'm reselling. If I spend $3000 on the system I really want, that one's theoretically just as liable to break as the $1800 system I'm trying to flip.

Yes but your potential flip could also be a potential backfire as well. If you're starting a business and you need to justify overhead or bulk expenses that's one thing. But if you're just a random guy trying a get rich quick scheme that's another thing entirely. Especially if your customers do their research on parts before buying.

Getting experience is the entire point of this whole idea. If there's a better way to do it, I'm all ears. I believe you when you say there isn't a great market for what I'm offering. I'm just trying not to throw away gratuitous amounts of money building systems that no one actually wants, not to mention wasting the parts themselves. Do I build the system, test it out, then resell the individual parts? That seems like an even worse idea.

No. And neither scenario sounds like it would be a good idea as far as selling systems to make money. Believe me I've thought about it as a way to make money, and if you're testing the waters as to starting a business, you'd have to have liability and collateral as well, which goes back to the point I made earlier about losing double your investment. You can't just dive in head first and expect to make a profit, because you will lose a lot of money on your initial investment and it will take a while to repay that back before you start making a profit.
 
Having built computers for a living for over 25 years now, I can say that its a very iffy proposition to get into building any computer. Most people anymore expect a high speed computer for basement prices, because they see all the ones at the big box stores selling cheap, and they don't understand the differences. You have to ask yourself, is your client base willing to pay $1800 or $3000 or more for a computer? You have to build your product to suit your client, or it will not sell.

If you don't sell, then you have excess inventory, which you have to sell at reduced price to get any of your money back.

If you have plenty of money saved, and can survive the first couple of years of losing money, you might have a chance. If not, don't try. There is a reason that most small businesses fail in the first year.
 

USAFRet

Titan
Moderator
Getting experience is the entire point of this whole idea. If there's a better way to do it, I'm all ears. I believe you when you say there isn't a great market for what I'm offering. I'm just trying not to throw away gratuitous amounts of money building systems that no one actually wants, not to mention wasting the parts themselves. Do I build the system, test it out, then resell the individual parts? That seems like an even worse idea.
You can't gain experience and make a profit on your very first custom water PC.
Especially as odd a system as a custom loop in a mini-ITX box.

I guess I could just put out a more networking-style ad, asking if anyone who is already planning to build their own rig wants a build buddy. It's not like I have zero experience building PCs. I just don't seem to personally know anyone right now who needs a new PC and is interested in the kind of system I want to build.

That's part of your marketing. Find your market. Convince someone else to shell out 2 or 3k to someone with zero experience.
And so far, I'm not being convinced.

Currently, you're going to buy $3k of parts, build something, and sell it for $1500.
 
I have to agree, what do you have to offer to the typical consumer that say a big box store cant offer? (remember, anyone with proper PC knowledge would probably prefer to do it themselves).
They see these fancy "Gaming" PCs at walmart, microcenter, cyperpower, etc all with custom loops for cheaper than yours (If you plan to make a profit), and wont think twice about it.

Yes a properly done custom system is worth the extra given the quality of components used by big PC manufacturers, but the typical consumer wont care.
Now if you had been doing this for 20 years and had a rep, charge the premium.
 

USAFRet

Titan
Moderator
Not that I would ever recommend buying a CyberPowerPC, but this is your competition:
ocIT8Vn.png


Add $50 for a liquid radiator with RGB fans.
 
Just buy a CPU block, Radiator, Pump and Reservoir combo. Buy lot many hard tubes acrylic get heat gun and that is all you need to test run the loop and practice hard tube bending as many times as you want. You won't be requiring a PC to do that at all. You connect all the components fill the reservoir loop with simple distilled water and run it for days to check if you did it perfectly or are there any leaks. Try making the loop as small and tight as possible as you progress with tight bends.
 
Last edited:
  • Like
Reactions: helper800