Mobile Gpu Lineup Leaks for Nvidia + Amd

[VEGGIM]

A Nvidia leak has revelaed that the rtx mobile series will be very more different than desktop compared to how 30 series was. 
4090m ad103
4080TIm ad104
4070m ad106
4060ti/4060m ad107

Meanwhile for AMD, there has been leaks that the top-of-the-line 7000m series gpu is as powerful as a RTX 3090/6950 XT. No specs have been revealed or any power limits.

[VEGGIM]

There is also something else here: 

 

[jaybee83]

21 hours ago, VEGGIM said:

There is also something else here: 

 

geez, were talking 50-60% of the 4090 desktop cuda cores and that at half to a third of the available TGP. might be performing like a 4070, if not worse... they might as well use the desktop counterpart names to stop confusing customers. but i get it, customer is happier when he has a "higher number" in his machine, even works on a lot of us here in the forum 😅

[Meaker]

On 10/23/2022 at 10:37 PM, jaybee83 said:

geez, were talking 50-60% of the 4090 desktop cuda cores and that at half to a third of the available TGP. might be performing like a 4070, if not worse... they might as well use the desktop counterpart names to stop confusing customers. but i get it, customer is happier when he has a "higher number" in his machine, even works on a lot of us here in the forum 😅

 

The 4090 is on the ragged edge of scaling, for mobile scaling the TGP back to that point will put it in the efficiency sweet spot. Though the top card should just be a 4070ti mobile in name.

[jaybee83]

6 hours ago, Meaker said:

 

The 4090 is on the ragged edge of scaling, for mobile scaling the TGP back to that point will put it in the efficiency sweet spot. Though the top card should just be a 4070ti mobile in name.

exactly, they gotta follow the efficiency curve for mobile hardware, thats why its dumb for them to use the desktop naming. better go back to the "M" days or smth along those lines.

[1610ftw]

On 11/12/2022 at 7:32 AM, jaybee83 said:

exactly, they gotta follow the efficiency curve for mobile hardware, thats why its dumb for them to use the desktop naming. better go back to the "M" days or smth along those lines.

 

Same chip but less power - understandable.

Different chip but using the name of the top of the line GPU - disingenuous at best.

 

I think we all agree that names matter and using the name of a GPU that in this generation may end up being twice as fast is getting ridiculous.

 

And I will say again that in general there is a complete lack of ambition with regard to mobile products these days - how they did not even have the guts to up the TDP for this generation is beyond me. Even something like this would have made a nice difference when paired with the chipsets of the desktop cards:

 

4090 Mobile (AD102): up to 300W

4080 Mobile (AD103): up to 225W

4070 Ti Mobile (AD104): up to 150W

 

No problem dissipating up to 400 or 450W for GPU and CPU in combined workloads with a proper vapor chamber solution and in fact cooling performance like that has been done before by Acer, Asus, Clevo and MSI.

 

That would be more honest and if for some reason manufacturers think they cannot put the AD102 in a laptop they do not have to - in that case just do not offer a mobile 4090.

 

 

[jaybee83]

14 hours ago, 1610ftw said:

 

Same chip but less power - understandable.

Different chip but using the name of the top of the line GPU - disingenuous at best.

 

I think we all agree that names matter and using the name of a GPU that in this generation may end up being twice as fast is getting ridiculous.

 

And I will say again that in general there is a complete lack of ambition with regard to mobile products these days - how they did not even have the guts to up the TDP for this generation is beyond me. Even something like this would have made a nice difference when paired with the chipsets of the desktop cards:

 

4090 Mobile (AD102): up to 300W

4080 Mobile (AD103): up to 225W

4070 Ti Mobile (AD104): up to 150W

 

No problem dissipating up to 400 or 450W for GPU and CPU in combined workloads with a proper vapor chamber solution and in fact cooling performance like that has been done before by Acer, Asus, Clevo and MSI.

 

That would be more honest and if for some reason manufacturers think they cannot put the AD102 in a laptop they do not have to - in that case just do not offer a mobile 4090.

 

 

 

u forget one very important thing in these calculations: the machine cannot be THIN anymore! and, as we all know, this is the absolute most important aspect of all mobile machines 😛 

[1610ftw]

4 hours ago, jaybee83 said:

 

u forget one very important thing in these calculations: the machine cannot be THIN anymore! and, as we all know, this is the absolute most important aspect of all mobile machines 😛 

 

Oh no, forrgot about that,  let me correct my mistake:

 

4090 MEGA (AD104): up to 125W

4080 ULTRA (AD105): up to 100W

4070 GIGA (AD106): up to 75W

 

That should do it as long as we add enough RGB 😁

[KING19]

On 11/15/2022 at 8:06 AM, 1610ftw said:

 

Oh no, forrgot about that,  let me correct my mistake:

 

4090 MEGA (AD104): up to 125W

4080 ULTRA (AD105): up to 100W

4070 GIGA (AD106): up to 75W

 

That should do it as long as we add enough RGB 😁

 

Perfect calculations! 😁

 

Dont forget about the soldiered RAM!. Fits well with the future of gaming laptops.

[jaybee83]

On 11/17/2022 at 11:36 PM, KING19 said:

 

Perfect calculations! 😁

 

Dont forget about the soldiered RAM!. Fits well with the future of gaming laptops.

 

oh oh, wait! i got another one!

 

- soldered power cable

- soldered display connection

- soldered keyboard

- abandon any and all ports, go fully wireless

 

the future looks grand!

 

/s

[Bullit]

You forgot: rent only.

[VEGGIM]

On 11/14/2022 at 1:17 PM, 1610ftw said:

 

Same chip but less power - understandable.

Different chip but using the name of the top of the line GPU - disingenuous at best.

 

I think we all agree that names matter and using the name of a GPU that in this generation may end up being twice as fast is getting ridiculous.

 

And I will say again that in general there is a complete lack of ambition with regard to mobile products these days - how they did not even have the guts to up the TDP for this generation is beyond me. Even something like this would have made a nice difference when paired with the chipsets of the desktop cards:

 

4090 Mobile (AD102): up to 300W

4080 Mobile (AD103): up to 225W

4070 Ti Mobile (AD104): up to 150W

 

No problem dissipating up to 400 or 450W for GPU and CPU in combined workloads with a proper vapor chamber solution and in fact cooling performance like that has been done before by Acer, Asus, Clevo and MSI.

 

That would be more honest and if for some reason manufacturers think they cannot put the AD102 in a laptop they do not have to - in that case just do not offer a mobile 4090.

 

 

Question. What laptops have existed that put out 300w on a single gpu die. There's a reason why the highest performing laptops back then was SLI. It was easier to manage heat by having 2 die's help with the laptop. At this point for the 300w thing to even work you would need that liquid cooling shenanigan.

There is also no standard oems need to follow for least amount of wattage combined on load. Nor on temps and wattage. The way how laptops have different kinds of design dreams for different oems make this impossible.

[jaybee83]

1 hour ago, Bullit said:

You forgot: rent only.

 

aaah yes! basic display with wifi adapter connected to servers that do the actual computing. chomebooks FTW 😄 

[Bullit]

45 minutes ago, VEGGIM said:

Question. What laptops have existed that put out 300w on a single gpu die. There's a reason why the highest performing laptops back then was SLI. It was easier to manage heat by having 2 die's help with the laptop. At this point for the 300w thing to even work you would need that liquid cooling shenanigan.

There is also no standard oems need to follow for least amount of wattage combined on load. Nor on temps and wattage. The way how laptops have different kinds of design dreams for different oems make this impossible.

 

Indeed

 

My 3060 draws 130w and heats to 70ºC

My 5800H draws 80w and heats to 95ºC

 

Die size matters, too much concentrated stuff cannot be cooled cheaply.

 

[1610ftw]

25 minutes ago, VEGGIM said:

Question. What laptops have existed that put out 300w on a single gpu die. There's a reason why the highest performing laptops back then was SLI. It was easier to manage heat by having 2 die's help with the laptop. At this point for the 300w thing to even work you would need that liquid cooling shenanigan.

There is also no standard oems need to follow for least amount of wattage combined on load. Nor on temps and wattage. The way how laptops have different kinds of design dreams for different oems make this impossible.

 

Wattage for all cards would be max wattage, obviously manufacturers could choose to support less. The 4090 die would scale very well straight up to 300W so if a manufacturer finds a way to make it possible they will come out on top.

[VEGGIM]

2 hours ago, 1610ftw said:

 

Wattage for all cards would be max wattage, obviously manufacturers could choose to support less. The 4090 die would scale very well straight up to 300W so if a manufacturer finds a way to make it possible they will come out on top.

If im correct we call that dynamic boost. The issue is the moment we get that thinkness in a laptop is workstation level. But if the thickness gets to the point where you could find an itx console type system thickness, you start to have problems with the exception of not having a battery. No matter what you do you cannot break the laws of therodynamics. They will be obeyed wheter anyone likes it or not.

The reason why the thin and light market is starting to exist ore and more cuz alot of programs now that people are using don't need massive amounts of power and arent benchmark obsessed. The only way people would go back to wanting more thicker laptops, if for some strange reason most programs started to become obsessive unefficient power hogs.