NotebookCHECK - Notebook Forum

There are 28 variants of GeForce RTX 30 Laptop GPUs - VideoCardz.com
videocardz.com/newz/there-are-28-variants-of-geforce-rtx-30-laptop-gpus

>Laptop makers have ignored NVIDIA's request to list TGP values on their websites. As a result, consumers are faced with a lottery of 28 different SKU

After reading this: https://www.notebookcheck.net/Buyer-beware-Nvidia-has-at-least-28-variants-prepared-for-the-mobile-RTX-3000-GPU-stack.517112.0.html

The actual article sounds like payed content to me... Especially "argument" No. 2: Mobile GeForce Names Will Be Much Simpler For Laymen Shoppers.

I wonder who inspired the author to write such a thing. De facto it is not (NOT) getterin any simpler - only more obscure!

best regards

phila

P.S.: I am thinking of linking this article to my main tech forum, just to make fun of this kind of argumentation.

Quote from: phila_delphia on January 28, 2021, 07:46:21
After reading this: https://www.notebookcheck.net/Buyer-beware-Nvidia-has-at-least-28-variants-prepared-for-the-mobile-RTX-3000-GPU-stack.517112.0.html

The actual article sounds like payed content to me... Especially "argument" No. 2: Mobile GeForce Names Will Be Much Simpler For Laymen Shoppers.

I wonder who inspired the author to write such a thing. De facto it is not (NOT) getterin any simpler - only more obscure!

best regards

phila

P.S.: I am thinking of linking this article to my main tech forum, just to make fun of this kind of argumentation.

We are legally obligated to tell you if an article is sponsored or paid. We're also well aware of the cons which we will get to in another article as mentioned :)

One of these bad notebookcheck articles again. Let's destroy all of these arguments, shall we?

1) some max q variant beats non max q counterparts.
-I'm sure comparing a high-end 17-inch laptop with a low to mid-end 15-inch laptop is a fair comparison. Just compare the max q average with the non max q averages and we shall see.
-of course, there will be exceptions when you have such a big variety of laptops out there. finding one case means absolutely nothing. if this happens in every brand then this might be an argument.

2)It would be simpler if there were fewer variants
-let's first break the "evidence." article claims 4 2080 super variants and it's confusing. I agree but it's only because Nvidia brought the super series to laptops. why not blame Nvidia bring out a new GPU that only got us 5-10% performance instead of blaming the branding that separated parts that have a higher % difference?
-the other comments already said this but there an article that claims that the 30 series still have 28 variants of the 3000 series. how are you going to tell them apart without the branding without OEMs telling you?
-so it is confusing to laymen buyers but the pc enthusiast community has grown monumentally during 2020 and it is the buyer's responsibility to know what they are buying. why not label everything laptop and call it a day? why punish the people who actually have knowledge? why is giving a hint to the performance characteristics of a chip a bad thing? the OEM clarification of the TGP is bad enough. how are you going to tell chips apart without reviews? (at least max q helped)

3) Hasten production and reduce QA costs
-do you see laptops decreasing prices? Nvidia and OEMs now have more margin.
-"Removing Max-Q would ... give more freedom to OEMs when it comes to adjusting GPU clock rates, TGP, fan noise, and chassis design."
this makes it worse for people to tell the performance of a chip without a detailed review. why are we, again, sacrificing the knowledge to those that do not do research on a product? If you are paying over 1500 dollars for a product, it should be your responsibility to use the money wisely and know what you are buying. Making it harder for everyone is not a solution and deserves no excuses.

Come on! -- so they had to flood the market with 28 "variants" of mobile GPUs? It doesn't help the consumer even a little bit, a complete mess. So now before purchasing a laptop one would have to sort what should have been sorted out by (scam)NVIDIA. It feels like Intel and NVIDIA scheme is to confuse the buyers and reviewers like you, Allen, so that no one literally understands what chips are what.

Nvidia's line up is still a mess but a different kind of mess. Now we have to do even more homework before buying.

Quote from: phila_delphia on January 28, 2021, 07:46:21
The actual article sounds like payed content to me... Especially "argument" No. 2: Mobile GeForce Names Will Be Much Simpler For Laymen Shoppers.

Perhaps a problem lies in your understanding. The problem won't get any simpler, but naming will (assuming they truly drop Max-Q, which still figures in the table in that article) - by obfuscating the detail. The real problem, uncertain performance, hasn't disappeared and I would argue that it hasn't gotten any bigger either. Yes, it looks like there is a bigger range of TGPs to choose from (I don't know how many there were before), but the fundamental limitation is cooling.

It would be of more use to me if a manufacturer could specify how many watts can the cooling system remove from GPU/ CPU under steady state conditions in a given environment and at what die temperature and noise. Sure, they could use a weaker version of a GPU than what the chassis can support (the only good reason I can think of is to ensure big enough steps between 3060/ 3070/ 3080). TGP would be interesting as a secondary information (they could even offer multiple TGPs). But the more common problem is the opposite. That the cooling capacity is the actual limiting factor. I won't even get into the fact that different manufacturers can target different temperatures and noise levels and that you might wonder how will they fare if you tune them to a similar level that suits you.

And it's not really as simple as that. Because CPU and GPU typically don't have completely separate cooling systems. At least they're linked by a heat pipe. It's not an unsolvable problem. But I'm looking at it as someone who can read a datasheet and do basic thermal design calculations. In the end, it might be most consumer friendly to simply advertise frame rates/ benchmark results that you should see (given configuration and settings). Both on battery and mains power. And again, the laptop should be properly heated up. If it doesn't meet it by a significant margin (around here, standard practice is 10 %), it's defective and should be returned.

Personally, I have tried elevating myself above it and simply accept that I might lose money. It doesn't have a good solution and it just brings you stress and you waste your time and energy. I say, get a good return policy.

Have you considered the possibility that some laptops might lack Max-Q branding because they don't have the certification?

I find this recent trend where you don't know what you are buying decisively anti-consumer. First Intel with their ambiguous Tiger Lake, now Nvidia... I hope something will be done to stop this shady marketing.

This is a terrible article that sounds like paid damage control. Since other comments have already done a great job rebuking every single point I won't bother, but this stuff is shameful, truly

Quote from: Allen.Ngo on January 28, 2021, 07:54:14

Quote from: phila_delphia on January 28, 2021, 07:46:21
After reading this: https://www.notebookcheck.net/Buyer-beware-Nvidia-has-at-least-28-variants-prepared-for-the-mobile-RTX-3000-GPU-stack.517112.0.html

The actual article sounds like payed content to me... Especially "argument" No. 2: Mobile GeForce Names Will Be Much Simpler For Laymen Shoppers.

I wonder who inspired the author to write such a thing. De facto it is not (NOT) getterin any simpler - only more obscure!

best regards

phila

P.S.: I am thinking of linking this article to my main tech forum, just to make fun of this kind of argumentation.

We are legally obligated to tell you if an article is sponsored or paid. We're also well aware of the cons which we will get to in another article as mentioned :)

I know it is not paid content  ;)
Yet the argument (especially No.2) seems so (SO) odd that it SEEMED to ME like it was.

Why not say: Removing the numbers from the cards leaving only the RTX would make the naming even more simple  :D

I mean come on...

And while we are at it. Let's give the CPUs the same treatment naming wise. Then the consumer could opt for a Notebook with RYZEN (any) with RTX (any) or Intel (any) with RTX (any).

Really simple! You just have to inform yourself a little better or, like the previous speaker said, check the return policies.

best regards

phila

Why stop at removing Max-Q?
1. Some GeForce Max-Q GPUs Were Already Outperforming Their Non-Max-Q Counterparts
A 3080 80W is outperformed by a 3060 120W and that in turn is outperformed by a high end 2080, lets get rid of all those number and stop the confusion.

2. Mobile GeForce Names Will Be Much Simpler For Laymen Shoppers
See above.

3. It Will Likely Hasten Production And Reduce QA Costs For Both Nvidia and OEMs
Image OEMs could swap in and out GPUs as they like, what a boon for production. Bonus points for shipping the faster GPUs to reviewers and use slower ones for customers units. Incredible cost savings possible!

Quote from: Atavism on January 28, 2021, 13:06:273. It Will Likely Hasten Production And Reduce QA Costs For Both Nvidia and OEMs
Image OEMs could swap in and out GPUs as they like, what a boon for production. Bonus points for shipping the faster GPUs to reviewers and use slower ones for customers units. Incredible cost savings possible!

Mate, you almost killed me -> death by laughter  ;D
I`d nerver thought about this option.
And I sincerely hope that you did not inspire manufacterers.
Yet if so, the should hand over some of the profit to you!

All the best

phila

Quote from: dude what on January 28, 2021, 13:01:18
This is a terrible article that sounds like paid damage control. Since other comments have already done a great job rebuking every single point I won't bother, but this stuff is shameful, truly

I was gonna write something too but yes everybody else said it already.
And the opinion/column declaration was omitted.

I'd be really interested in an article on Whisper Mode 2.0 when you have a selection of 30-series laptops to compare.

Specifically, if I want a laptop with decent performance that doesn't make too much noise, should I buy a laptop with a low TDP (traditional Max-Q design) or one with a high TDP & configure it to run at lower performance in the software?

While I'm not looking forward to the confusion about what power a given GPU can support, the potential of a Max-P laptop with a Max-Q option when needed sounds perfect.

Classic allen ngo giving me a good laugh and disinforming every living creature on Earth

Tbh, partially I still visit this website because of his carptastic articles. Good job

One thing that a lot of angry people in this thread don't understand is that gpu barely guarantees anything by itself. The only thing we should care about is actual level of performance we get for our money. And this depends mainly on two factors: target temperature (cpu, gpu, and chassis itself) set  by manufacturer and corresponding noise level generated by fans (basically level of heat dissipation by particular chassis).

That is why these two mentioned laptops are great example:
comparable noise level during gaming: 54.1 db for Asus and 53 db Razer, and surface temperature 34.1 °C Asus and 35.4 °C Razer. But they deliver around the same fps numbers and average gpu core frequency, even though Razer employs 100W version of gpu and low-power gpu memory, and Asus instead features full 115W mobile 2070. The only difference is in average gpu temperature (74 vs 70: this allow nvidia gpu to boost higher), that is why lower temperature on Razer and less memory energy budget catch up with 15W tdp deficit. And, of course, the fact that Razer was tested with optimus disabled.

This is also why best implementations of 2060 115W like MSI GE75 Raider 10SE and cheap Walmart EVOO Gaming 17 perform on par with at least two laptops with full 2070 like ASUS ROG Strix SCAR III G731GW and Asus Strix Scar III G531GW, and also better than at least 15 others with 2070 Max-Q.

THIS IS A MESS. Consumers are just duped by manufactures every god damn time. It's like laptops manufacturing is an ART and no one can judge them objectively with a tangible score. The one we see here, at notebookcheck, is barely ok. What all these numbers tell us? Nothing really. How would one buy just the best ones?

I wonder when reviews would stop being just a stupid flavor reviews. There is a clear need to devise a proper comparison methodology. No amount of certification from nvidia with labels and perceived QA will guarantee anything. We may only know this for sure after proper analysis is performed and not just numbers are measured.

So it is going to be interesting when this all will be understood and implemented fully. Nvidia's policy change just forces this and opens our eyes on how much trust we should have for manufacturers really.

ComputerBase (those racist kebab heads who delete your comments if it contains China, Taiwan, freedom, independence or democracy) claim that the minimum for CPUs in laptops for gaming is 6 cores, better 8.

Maybe the 8 core i7 could process more GPU commands thanks to higher bandwidth?

Well, given that apparently this is officially not paid content, I guess it's just an example of some painfully bad motivated reasoning.

Nvidia abandoning Max-Q branding would only improve things if they had replaced it with a way for the customer to actually know what they're getting - only, they haven't. Instead, they've abdicated any duty of care to the end user and made an already-bad situation worse.

Previously the "Max-Q" branding at least acted as a kind of warning that you were getting an overpriced chip that would under-perform for its class. The existence of "Max-P" designs with insufficient power delivery and cooling to reach their rated performance complicated things, but it was entirely within Nvidia's power to implement some regulations that would prevent that.

Instead, they've gone and made it clear that they don't give a s*** about the experience of the end-user. At least they aren't trying to dress it up as an improvement, unlike Allen Ngo. 🤷‍♂️

Class masterstroke from Allen. I come to NBC only to read funny and retarded articles. No serious content here.

It is tough to find a flip-flop artist and apologist like Allen. Congratulations Nvidia, you found your guy, who is trolling others on your behalf, for free!