NEW SFF GRAPHICS CARD MAKER

A

AleksandarK

n00b
Joined
May 20, 2017
Messages
23
Dear forum members. Today is a big day for me. Today i am announcing officaly that i need your help. I would like to open up a new company that makes small form factor graphics cards.

Now i what i am planing for this company? I am planing to make all king of small form factor GPUs that i will try to maintain under 170mm. If i get enough funds i think i can start with manufacturing by late 2017 or early 2018.

Funding will go throught Kickstarter.

I have a few questions for the comunity:
How should i name my company? Sugest a few names.
Are you interested in this?

Also if you have any questions feel free to ask me.
 
by "make", as in licensing + mfg yourself?

or

raise enough fund for another company to do so?
 
wahaha360 said:
by "make", as in licensing + mfg yourself?

or

raise enough fund for another company to do so?
Click to expand...
Yes licence and manufacture my self if i raise enouqh funds, but if i dont i will contact and hire another company to do it for me.
 
No offense but this seems sketch, and the company has a lot of competition already. Do you have any other successful Kickstarter campaigns?
 
1) Do you have an OEM that agreed to make the boards for you or are you planning to do them in your own factory?
2) Have you contacted AMD/Nvidia about this yet? Do you have a green light from them on this?
3) What is your/your team's experience in development of electronics?
4) How big of a budget you want to gather in crowdfunding for a production? How many units you want to make for a first run?
5) What specific features of your cards will bring the enthusiasts to back your campaign?
 
SaperPL said:
1) Do you have an OEM that agreed to make the boards for you or are you planning to do them in your own factory?
2) Have you contacted AMD/Nvidia about this yet? Do you have a green light from them on this?
3) What is your/your team's experience in development of electronics?
4) How big of a budget you want to gather in crowdfunding for a production? How many units you want to make for a first run?
5) What specific features of your cards will bring the enthusiasts to back your campaign?
Click to expand...
1. I contacted EVGA and i am currently waiting for a response
2. I contacted NVIDIA and i got my info redirected to concern team and i am waiting for a response.
3. I can solder and de solder capacitor. :D:D:D
4. I am thinking 1 million at maximum and i would like to make 5000 units in a first run
5. Size, cooling and looks
 
AleksandarK said:
1. I contacted EVGA and i am currently waiting for a response
2. I contacted NVIDIA and i got my info redirected to concern team and i am waiting for a response.
3. I can solder and de solder capacitor. :D:D:D
4. I am thinking 1 million at maximum and i would like to make 5000 units in a first run
5. Size, cooling and looks
Click to expand...

1) Even if EVGA will want to work with you, I bet they would only sell you the boards they are already manufacturing. Getting a custom board manufactured before the campaign is out of your reach without the campaign funds, I think. Without a prototype tested, you won't get money from anyone. With a chassis it's a form, shape, etc so it's easier than design of custom PCB. With that, the only thing you can really do is custom cooling solution and people will not want to buy the whole card from you if the only thing they want is a small cooler.

2) You're got huge road ahead then, because I think EVGA may not be able to sell you the boards as unfinished product for you to complete it with your cooling until you get some corporate contract with nvidia for that. Nvidia as other chip vendors has most likely some rules for the cards/boards to be following their specs, to be certified etc so the market won't be flooded with faulty products based on their chips.

3) Without a team of engineers experienced in design of consumer electronics you will not succeed. Figure out how to get them.

4) 5000 units is huge amount for such a niche product that poses huge risk of failure and at the same time if you want a custom board and/or cooling you're looking at requirement of few thousand units to at least order production.

5) Manufacturers are doing their best to make small ITX sized cards that will still be powerful enough in their price segment and stable/not overheating. I don't really see where you could go with a standard ITX sized GPU to make it smaller with no portfolio/experience...

Interesting subject though.

I wonder if you could for example start selling BOX MXM GPUs for the SFF boards like the one that asrock launched recently and at the same time as replacement for laptops. If you'd have enough cracking skills to provide modded laptop bioses for your newer GPUs, it'd could be a profitable business.
 
wahaha360 said:
Can you explain how you arrived at this number? Sources?
Click to expand...
Firstly i need to build the actuall factory. Then i have to buy expensive machines and many many components. Secondly i have to hire an entire team to desing the cutom PCB wich can take up to 6 months. After that i need to pay for licencing and payment for the workers that will help me create the graphic cards. If i contact an OEM to help me out cost will be much lower.
 
AleksandarK said:
1. I contacted EVGA and i am currently waiting for a response
2. I contacted NVIDIA and i got my info redirected to concern team and i am waiting for a response.
3. I can solder and de solder capacitor. :D:D:D
4. I am thinking 1 million at maximum and i would like to make 5000 units in a first run
5. Size, cooling and looks
Click to expand...

You are going to get laughed at until you can produce your product own PCB stack up designs for these cards to get quotes with.

And $1 million isn't enough to secure design and preproduction work for a custom GPU.
 
AleksandarK said:
Firstly i need to build the actuall factory. Then i have to buy expensive machines and many many components. Secondly i have to hire an entire team to desing the cutom PCB wich can take up to 6 months. After that i need to pay for licencing and payment for the workers that will help me create the graphic cards. If i contact an OEM to help me out cost will be much lower.
Click to expand...

No OEM is going to offer you custom design services for your pet project.
 
SaperPL said:
1) Even if EVGA will want to work with you, I bet they would only sell you the boards they are already manufacturing. Getting a custom board manufactured before the campaign is out of your reach without the campaign funds, I think. Without a prototype tested, you won't get money from anyone. With a chassis it's a form, shape, etc so it's easier than design of custom PCB. With that, the only thing you can really do is custom cooling solution and people will not want to buy the whole card from you if the only thing they want is a small cooler.

2) You're got huge road ahead then, because I think EVGA may not be able to sell you the boards as unfinished product for you to complete it with your cooling until you get some corporate contract with nvidia for that. Nvidia as other chip vendors has most likely some rules for the cards/boards to be following their specs, to be certified etc so the market won't be flooded with faulty products based on their chips.

3) Without a team of engineers experienced in design of consumer electronics you will not succeed. Figure out how to get them.

4) 5000 units is huge amount for such a niche product that poses huge risk of failure and at the same time if you want a custom board and/or cooling you're looking at requirement of few thousand units to at least order production.

5) Manufacturers are doing their best to make small ITX sized cards that will still be powerful enough in their price segment and stable/not overheating. I don't really see where you could go with a standard ITX sized GPU to make it smaller with no portfolio/experience...

Interesting subject though.

I wonder if you could for example start selling BOX MXM GPUs for the SFF boards like the one that asrock launched recently and at the same time as replacement for laptops. If you'd have enough cracking skills to provide modded laptop bioses for your newer GPUs, it'd could be a profitable business.
Click to expand...
Thank you for detailed response. It relly means to hear some wisdome from someone that has experience in the industry.
I actually thought about MXM modules and i was like no, becouse ut is used mainly for the laptops.
 
Ocellaris said:
You are going to get laughed at until you can produce your product own PCB stack up designs for these cards to get quotes with.

And $1 million isn't enough to secure design and preproduction work for a custom GPU.
Click to expand...
Maybe 1 milion isnt enough but all i can do is to try.
 
First of all, before drawing up the huge plans for business, show off what kind of GPU/Cooling you want to make. People here will give you feedback and you will see if that's even worth chasing.

Most of the people with success stories in crowdfunding were focused on precisely one thing they wanted to make, that was solving a specific problem/task and not because they wanted to be manufacturers. Crowdfunding campaigns succeed because people want specific product to be made and not to kick off some startup.

Finally, if you came here to ask "what kind of product would you like to get" and gather metrics on what you should do, stop. It's not how this works. It's the sole reason huge companies make products that are indistinct. This way we get AAA games that have female, black and gay characters, open world online crafting RPG shooters etc while there is nothing fresh in the game except for the scenario which is also far from innovative. The same goes for phone companies and laptops - they are optimising to maximise their sales.

If you want to succeed in crowdfunding with help of community, you need to set a course yourself and attract people that like your idea and then they can give you feedback to tweak your concept.
 
Start by drawing up a concept of what you'd like to change in the GPU to make it better for a SFF build, when you do this, you'll get some feedback on the subject.
 
SaperPL said:
First of all, before drawing up the huge plans for business, show off what kind of GPU/Cooling you want to make. People here will give you feedback and you will see if that's even worth chasing.

Most of the people with success stories in crowdfunding were focused on precisely one thing they wanted to make, that was solving a specific problem/task and not because they wanted to be manufacturers. Crowdfunding campaigns succeed because people want specific product to be made and not to kick off some startup.

Finally, if you came here to ask "what kind of product would you like to get" and gather metrics on what you should do, stop. It's not how this works. It's the sole reason huge companies make products that are indistinct. This way we get AAA games that have female, black and gay characters, open world online crafting RPG shooters etc while there is nothing fresh in the game except for the scenario which is also far from innovative. The same goes for phone companies and laptops - they are optimising to maximise their sales.

If you want to succeed in crowdfunding with help of community, you need to set a course yourself and attract people that like your idea and then they can give you feedback to tweak your concept.
Click to expand...
Thanks for this.
 
Don't worry about that. It is better that you asked at all, and this thread may better inform future parties.
 
SaperPL said:
Most of the people with success stories in crowdfunding were focused on precisely one thing they wanted to make, that was solving a specific problem/task and not because they wanted to be manufacturers.
Click to expand...

That is THE biggest downfall of 90% of the kickstarter "success" stories. Somebody comes up with a crazy idea, collects the money then half way thru, realizes that the given stuff doesn't exist because it is not feasible. Impossible to make (like those water out of air craps), don't deliver the imagined performance (solar roadway, Ouya, that micro drone thing etc) or simply don't work on industrial scale (plastic out of air pollution). The "inventors" realize that organizing mass production and fighting with suppliers and the generic reality is not as hip as drawing up ideas on a flip board. Then they silently disappear with the money. And the only thing that remains are 100 "success story" articles and 2-3 less publicized ones about the failure.
 
Pendra said:
That is THE biggest downfall of 90% of the kickstarter "success" stories. Somebody comes up with a crazy idea, collects the money then half way thru, realizes that the given stuff doesn't exist because it is not feasible. Impossible to make (like those water out of air craps), don't deliver the imagined performance (solar roadway, Ouya, that micro drone thing etc) or simply don't work on industrial scale (plastic out of air pollution). The "inventors" realize that organizing mass production and fighting with suppliers and the generic reality is not as hip as drawing up ideas on a flip board. Then they silently disappear with the money. And the only thing that remains are 100 "success story" articles and 2-3 less publicized ones about the failure.
Click to expand...

I was talking about the fact that you need to know what product you want to sell before figuring out the whole thing about how to manufacture it. I never stated that you don't need to be prepared to fulfil the campaign promises :)

AleksandarK started talking about building a factory and hiring people before even knowing what he really want to do and I addressed that. He might not need to do all that stuff if he can order whole manufacturing at an OEM factory like for example dondan did with LianLi. Being prepared to deliver is a must, but preparing to deliver something you don't really know what it is, is waste of time and money...
 
SaperPL said:
I was talking about the fact that you need to know what product you want to sell before figuring out the whole thing about how to manufacture it. I never stated that you don't need to be prepared to fulfil the campaign promises :)

AleksandarK started talking about building a factory and hiring people before even knowing what he really want to do and I addressed that. He might not need to do all that stuff if he can order whole manufacturing at an OEM factory like for example dondan did with LianLi. Being prepared to deliver is a must, but preparing to deliver something you don't really know what it is, is waste of time and money...
Click to expand...

Hey, no need to further explain or anything. The mass production manufacturing is not simple or easy. You need to be in it for a while to learn it. Kickstarter guys rarely have this experience. Some don't even know high school physics, chemistry or biology. That usually doesn't stop them from "inventing" in said fields. That is one reason they can come up with crazy stuff and talk about it like making it was the easiest thing on the world. Most often than not, I see confidence out of ignorance paired with tons of wishful thinking.

I come up with crazy stuff every week. I vividly remember that 20 some years ago we wanted to make a computer store with a classmate in high school. 15 years ago I was totally into wind energy. I did all preparations to build a wind turbine but eventually, the reality caught on and I realized that it is pretty hard to raise 1 million USD. I still have 1-2k unsold boxes of my revolutionary board game (the educational topic just doesn't sell well). And I still have six airsoft Barrett M95s to sell. My special social networking webpage, even with the advanced features was DOA next to Facebook. My group buying/collective bargaining start-up worked pretty well back in 2006-2007 but Groupon and similar, much better capitalized companies kinda put an end to it.
Even now, I have more pet projects and ideas than fingers (no, I'm not a pirate or a failed carpenter or butcher). Unfortunately, I also have experience in manufacturing on multiple fields from manual labor organization to robotic assembly lines and programing, from hands-on to management. So after all these, I usually take a look at my crazy ideas, do some math and planning and then drop the whole thing because it defies reality. And then, some years later, someone else comes up with the same thing, collects money and acclaim, and then disappear when the reality catches up. It is somewhat interesting to see that I was right all along, but that doesn't give the money back to the people conned by these "inventors" and doesn't help the reputation of the crowdfunding in general.
 
Pendra said:
Hey, no need to further explain or anything. The mass production manufacturing is not simple or easy. You need to be in it for a while to learn it. Kickstarter guys rarely have this experience. Some don't even know high school physics, chemistry or biology. That usually doesn't stop them from "inventing" in said fields. That is one reason they can come up with crazy stuff and talk about it like making it was the easiest thing on the world. Most often than not, I see confidence out of ignorance paired with tons of wishful thinking.

I come up with crazy stuff every week. I vividly remember that 20 some years ago we wanted to make a computer store with a classmate in high school. 15 years ago I was totally into wind energy. I did all preparations to build a wind turbine but eventually, the reality caught on and I realized that it is pretty hard to raise 1 million USD. I still have 1-2k unsold boxes of my revolutionary board game (the educational topic just doesn't sell well). And I still have six airsoft Barrett M95s to sell. My special social networking webpage, even with the advanced features was DOA next to Facebook. My group buying/collective bargaining start-up worked pretty well back in 2006-2007 but Groupon and similar, much better capitalized companies kinda put an end to it.
Even now, I have more pet projects and ideas than fingers (no, I'm not a pirate or a failed carpenter or butcher). Unfortunately, I also have experience in manufacturing on multiple fields from manual labor organization to robotic assembly lines and programing, from hands-on to management. So after all these, I usually take a look at my crazy ideas, do some math and planning and then drop the whole thing because it defies reality. And then, some years later, someone else comes up with the same thing, collects money and acclaim, and then disappear when the reality catches up. It is somewhat interesting to see that I was right all along, but that doesn't give the money back to the people conned by these "inventors" and doesn't help the reputation of the crowdfunding in general.
Click to expand...
We all must start somewhere, i had to post this becouse if i havent i would have false hope for years and it would take so much of my time.
 
A good friend said to me: "The most people fail because they only focus on things that can fail and this prevent them to just start with the idea."

I have a split opinion about it: "In the first place there should be your idea but you should also be realistic. You have to address problems, but they are there to solve and not to stop you."


I think there are five key features to do it right:


- Believe in your idea but be realistic
- Work hard and on the same or higher professional level as the competitors ( master plan, website, content, fulfill standards, graphics, technical level s.o.)
- Solve problems and don't give up
- failure/mistakes happen, learn out of it and don't do the mistake again
- don't do thinks fast or halfheartedly do it right/ as perfect as possible


So in your case you should start with key feature one and two. Plan your project right before asking for help. Make drawings, pointing out the idea getting an oem s.o. If you ask if somebody is interested in an idea you should ask yourself if you believe in this. Also asking for a company name should not the first step you think about. If you are afraid in wasting time you are not believing in your idea.
 
Last edited:
AleksandarK said:
19 years an curently in high school.
Click to expand...
Well, if you have such dreams, enroll to a good university and complete an MSC in electronics engineering and specialize in micro electronics. During or after the university, take a job at a computer component developer and/or manufacturing company. If you are good enough, ~12 years from now, you will have the necessary skills and experience, maybe even the connections, to pursue your dream. I know, life is cruel, unforgiving and doesn't give a **** about one's dreams.
 
Last edited:
dondan said:
A good friend said to me: "The most people fail because they only focus on things that can fail and this prevent them to just start with the idea."

I have a split opinion about it: "In the first place there should be your idea but you should also be realistic. You have to address problems, but they are there to solve and not to stop you."


I think there are five key features to do it right:


- Believe in your idea but be realistic
- Work hard and on the same or higher professional level as the competitors ( master plan, website, content, fulfill standards, graphics, technical level s.o.)
- Solve problems and don't give up
- failure/mistakes happen, learn out of it and don't do the mistake again
- don't do thinks fast or halfheartedly do it right/ as perfect as possible


So in your case you should start with key feature one and two. Plan your project right before asking for help. Make drawings, pointing out the idea getting an oem s.o. If you ask if somebody is interested in an idea you should ask yourself if you believe in this. Also asking for a company name should not the first step you think about. If you are afraid in wasting time you are not believing in your idea.
Click to expand...
Thanks for the wisdome.
 
Pendra said:
Well, if you have such dreams, enroll to a good university and complete an MSC in electronics engineering and specialize in micro electronics. During or after the university, take a job at a computer component developer and/or manufacturing company. If you are good enough, ~12 years from now, you will have the necessary skills and experience, maybe even the connections, to pursue your dream. I know, life is cruel, unforgiving and doesn't give a **** about one's dreams.
Click to expand...
You are right and i maybe 10-15 from now i will have enough knowlage to do something.
 
If I wanted to do custom boards I would build a graphics card with the components on what's now called the back of the card and have standard Intel mounting holes and then use a simple 90 riser and sell it with an appropriate case so that you can have your favourite tower cooler on your GPU because how insane it is that we use 1.3kg worth of metal in a roughly 15cm cubed shape for the CPU and a low profile bad airflow cooler for the GPU which now is always much, much larger TDP. Now see that would be a product to buy.
 
  • Like
Reactions: ceski
like this
chx said:
If I wanted to do custom boards I would build a graphics card with the components on what's now called the back of the card and have standard Intel mounting holes and then use a simple 90 riser and sell it with an appropriate case so that you can have your favourite tower cooler on your GPU because how insane it is that we use 1.3kg worth of metal in a roughly 15cm cubed shape for the CPU and a low profile bad airflow cooler for the GPU which now is always much, much larger TDP. Now see that would be a product to buy.
Click to expand...
You would probably really like the MXM form factor, if it ever takes off.

15879351_1488462101168246_396735554_n.png

See that tiny thing at the top with a fan? That's the GPU and it slides in horizontally into the motherboard. Some suppliers are selling up to a GTX 1080 in the size.

Pg8ZiZr.jpg

Imagine a decent sized pair of heatsinks or a custom one piece heat sink that covered both the CPU and GPU. Some good reading material if you're inclined:

https://www.cool3c.com/article/116315

https://smallformfactor.net/news/asrock-z270m-stx-mxm-pictured

https://smallformfactor.net/forum/threads/custom_mod-nano-mxm-1-19l-rev-1.1557/page-5#post-41139
 
I am well aware of that but I have also seen the price of the MXM cards and also they are pitched for very small factor which is silly. The case should be like a cube with nice large heatsinks.

Also, how long do we need to wait for that? It was shown in January it's almost June. Maybe it'll be shown again at Computex and we will hear a price and release date or it remains vaporware.
 
AleksandarK said:
Dear forum members. Today is a big day for me. Today i am announcing officaly that i need your help. I would like to open up a new company that makes small form factor graphics cards.

Now i what i am planing for this company? I am planing to make all king of small form factor GPUs that i will try to maintain under 170mm. If i get enough funds i think i can start with manufacturing by late 2017 or early 2018.

Funding will go throught Kickstarter.

I have a few questions for the comunity:
How should i name my company? Sugest a few names.
Are you interested in this?

Also if you have any questions feel free to ask me.
Click to expand...
fixing the English of your first post would be a great start
 
You must log in or register to reply here.
Back
Top