nvidia/OEMs sneaking slower-clocked MX150 variants into ultrabooks without disclosure..

[Picao84]

In the case of the 14CU RX560 they probably just had a bunch of RX460 chips laying around and the AIBs rebranded them as RX560.
We don't know in which hands these 14 CU chips were (AMD or AIBs) when they decided to sell the 14CU RX560, so you can't really trace the (original) fault back to AMD or AIBs.
You only know that AMD in the end allowed them to sell these cards as RX560, which IMO is bad enough.

Yes I can, AMD themselves admitted to doing it.

It’s correct that 14 Compute Unit (896 stream processors) and 16 Compute Unit (1024 stream processor) versions of the Radeon RX 560 are available. We introduced the 14CU version this summer to provide AIBs and the market with more RX 500 series options.

In the case of the 10W MX150 with 30% lower performance it's definitely not a case of laptop makers buying regular MX150 chips and downclocking them without nvidia knowing about it, because they have different chip IDs.

This is the GPU-Z reading for the normal MX150:

And this is the same reading for the lower-performing MX150:

nvidia is selling the "10DE 1D10" with very different base/boost clocks from the "10DE 1D12", as you can see.
So unless you think laptop makers are somehow changing the chip ID microcode, nvidia is definitely in on this.


And again: the timing of these laptops appearing with the 1D12 MX150 is very curiously matching the appearance of Raven Ridge solutions in the market.




The problem is not that nvidia released a 10W version of GP108 to counter Raven Ridge.
The problem is they're calling it MX150, so reviewers have been comparing Ryzen Mobile's gaming performance to laptops with the old, higher-performing and more power-consuming MX150.

That means squat. Even chips with exactly the same specs, including clock speeds, show different Device IDs on GPU-Z all the time. They just have to be manufactured later with small, meaningless revisions to end user, to have different IDs.

 

[Deleted member 13524]

Yes I can, AMD themselves admitted to doing it.

That doesn't answer the question of who had the 14CU chips in their hands when they decided to go forth with selling 14CU RX560.
Regardless, we all agree it was wrongdoing on AMD's part so that point is moot.

That means squat. Even chips with exactly the same specs, including clock speeds, show different Device IDs on GPU-Z all the time. They just have to be manufactured later with small, meaningless revisions to end user, to have different IDs.

According to notebookcheck's article, all 1D10 GPUs have 1469-1502MHz core clocks and all 1D12 GPUs have 937-1253MHz core clocks.

 

[Picao84]

The problem is they're calling it MX150, so reviewers have been comparing Ryzen Mobile's gaming performance to laptops with the old, higher-performing and more power-consuming MX150.
Had reviewers been comparing the Ryzen 2700U/2500U models with the lower-clocked 1D12 MX150 (which is the model that fits into the formers' power/heat demands) then the comparisons would probably be very different, as I pointed out in the first post.

Concerning that:
1 - There are so few laptops around with 2700U/2500U that the effect this can have is essentially meaningless.
2 - Regarding the effect, looking at benchmarks, the Ryzen 2700U is as fast as a GeForce 940MX. The original MX150 is around 50% faster than the 940MX. If the "new" MX150 is 30% slower than the original one, it will still be faster than the Ryzen 2700U.
3 - Given that is the really low end we are talking about here, those 30% will for sure not make the difference between a game being playable or not.

Yes, this is wrong, it should not happen, regardless if the fault is with NVidia, AIBs or both. However, given the really short potential for this machines to be used for anything but light gaming, I don't see it as bad as what AMD did with the RX560 (which was what made answer to this thread in the first place! Without that comparison I would not have bother to answer...).

 

[Picao84]

According to notebookcheck's article, all 1D10 GPUs have 1469-1502MHz core clocks and all 1D12 GPUs have 937-1253MHz core clocks.

https://zhuanlan.zhihu.com/p/28344462

̶A̶b̶o̶v̶e̶ ̶y̶o̶u̶ ̶c̶a̶n̶ ̶s̶e̶e̶ ̶n̶o̶t̶ ̶o̶n̶e̶,̶ ̶b̶u̶t̶ ̶s̶e̶v̶e̶r̶a̶l̶ ̶v̶e̶r̶s̶i̶o̶n̶s̶ ̶o̶f̶ ̶t̶h̶e̶ ̶M̶X̶1̶5̶0̶ ̶f̶o̶r̶ ̶d̶i̶f̶f̶e̶r̶e̶n̶t̶ ̶v̶e̶n̶d̶o̶r̶s̶,̶ ̶i̶n̶c̶l̶u̶d̶i̶n̶g̶ ̶t̶h̶e̶ ̶1̶D̶1̶2̶.̶ ̶T̶h̶e̶y̶ ̶a̶l̶l̶ ̶h̶a̶v̶e̶ ̶d̶i̶f̶f̶e̶r̶e̶n̶t̶ ̶d̶e̶v̶i̶c̶e̶ ̶I̶D̶s̶ ̶a̶n̶d̶ ̶d̶i̶f̶f̶e̶r̶e̶n̶t̶ ̶c̶l̶o̶c̶k̶ ̶s̶p̶e̶e̶d̶s̶.̶ ̶D̶o̶ ̶y̶o̶u̶ ̶t̶h̶i̶n̶k̶ ̶N̶V̶i̶d̶i̶a̶ ̶c̶r̶e̶a̶t̶e̶s̶ ̶d̶o̶z̶e̶n̶s̶ ̶o̶r̶ ̶h̶u̶n̶d̶r̶e̶d̶s̶ ̶o̶f̶ ̶S̶K̶U̶s̶ ̶e̶a̶c̶h̶ ̶w̶i̶t̶h̶ ̶d̶i̶f̶f̶e̶r̶e̶n̶t̶ ̶c̶l̶o̶c̶k̶ ̶s̶p̶e̶e̶d̶s̶ ̶f̶o̶r̶ ̶e̶a̶c̶h̶ ̶v̶e̶n̶d̶o̶r̶?̶ ̶I̶t̶ ̶i̶s̶ ̶w̶a̶y̶ ̶m̶o̶r̶e̶ ̶b̶e̶l̶i̶e̶v̶a̶b̶l̶e̶ ̶t̶h̶a̶t̶ ̶O̶E̶M̶s̶ ̶t̶h̶e̶m̶s̶e̶l̶v̶e̶s̶ ̶c̶u̶s̶t̶o̶m̶i̶z̶e̶ ̶t̶h̶e̶m̶ ̶f̶o̶r̶ ̶t̶h̶e̶i̶r̶ ̶d̶e̶s̶i̶g̶n̶ ̶t̶a̶r̶g̶e̶t̶s̶,̶ ̶s̶i̶n̶c̶e̶ ̶N̶V̶i̶d̶i̶a̶ ̶d̶o̶e̶s̶ ̶n̶o̶t̶ ̶d̶e̶s̶i̶g̶n̶ ̶l̶a̶p̶t̶o̶p̶ ̶c̶h̶a̶s̶s̶i̶s̶ ̶a̶n̶d̶ ̶c̶o̶o̶l̶i̶n̶g̶ ̶s̶y̶s̶t̶e̶m̶s̶.̶

EDIT - I misread the link above (its Chinese, so translation is rough).

After further investigation on some reviews, I do accept the 1D12 is NVidia's own specs provided to OEMs.

 

[CSI PC]

Anyone have a few benchmark reviews for gaming performance and battery life just to provide some reference between 2500u and MX150 generally.
Not sure if I am reading this correctly but it seems the 1D12 MX150 still has a notable performance advantage in games over the 2500u Vega, just saying from a performance perspective.
Something that is really important is do partners charge the same for both types of MX150 (bad) or used this as a way to increase pricing (even worst).

1d12 MX150 review: https://www.notebookcheck.net/Asus-ZenBook-13-UX331UN-i7-8550U-MX150-Laptop-Review.277586.0.html
2500u Swift3 review: https://www.notebookcheck.net/Acer-...Vega-8-256-GB-FHD-Laptop-Review.277262.0.html

Unfortunately can only compare 2 games though; Bioshock Infinite, The Witcher 3.
And more unfortunate for some reason their review for both is without power consumption while importantly running Witcher 3; would had been nice to see as part of gaming performance/envelope behaviour.

Anyway, is there an easy way to try and see how the two MX150 models influenced partner's model pricing as that could also be a negative for consumers.

 

[Deleted member 13524]

Anyone have a few benchmark reviews for gaming performance and battery life just to provide some reference between 2500u and MX150 generally.
Not sure if I am reading this correctly but it seems the 1D12 MX150 still has a notable performance advantage in games over the 2500u Vega, just saying from a performance perspective.

Say no more fam!
I was just working on it.

These come from the results of the Asus UX331UN, the Acer Swift 3 with 15W TDP and the X360 with 25W mXFR:

The low bandwidth of the APU hits differently depending on which game.
Tomb Raider 2013 doesn't get great results on the APUs across the board compared to the MX150
Bioshock Infinite seems to favor the Vega 8 at lower settings, eventually losing to the MX150 with dedicated memory at higher (though unplayable) settings.
Witcher 3 looks like almost a wash between the 25W mXFR and the MX150 1D12.
Metal Gear Solid V is a wash at low settings (probably frame-capped) and at high it loses to the dGPU.
Evil Within loses.
CoD WW2 wins.
NFS and Battlefront 2 loses.


One problem with these benchmarks is they were taken at the time of release of each laptop, which was 1 month ago for the Asus/MX150 and December/January for the other two using release drivers.
Moreover, AMD for some reason isn't supporting Raven Ridge iGPUs with the Adrenalin driver releases at the moment, so the Ryzen Mobile laptops are all still using their release drivers.

There are so few laptops around with 2700U/2500U that the effect this can have is essentially meaningless.

IIRC, AMD claimed a total of 60 design wins on laptop designs until the end of the year.

Regarding the effect, looking at benchmarks, the Ryzen 2700U is as fast as a GeForce 940MX. The original MX150 is around 50% faster than the 940MX. If the "new" MX150 is 30% slower than the original one, it will still be faster than the Ryzen 2700U.

See above.

Given that is the really low end we are talking about here, those 30% will for sure not make the difference between a game being playable or not.
(...)
However, given the really short potential for this machines to be used for anything but light gaming

What I see here is your own personal lack of interest in gaming on portable and light devices standing in the way of your judgment in the matter.

I'm all in for a gaming capable 13" 2-in-1 with the gaming capabilities of a 2500U.

Above you can see not one, but several versions of the MX150 for different vendors, including the 1D12.

Nope. That link only shows 1D10 and 1D12 for the MX150. If you use a translator, you'll see the other GPU with a different ID refers to a desktop GT1030.

 

[Picao84]

IIRC, AMD claimed a total of 60 design wins on laptop designs until the end of the year.

Nvidia also claimed to have lots of design wins for Tegra haha! I don't see many 13 inch laptops on the market place with Ryzen U. I waited a little bit before settling on Dell XPS 13 and nothing I've seen with it was worth it (heavy and smaller batteries). Not that I care much as a decent GPU was not on the top of my priorities.

What I see here is your own personal lack of interest in gaming on portable and light devices standing in the way of your judgment in the matter.

I'm all in for a gaming capable 13" 2-in-1 with the gaming capabilities of a 2500U.

Fair enough. I'll just add: 2-in1 eeewww, kill them with fire! I had one and found it pointless, unless you use a pen to draw or something like that. I'm definitely old school on my tastes/needs (who needs a touch screen to code really?) .

Nope. That link only shows 1D10 and 1D12 for the MX150. If you use a translator, you'll see the other GPU with a different ID refers to a desktop GT1030.

Yes I know, I noticed that before your reply and edited my post.

 

[Deleted member 13524]

Picked up by Anandtech:

https://www.anandtech.com/show/12565/nvidia-silently-rolls-out-slower-mx150-for-ultrabooks

 

[Picao84]

Picked up by Anandtech:

https://www.anandtech.com/show/12565/nvidia-silently-rolls-out-slower-mx150-for-ultrabooks

A comment there corroborates how hard is to find Ryzen U laptops

- Only Ryzen option is the ACER in my country and costs almost $990! (800 Euro). And that's for the 2500U.

 

[Deleted member 13524]

Extremetech
https://www.extremetech.com/computing/266286-nvidia-mx150-gpu-massively-slowed-laptops

Tom's Hardware
http://www.tomshardware.co.uk/nvidia-geforce-mx150-two-variants,news-58100.html

 

[CSI PC]

Has any articles that found both examples mentioned whether the same model names (such as the Asus Zenbook13 UX430UN) are same price or different depending upon which MX150 is used.
This is the real news IMO if Asus is selling both versions at same price or worse even used it to increase prices, or if a model (such as the Asus UX430UN) has both 1D10 and 1D12 available going forward.
Would be less of an issue/news item if they price the 1D12 cheaper (still consumer confusion though) and less so if also combined with the move away from 1D10 to 1D12 in those models in future .

Shame they are only reaching out to Nvidia and not also the AIB partners doing this.
Journalists will be more likely to get answers from the AIB partners doing this rather than from Nvidia supplying them with the part.
I bet Nvidia will not be able to share much more information due to the partner agreements between them and the AIB partner, although maybe going forward they will force it to be specc'd like Max-Q meaning less flexibility by AIB partners - would be the ideal situation for consumers going forward and will be interesting to see if it ends up going that way, albeit with less flexibility on efficiency/battery life by AIB partner's product but worth it for greater transparency.

Edit:
Just wondering Nvidia has recently manufactured the 1D12 "Max-Q" type version of the MX150 and partners/OEM are using this to replace the existing 1D10 in 13" size or smaller but instead of renaming the models they just kept as is, naughty if so by the partners.
All depends if there is a mix of 1D10 and 1D12 still available going forwards in use at 13" for a specific model such as the UX430UN or lower to tell what the action is about, but still should be renamed product if they are moving to 1D12 as a replacement or an alternative at same size.
And could be argued if they are replacing with the 1D12 Max-Q type it should possibly be cheaper, maybe part of the reason the partners/OEM are not renaming the models to keep current prices on the quiet.

However and importantly this situation could had been totally blocked if Nvidia had specc'ed/branded the 1D12 MX150 in similar way they do for trational 'Max-Q'.
So they do share some of the blame but how much is to be seen as it could come back mostly to the partner/OEM.

 

[Deleted member 13524]

Notebookcheck updated their article with the news of another model caught with the 1D12 variant, the ThinkPad T480.
https://www.notebookcheck.net/Nvidi...12-variant-onto-some-Ultrabooks.289358.0.html

Still no official response from nvidia anywhere, AFAIK.

A comment there corroborates how hard is to find Ryzen U laptops

There are plenty of other threads with information/discussion about Ryzen Mobile laptops where you can complain about the lack of a product for which you openly declared you have no interest in.
You could even start a thread about the lack of different laptops with Ryzen Mobile if you want.

Shame they are only reaching out to Nvidia and not also the AIB partners doing this.

Did "they" reach AMD's partners when the 14CU RX560 cards started to come up?
It was AMD who had to apologize for the AIBs not disclosing the CU count in the end.
Which is correct. AMD should have created a different name for the 14CU part. AMD didn't and that fault is on them.

However and importantly this situation could had been totally blocked if Nvidia had specc'ed/branded the 1D12 MX150 in similar way they do for trational 'Max-Q'.
So they do share some of the blame but how much is to be seen as it could come back mostly to the partner/OEM.

nvidia sells two GPUs to OEMs with a very distinct real-world performance but exact same marketing name. But you say most of the blame is on the OEMs, not nvidia.
And you think the OEMs should be the ones being asked for clarification, and not nvidia who is selling the 1D10 and 1D12 models that carry the exact same MX150 name.

The obvious reaction here is to ask nvidia why they're selling two different GPUs with distinct performance targets under the exact same name.
All OEMs do is buy the GPUs from nvidia, stick them into their motherboards and cooling solutions according to the spec sheet and name them the name nvidia tells them to put there.
It's not Asus' job to put a sticker in the UX331 saying "MX150 slower edition". It's nvidia's job to name it differently, like MX150 Max-Q or MX140.

Your opinion on the matter is the complete opposite of every single tech-news/gaming/enthusiast outlet.

And then you act all surprised when users accuse you time and again for being an apologist.
¯\_(ツ)_/¯

 

[Picao84]

There are plenty of other threads with information/discussion about Ryzen Mobile laptops where you can complain about the lack of a product for which you openly declared you have no interest in.
You could even start a thread about the lack of different laptops with Ryzen Mobile if you want.

My comments on Ryzen follow exclusively your conspiracy theory about this laptops showing up after Ryzen 2500u / 2700u were launched. You were the one bringing them into this discussion, not me. As this is still a discussion, I limited myself to point out that availability of such laptops is limited to begin with, so there is not much ground to claim AMD Ryzen U laptops are being hurt by this, as they are hard to get in the first place. It is not my fault that reality does not follow your theories

 

[CSI PC]

Your opinion on the matter is the complete opposite of every single tech-news/gaming/enthusiast outlet.

And then you act all surprised when users accuse you time and again for being an apologist.
¯\_(ツ)_/¯

There you go again...
I know Jack Russell Terriers that ankle bite less than you

 

[Picao84]

There you go again...
I know Jack Russell Terriers that ankle bite less than you

He is right though, nVidia is at fault here.

 

[Deleted member 13524]

My comments on Ryzen follow exclusively your conspiracy theory about this laptops showing up after Ryzen 2500u / 2700u were launched.

The 1D12 did appear at the same time as the Ryzen Mobile laptops started to come up. Ryzen Mobile laptops are benched side-by-side with the MX130/MX940 and the MX150.

It's perfectly expected for nvidia to come up with a discrete GPU that competes with the 15-25W Ryzen Mobile APUs in perf/Watt when paired with a Core U.
Companies launch reactionary products all the time, that's not a conspiracy.

It's calling it the same name as a significantly more powerful and power-consuming GPU that is misleading. It misled reviewers to compare the Raven Ridge-powered Acer Swift 3 and HP X360 side-by-side with laptops using the 1D10 MX150 instead of the 1D12 version it's meant to compete with.

I limited myself to point out that availability of such laptops is limited to begin with, so there is not much ground to claim AMD Ryzen U laptops are being hurt by this, as they are hard to get in the first place.

Despite your so-called "proof" from a single comment in the internet, availability isn't limited. You can find the Swift 3 (15W) and X360 (25W mXFR) in all the key markets right now.
Besides those, there are also confirmed:
- Lenovo Ideapad 720S (to avoid because it uses only single-channel memory)
- Lenovo Thinkpad A275, A475, E485, E585
- Acer Aspire 3
- HP Elitebook 735 G5, 745 G5
- Dell Inspiron 5575, Inspiron 17 5000
- Smach-Z gaming console

AFAIK this is already quite a bit more than what Carrizo and Bristol Ridge could get in their lifetimes, and there's little reason to doubt Lisa Su's statements about Ryzen Mobile getting several dozens of design wins throughout 2018.

 

[CSI PC]

(March 25, 2018 update: The ThinkPad T480s has been discovered shipping with the slower GeForce MX150 Max-Q GPU. This system has been added to our 3DMark benchmark list below

So this seems to be more looking like the 1D10 is being replaced with the 1D12 variant without changing model name/structure.
Which is one aspect I raised as a possibility in my post.

Just wondering Nvidia has recently manufactured the 1D12 "Max-Q" type version of the MX150 and partners/OEM are using this to replace the existing 1D10 in 13" size or smaller but instead of renaming the models they just kept as is, naughty if so by the partners.
.....
And could be argued if they are replacing with the 1D12 Max-Q type it should possibly be cheaper, maybe part of the reason the partners/OEM are not renaming the models to keep current prices on the quiet.

However and importantly this situation could had been totally blocked if Nvidia had specc'ed/branded the 1D12 MX150 in similar way they do for trational 'Max-Q'.
So they do share some of the blame but how much is to be seen as it could come back mostly to the partner/OEM.

Nvidia does share some of the blame like I mentioned but different POVs/opinion will decide how much (personally I think the partners had too much leeway and made the most of it), and yeah Nvidia should apologise as well tbh like AMD did.

 

[Picao84]

Despite your so-called "proof" from a single comment in the internet, availability isn't limited. You can find the Swift 3 (15W) and X360 (25W mXFR) in all the key markets right now.
Besides those, there are also confirmed:
- Lenovo Ideapad 720S (to avoid because it uses only single-channel memory)
- Lenovo Thinkpad A275, A475, E485, E585
- Acer Aspire 3
- HP Elitebook 735 G5, 745 G5
- Dell Inspiron 5575, Inspiron 17 5000
- Smach-Z gaming console

AFAIK this is already quite a bit more than what Carrizo and Bristol Ridge could get in their lifetimes, and there's little reason to doubt Lisa Su's statements about Ryzen Mobile getting several dozens of design wins throughout 2018.

I did not say it was proof, it simply echoes my own experience. As I said before, I was looking forward to laptops with Ryzen U (go and check my posts on the thread about Ryzen Mobile) before getting a Dell XPS 13 with Core i7 8550u. I did a Google / Amazon search everyday and the most I could find were the 3 laptops AMD showed at the 2500u/2700u presentation. Compared that with a Google / Amazon search for 8250u and 8550u and compare. Night and Day difference in availability.

Carrizo is not a good point of comparison for anything really. If that is your target to beat, then you all but prove my point.

A search for 2500u on Amazon UK turns out three lonely results! A search for 2700u brings ONE single result!
A search for 8250u brings 7 pages of results. A search for 8550u brings 3 pages of results on a grid (rather than listed) with 24 laptops per page!

Its not even a competition, AMD does not have even 10% of the laptops Intel has with 8250U and 8550u!

But this is not the thread to discuss this, I give you that, and that is why this is my last post on the matter.

 

[Deleted member 13524]

So this seems to be more looking like the 1D10 is being replaced with the 1D12 variant without changing model name/structure.

Oh for crap's sake...
Just a couple of days ago notebookcheck reviewed two new laptops with the 1D10 MX150: HP Envy 17 and Zenbook UX430UN.

You keep trying to find ways to excuse nvidia from the fact that they're selling right now, at the same time, two different GPUs with different performance levels while carrying the exact same name.
And now you theory is nvidia is simply replacing the 1D10 with the 1D12 and the OEMs are the baddies who took advantage of that to underclock their GPUs.

Don't you see how desperate that sounds?!

 

[CSI PC]

The Zenbook UX430 with MX150 is not new, launched before 2018.
When did 1D12 come available to market as launch, asking as I am not sure myself.
Even if the UX430 was swapping to 1D12, you would have a mix of old/new stock for a little while.
Logically, why would an OEM complicate their logistics and roll a dice to decide which MX150 variant they implement for same model number going forward.
Would they mix both variants into the same logistics-manufacturing pool, could be possible they do not care (but it changes the spec-behaviour-support of their product) - which brings it back to journalists engaging the partners for their side of what is happening.