Nvidia GeForce RTX 4090 Laptop vs. Desktop GPU

Even better, Nvidia should name all GPU's the same.
4090 96GB
4090 48GB
4090 32GB
4090 24GB
4090 20GB
4090 16GB
4090 12GB
4090 10GB
4090 8GB
4090 6GB
4090 4GB
Makes sense?

And then just have multiple 4090 8/12gb for all the mid tier models with the same amount of VRAM. Yea sounds like something nvidia could do! And as a shareholder I really hope they explore this venue of increasing profit margin. In fact I believe it to be gross negligence towards their customers (e.g. us shareholders) that they haven’t as of yet!
 
Laptop GPU nomenclatures have been a marketing scam for the last decade...I'm amazed they can get away with it, TBH.
Pascal was the last "good" Nvidia laptop generation. A desktop 1070 and a laptop 1070 were the same chip with tailored configurations for power envelope.

I'd like to see Tim twist the knife by putting the laptop "4090" against a real 4090 limited to 150W.
 
As expected, this is almost always the case: desktop>laptop

They did this on purpose to sell more laptops, the average consumer is ignorant and all they see is a name that's the same.
 
""give the laptop GPU a slightly different name. This should be an RTX 4090M – or even better, the RTX 4080M ""

nah ! it would mean, being honest, and it's about nvidia we're speaking here... LOL

That quote pretty much nails it.
 
It's literally sold as a '4090 Laptop GPU'. What's with the obnoxious take?

It has 4080 specced silicon in it.

The '4090' naming exists exclusively to obfuscate performance expectations in order to charge more money. Same reason Nvidia tried to name the 4070 Ti the '4080 12GB'. Exclusively to fleece people out of more money by using bigger number.
 
Last edited:
AD103 will be the top mobile GPU this generation. It is named the 4090 Laptop GPU.

The supposed confusion is a pretend problem.

I disagree.

It's intentionally labeled with a number to conflate performance with a much higher performing part, using a familiar numbering system that everyone uses to assess relative performance.

In the end it's just marketing a number and many clearly see through this obfuscation, but this is designed to take advantage of people who want high performance but are less technically savvy.

I appreciate the cleverness from a business perspective but also criticize it for it's intentional deception.
 
Would be more interesting if you compared it with RTX 4080 desktop since it's the same chip as RTX 4090 laptop so I hope you would compare them in a future article (or update this one).
 
It is performing on par with a 3080. All the results at 1080p and 1440p are useless due to driver overhead.
 
Would be more interesting if you compared it with RTX 4080 desktop since it's the same chip as RTX 4090 laptop so I hope you would compare them in a future article (or update this one).

You probably missed the part where it says that it is about 60% slower at 2160p in GPU bottleneck... this is 3080-3090 territory.
 
I don't get NVidia these days. Why don't they just call this thing then RTX 4080 mobile? Same core count, same amount of RAM, just clocked lower to get the power usage under control. I think calling this an RTX 4090 is deceptive (on the other hand... in a sense it doesn't matter, notebook parts are not going to hit the performance of the desktop equivalent whether you name it properly or not.)
 
They are basically just same parts, fed with different power, maybe nVidia should just make 4090 laptops be able to be hooked up a giant heatsink at home and be turned into a 4090 desktop
 
They are basically just same parts, fed with different power, maybe nVidia should just make 4090 laptops be able to be hooked up a giant heatsink at home and be turned into a 4090 desktop
That's the thing though, it's not the same part. The 4090 laptop is using the same chip design as a 4080 desktop (which is why I wonder why they didn't just call it a mobile RTX 4080 since really that's what it is?)

There are people who've done this in the past -- they'd take that mobile GPU (which back then was like a 40 or 50W GPU), overclock the piss out of it to get it up to desktop clockspeeds... and either add massive cooling to the system, or not do that and then wonder why it was crashing after about 30 seconds. Of course now, there's already a laptop "4090" spec'ed at 150W so I don't know how much higher you could really go -- I do wonder how many people overclocked those GPUs back then only to blow caps and power ICs on their laptops by drawing like 2x the power it was designed to supply.
 
That's the thing though, it's not the same part. The 4090 laptop is using the same chip design as a 4080 desktop (which is why I wonder why they didn't just call it a mobile RTX 4080 since really that's what it is?)

There are people who've done this in the past -- they'd take that mobile GPU (which back then was like a 40 or 50W GPU), overclock the piss out of it to get it up to desktop clockspeeds... and either add massive cooling to the system, or not do that and then wonder why it was crashing after about 30 seconds. Of course now, there's already a laptop "4090" spec'ed at 150W so I don't know how much higher you could really go -- I do wonder how many people overclocked those GPUs back then only to blow caps and power ICs on their laptops by drawing like 2x the power it was designed to supply.
Yeah, maybe they should just design it in a way that it can be fed with 400W, so you can still have desktop level performance at home just with your laptop, it would be really awesome.
 
Why aren't you testing the 4090 laptop against the 4080 desktop as they are the same card? Seeing how much effect the lower power envelope of the laptop part has on performance for other wise identical parts would be far ore interesting. Pointless to test it against a card it has nothing to do with.
 
Why aren't you testing the 4090 laptop against the 4080 desktop as they are the same card? Seeing how much effect the lower power envelope of the laptop part has on performance for other wise identical parts would be far ore interesting. Pointless to test it against a card it has nothing to do with.
I agree. But on the other hand if NVidia is going to market this (really) mobile RTX 4080 as a 4090, that makes it open to be compared against other 4090s.
 
I agree. But on the other hand if NVidia is going to market this (really) mobile RTX 4080 as a 4090, that makes it open to be compared against other 4090s.

Techspot users should be more savvy. Leave the marketing bs for engadget and the like.
 
Back