What do you think of a hardware Kickstarter? External GPUs DOA??

[JrPgFaN83]

I was looking over some articles for external GPUs and was noticing a lot of cost issues and tech not being up to par. What if a team of hardware designers got together and started a kickstarter to fund the project? I guarantee they would get funding. To develop a high end external graphics solution for laptops and PCs that dont normally have the graphics power? Gamers and editors alike would pay for that. Instead of having a laptop and a powerful rendering PC to maintain, they would just have their workhorse laptop and only use the external GPU for what is needed. I think its a plausible idea. I'm just hoping the cons dont outweigh the pros. I know it could get funding though. I know corporations are afraid to really step out and take the risk. But imagine the possibilities if this were to become a reality? I want to see the discussion from you guys. What is everyones thoughts on this?

 

[Newbbuilder11]

I was looking over some articles for external GPUs and was noticing a lot of cost issues and tech not being up to par. What if a team of hardware designers got together and started a kickstarter to fund the project? I guarantee they would get funding. To develop a high end external graphics solution for laptops and PCs that dont normally have the graphics power? Gamers and editors alike would pay for that. Instead of having a laptop and a powerful rendering PC to maintain, they would just have their workhorse laptop and only use the external GPU for what is needed. I think its a plausible idea. I'm just hoping the cons dont outweigh the pros. I know it could get funding though. I know corporations are afraid to really step out and take the risk. But imagine the possibilities if this were to become a reality? I want to see the discussion from you guys. What is everyones thoughts on this?

While i see the pro's of this idea i also see that it could potentially be bad for the GPU since it would be exposed and could get damaged. Also we have to keep in the idea of what the laptop specs are and if they can handle the cards because the cards need specific power from the CPU's to not bottleneck and the GPU's need a certain wattage of power to run. The hassle is just to high, unless they design a Non PSU GPU like the 750 TI or the like... but that card wouldn't make the laptop a "powerhouse" since it wouldnt get the task of 3D modeling, maybe gaming. But aside from that we would probably need software to override the intergrated GPU in order to read and utilize the GPU implanted into External GPU. But anyway that is a good idea and I hope someone tries to do this ..[strike] also I think this might be the wrong thread for this ^.^;
[/strike]

Disregard the crossed out section

 

[Ro-Tang Clan]

I was looking over some articles for external GPUs and was noticing a lot of cost issues and tech not being up to par. What if a team of hardware designers got together and started a kickstarter to fund the project? I guarantee they would get funding. To develop a high end external graphics solution for laptops and PCs that dont normally have the graphics power? Gamers and editors alike would pay for that. Instead of having a laptop and a powerful rendering PC to maintain, they would just have their workhorse laptop and only use the external GPU for what is needed. I think its a plausible idea. I'm just hoping the cons dont outweigh the pros. I know it could get funding though. I know corporations are afraid to really step out and take the risk. But imagine the possibilities if this were to become a reality? I want to see the discussion from you guys. What is everyones thoughts on this?

The idea is great and it would be amazing to do that, especially for the average laptop that doesn't have dedicated GPU's to start with. But I think that's all it will ever be, just a great idea.

In reality it just wouldn't work regardless of the funding. The issue that Newbbuilder11 raised could actually be fixed by GPU's with power bricks like external 3.5inch HDD's. However the problem lies with our current technology. Trying to integrate an external GPU with technology is going to be too much like a gimmick IMO and here's the reason why:

Connectivity. To make this project work you need a way of connecting the external GPU to the laptop or desktop PC, and getting PC manufacturers to change their hardware configuration for an add-on product is just not going to happen. So you have to work with what you've got. So exactly what is that?

On an average laptop all the connectivity sockets are 3x USB 3.0, 1x HDMI and 1x headphone and mic ports. The only feasible socket is USB but the problem inhibited by that is the latency. Like when SLI/X-fire'ing cards, you're never truly going to get double the performance. In theory the numbers add up, but in practical use it's not how it happens. If I remember correctly, one of the main issues for dual GPU technology at it's beginning was the latency between the two cards. Because there was a slight data delay when the cards tried to 'talk' to each other, it had a major performance impact as the cards could never work in tandem correctly. Bear in mind both cards for SLI/X-fire are connected on the mainboard and the delay is still present.

Now imagine that latency but 10times worse over USB. You're never really going to get a decent performance gain because of the interface the card is connected to. In theory, this could be half-fixed if an external GPU could be connected to the mainboard directly, but having an average consumer open up their PC is a bad idea. Not to mention this would void the warranty AND to keep it an external product you would still need an interface and cable of some sort to connect it to. Which is why I say the problem would be only half fixed; you would still have latency of some sort preventing you from getting the performance you need.

Ultimately, the goal would be to create an external GPU that gives a nice performance boost without being too pricey as if it's too costly the consumer might as well just buy a new laptop/dedicated GPU for desktop for the same price. To make this happen you would need a solution which doesn't take too much research and development as R&D is expensive especially if you're trying to develop a new technology. The pricier the development, the more expensive the product is going to become. To keep R&D costs low you need to make something that works with current hardware and as I explained in the previous paragraph the only way is via USB. Which makes this a gimmick product.

Developing a whole new technology which gives a real life performance boost would be far too costly and result in a product that would actually work, but for the price you can forget its existence.