Socketed GPU?

[WhitetailAni]

1 minute ago, olliecoding321 said:

I didn't know that it is impossible for people outside of pc part companies to to come up with new ideas for tech

No, it's not that it's a bad idea.

It's that it's a bad idea right now because we don't need to.

Changing something that's almost 20 years old now for no reason other than that "it's new" is not a good reason.

[olliecoding321]

5 minutes ago, HelpfulTechWizard said:

You need some more about math because half of 1 amp will be too much for something that needs .05a.

 

5ma is 0.005a

0.05a is 50ma

[Helpful Tech Witch]

2 minutes ago, olliecoding321 said:

I am sorry that I suggested the creation of a new type of technology I didn't know that it is impossible for people outside of pc part companies to to come up with new ideas for tech

It's just a bad idea. Now good ideas that non tech people have come up with:

1. Automatic telephone system- created by a (very petty, might I add) undertaker
2. Intel not being so sh*tty with there mid and low end boards- all consumers ever

[Helpful Tech Witch]

1 minute ago, olliecoding321 said:

5ma is 0.005a

0.05a is 50ma

I said 1a and .05a

You said .05a was half of 1a. 

That's not true.

[WhitetailAni]

7 minutes ago, olliecoding321 said:

I am suggesting a new type of PCIe card

So MXM 4.0?
These failed.

Failed HARD.

[olliecoding321]

Just now, HelpfulTechWizard said:

It's just a bad idea. Now good ideas that non tech people have come up with:

1. Automatic telephone system- created by a (very petty, might I add) undertaker
2. Intel not being so sh*tty with there mid and low end boards- all consumers ever

I am sorry but I like to think of new ways of doing things

[olliecoding321]

Just now, HelpfulTechWizard said:

I said 1a and .05a

You said .05a was half of 1a. 

That's not true.

i said half of 1a is 0.5

[Helpful Tech Witch]

Just now, olliecoding321 said:

I am sorry but I like to think of new ways of doing things

Yeah. And you like to argue with people who point out flaws.

 

I'm all it for New ideas and ways to do things, but the people need to realise oitations and adapt.

[olliecoding321]

1 minute ago, FakeKGB said:

So MXM 4.0?
These failed.

Failed HARD.

just because one new thing failed doesn't mean that every other new thing has to.

I never knew that we aren't able to think beyond what we have now and believe in new things

[olliecoding321]

Just now, HelpfulTechWizard said:

Yeah. And you like to argue with people who point out flaws.

 

I'm all it for New ideas and ways to do things, but the people need to realise oitations and adapt.

because I try to modify my ideas to get rid of flaws and to try to think of something that could work

[olliecoding321]

Just now, olliecoding321 said:

because I try to modify my ideas to get rid of flaws and to try to think of something that could work

it's called inventing

[WhitetailAni]

Just now, olliecoding321 said:

just because one new thing failed doesn't mean that every other new thing has to.

There were 3 revisions of MXM.

Each one never succeeded.

Once something fails 3 times, it won't succeed again. There's a building near me that has had 19 restaurants in it now. Each one has failed, whether it be in weeks, months, or years.

But they always have gone out of business before 2 years.

1 minute ago, olliecoding321 said:

I never knew that we aren't able to think beyond what we have now and believe in new things

Who said we aren't innovating?
We have PCIe 6.0 in the works, with 128 GB (that's GigaBYTES) per second theoretical transfer speed.

[Helpful Tech Witch]

2 minutes ago, olliecoding321 said:

just because one new thing failed doesn't mean that every other new thing has to.

I never knew that we aren't able to think beyond what we have now and believe in new things

I'm done at this point. Someone who takes criticism and turned it into saying you can't have ideas is someone I don't want to have a conversation with.

 

User ignored.

[The_russian]

9 minutes ago, olliecoding321 said:

I am sorry that I suggested the creation of a new type of technology I didn't know that it is impossible for people outside of pc part companies to to come up with new ideas for tech

 

2 minutes ago, olliecoding321 said:

I am sorry but I like to think of new ways of doing things

There's nothing wrong with coming up with new ideas but if you want the ideas to get any further than just a thought or fantasy, you have to be able to take criticism and actually address the problems that come along with the idea. Another thing to think about is considering "should we" do something before "can we" do something. Just because something is possible to do, doesn't mean it is a good idea. This idea has been brought up countless times over the last couple decades and is by no means a new idea. If it was as good a solution as you seem to think it is, it would have been done already.

[trag1c]

5 minutes ago, HelpfulTechWizard said:

You need some more about math because half of 1 amp will be too much for something that needs .05a.

 

I am not the least bit on board with this guys socket idea but he is correct in saying that power supply amperage value is the maximum number of amps that can be drawn from it within the thermal constraints of the design.  At the end of the day Ohms law doesn't lie, I = V / R. The voltage out of your power supply will be constant so the only thing that can change the current draw is the resistance which is also constant because the resistance is determined by the sum of the series parallel resistance of the entire device.

 

TLDR; PS amp rating =/= current draw. The only 2 ways to change current draw are to increase voltage and/or decrease resistance. 

 

(Every thing here assumes theoretical speak. Real world issues would persist but for the basis of this conversation it is not needed.) 

[akio123008]

8 minutes ago, HelpfulTechWizard said:

I'm done at this point. Someone who takes criticism and turned it into saying you can't have ideas is someone I don't want to have a conversation with.

So when are you going to stop ignoring my attempts to explain ohm's law and take criticism yourself?

(or have I been ignored too haha)

 

BTW, I take no stance on socketed GPUs, I think GPUs are fine I guess? I'm not into GPUs enough to bother.

 

 

[Elisis]

26 minutes ago, HelpfulTechWizard said:

You need some more about math because half of 1 amp will be too much for something that needs .05a.

Good thing no PC components (or really, any components outside of toy circuits or niche uses) run constant current! Current is drawn, voltage is forced 

[Elisis]

9 minutes ago, trag1c said:

At the end of the day Ohms law doesn't lie, I = V / R.

Oh, Ohm's Law lies all the time. Inductors, caps, and semiconductors do exist, after all.

[trag1c]

4 minutes ago, Elisis said:

Oh, Ohm's Law lies all the time. Inductors, caps, and semiconductors do exist, after all.

True but for the purpose of the conversation at that point it doesn't lol. Thats why I put the disclaimer at the bottom.  

6 minutes ago, Elisis said:

Good thing no PC components (or really, any components outside of toy circuits or niche uses) run constant current! Current is drawn, voltage is forced 

I have to agree that CV is typically way more useful and prevalent. Off the top of my head the best use for CC is LED's because they gotta be weird and change their resistance with voltage. 
 

[Helpful Tech Witch]

20 minutes ago, akio123008 said:

So when are you going to stop ignoring my attempts to explain ohm's law and take criticism yourself?

(or have I been ignored too haha)

 

BTW, I take no stance on socketed GPUs, I think GPUs are fine I guess? I'm not into GPUs enough to bother.

 

 

no, in fact I didn't pay much attention to your quote lol. i just thought it was one of op's

 

I didnt know that there was 2 ways to do it, like @Elisis pointed out,

[akio123008]

Just now, HelpfulTechWizard said:

i just thought it was one of op's

I realise I have the same generic profile picture haha

 

17 minutes ago, Elisis said:

Oh, Ohm's Law lies all the time. Inductors, caps, and semiconductors do exist, after all.

But if you take all those components' weird properties into account and pile them up into a new number called the impedance (aka Z) you can replace R in ohm's law for Z and everything works just fine. The equation becomes I = V / Z.

 

[Elisis]

14 minutes ago, akio123008 said:

you can replace R in ohm's law for Z and everything works just fine

Then it's not pure Ohm's Law. It's generalized to work with reactive components. That also does nothing to explain semiconductors.

[PlayerLorenzo]

the PCI-E slot already is a socket in their own right. BUT something close to the average CPU socket is a MXM GPU, They look like this, and are usually made for laptops and mini PCs that do not have space for a full-sized GPU in a PCI slot. But this is usually made for laptops and mini PCs, and you probably won't see that MXM slot on a regular desktop motherboard  SO, a PCI slot already is a socket in their own right

[RoseLuck462]

If GPUs became socketed then (among many things) mobo compatibility would be that much more complex.

 

Motherboards would need to account for Intel and AMD specific CPU sockets along with Nvidia and AMD GPU sockets as well.

[WhitetailAni]

@olliecoding321

Here's an example of what could happen if you put an RTX 3090 in a GPU socket designed for a GT 1030: