Nvidia Has Been Sneaking a Slower MX150 Variant Into Some Ultrabooks

rgMekanic

[H]ard|News
Joined
May 13, 2013
Messages
6,943
The good news just keeps coming for Nvidia. In an article from Notebookcheck, they are advising people purchasing an ultrabook with GeForce MX150 graphics to be on the lookout for an underclocked, less powerful GPU hiding under the same name, and not advertised as such. They have discovered two variants of the MX150 in various laptops, the standard has an ID of 1D10, and another with the ID of 1D12. The 1D12 has been found in 13 inch ultrabooks, and has a 36% lower base clock, and 20-25% less performance than the 1D10. Things like this make me wonder if someone thinks I'm an 1D10T.

While we don't have an issue with selling multiple variants of the same silicon, it's disappointing to not see any marketing material differentiating between the two versions especially since their performance deltas are more than marginal. Casual gamers or heavy users looking for full performance out of the MX150 should double-check new notebooks before committing to a full purchase.
 
I care about this and most people on this forum would care about this but the reality is, non IT people would never know or care. My dad or sister would just shrug I suspect if I told them their laptop video card was 25% slower than it was advertised to be.
 
FCUK nVidia. These constant deliberate actions to screw everyone while not disclosing, nor discounting the inferior iteration. Same reason I am not buying another Intel machine. Tired of the new processor? Then, buy a new motherboard and chipset, because, there is less money if we allow you to just use what you have already purchased even if it is pin compatible. Sorry, Rant.
 
This isn't another one of those where the branding is based on the SPECMark, is it? I read about something similar to this (from the very early 2000's) when both NVidia and ATI had a million different variants of their low end chips, and Radeon had three different memory architectures for the same (9500?) chip.

Anyway, I'm wondering if the 1D12's are all for 1366x768 screens where they can lower the power and the clock rate but still keep the same performance as a 1920x1080.


P.S. But in any case, fuck NVidia!

P.P.S. I am replaying Far Cry 4 to get ready for Far Cry 5, and I'd also like to say, "FUCK EAGLES!"

P.P.P.S. And fuck all those assholes who drive over the Sherpa while I'm running up to trade with him! Kyrat needs driverless cars.
 
The wonders of OEM selectable TDPs. I wonder when will [H]'s next target, Intel get roasted for that. AMD will get away scot-free, though.
 
meh, cudo's for lenovo getting by not labeling gpu part as a weaker chip.
 
This is directly related to the cost of the devices. Look it up, look up the price of the Lenovo and the Asus, and than bark at nV again. Yeah, at the moment they are shitheads and it's kinda modern to throw mud at them.

However, I am absolutely sure that this is OEM's choice ;)
 
This is directly related to the cost of the devices. Look it up, look up the price of the Lenovo and the Asus, and than bark at nV again. Yeah, at the moment they are shitheads and it's kinda modern to throw mud at them.

However, I am absolutely sure that this is OEM's choice ;)

I did a price on these just a moment ago and get your point. But the MX150 is in a TON of laptops, ultraportables and 2 in 1s so it is confusing branding.
 
SO that would put its performance below top end Ryzen APUs on mobile?


No ulterior motive here.
 
Pretty sure NVidia is not the problem here. OEM's know exactly what they buy. Now it is possible any given OEM could be slipped different parts and not be aware of it. All depending on how the quality assurance is setup at the OEM.

As much as I enjoy my tin foil hat (I do look smashing in it), I have to say there simply is not enough information to really know who did what to who
 
Pretty sure NVidia is not the problem here. OEM's know exactly what they buy. Now it is possible any given OEM could be slipped different parts and not be aware of it. All depending on how the quality assurance is setup at the OEM.

As much as I enjoy my tin foil hat (I do look smashing in it), I have to say there simply is not enough information to really know who did what to who

But the branding on the chip, that's on nVidia.
 
Whats the big deal.. It is just like thier 1060 part.. one twin is just a little slower, and cant remember as much as the other..
 
Pretty sure NVidia is not the problem here. OEM's know exactly what they buy. Now it is possible any given OEM could be slipped different parts and not be aware of it. All depending on how the quality assurance is setup at the OEM.

As much as I enjoy my tin foil hat (I do look smashing in it), I have to say there simply is not enough information to really know who did what to who

Where does the DeviceID come from?
Where does the default clock setting come from?

If both of these come from Nvidia then Nvidia is the problem. I notice the memory is Samsung vs Hynix (1D10 vs 1D12). I am guessing that lower speed memory is used in the 1D12 and that is the reason for the clock difference. If so then Nvidia is selling a product that is fundamentally inferior by design and should be sold using a different name to distinguish it from similar, but superior, products.
 
But the branding on the chip, that's on nVidia.

Where does the DeviceID come from?
Where does the default clock setting come from?

If both of these come from Nvidia then Nvidia is the problem. I notice the memory is Samsung vs Hynix (1D10 vs 1D12). I am guessing that lower speed memory is used in the 1D12 and that is the reason for the clock difference. If so then Nvidia is selling a product that is fundamentally inferior by design and should be sold using a different name to distinguish it from similar, but superior, products.

If the OEM specified they would accept slower parts with the new ID, then NVidia is in the clear. I have not seen anything from any OEM saying they have been screwed over, which would be the nail closing the coffin. The OEM could be the one screwing over its customers.

I am just playing devils advocate here, for the moment.

An independent site saying they are different is just that. It only shows different parts with the same family part number are being used in different laptops. Seems to be premature to jump on anyone yet.
 
I did a price on these just a moment ago and get your point. But the MX150 is in a TON of laptops, ultraportables and 2 in 1s so it is confusing branding.

As it is in a TON of laptops, why this is being discovered now? And why is there only this example ? To be honest, I can actually see Lenovo pulling a "quickee" on their customers as they have did numerous times before.

Don't get me wrong, the MX150 is nVidia branding and all of them should be the same. However, if an OEM comes to you and requests a lower TDP so they can "shove" it in their new precious that is supposed to go on the cheap...nah fuck it. Any self respecting human being will read what he is buying. Now, if there is no DATA about the chip itself being delivered to the customer from the OEM, than my dear this is called a Class A.
 
You honestly think its NVIDIA and not the OEM penny pinching on hardware lol.
 
is t he 150 actually fast enough to do any gaming where a slower version would affect it? I actually have one in my Dell but I'd never game on it just came with it.
 
Well.. We are fucked [H] until amd can compete....

AMD isn't immune from pulling bullshit either. Not too long ago they tried to stealth release a weaker card (less VRAM I believe) under the exact same name as a stronger card without any naming or numbering differences. Unfortunately, stuff like this is not all that uncommon.
 
AMD isn't immune from pulling bullshit either. Not too long ago they tried to stealth release a weaker card (less VRAM I believe) under the exact same name as a stronger card without any naming or numbering differences. Unfortunately, stuff like this is not all that uncommon.

I think it was a reduced core count rx560, but your point is well taken.

All we as consumers can do is call it out when we see it.
 
At the very least they could call the lower end product the MX145 or the higher end the MX150+. Anything to distinguish the two.

This is probably the most realistic approach. Obviously Nvidia probably figured there wasn't enough of a change to warrant this on their end. It's not like a GTX 1080 doesn't come in 100 flavors already, but most of the OEMs come up with their own markets scheme with hundreds of codes to try to signify the differences. If the actual chip itself is the same part, I don't think Nvidia is at fault at all if an OEM wants to underclock the part. They provide a reference clock, but it's not like 3/4s of the cards stick to that anyway. The difference here is that OEMs normally want to differentiate their product from stock because it's faster. There is no advantage to them coming out and saying they have a slower part than stock, so they don't market it as such. Instead they will just push that their battery life numbers are better than another OEM with the same specs.
 
Here's the thing about being top dog: failure is inexcusable. Success is expected. When you do fail, everyone starts to wonder where else you might be failing, and all of those seemingly innocuous decisions become massive foul-ups.

It might be time for Nvidia to hit reset and get back to when winning was a direct result of being better than the competition, and not just the windfall of previous successes.
 
Hmm I expect more from Nvidia to control their branding of the MX150, should really just brand differently like Thatguybil suggested. Though I would be interested to see which vendor has what?
 
I care about this and most people on this forum would care about this but the reality is, non IT people would never know or care. My dad or sister would just shrug I suspect if I told them their laptop video card was 25% slower than it was advertised to be.

Tell them in terms of $$monies$$. Now obviously a gpu isn't the whole system, so lets say 1/6th of a system, that -25% would be like $50-$100 difference compared to someone who paid the same amount for an identical system with the faster gpu. My dad would immediately complain about how that's like a months worth of gas money for him. Or tell me about how he once bought a complete sound system, new in box for $100 and still has and uses it to this day, some 50+ years later.
 
Same sort of shenanigans with regards to CPU on the ultrabooks (restricting TDP, possibly on the same ultrabooks).
 
Ultrabooks are only good for a few things no problem here really. People will buy anything based on reviews sight unseen.
 
Tons of tomfoolery goes on with nV in the mobile space in terms of rebranding, pricing, and now it seems, selling two different products under one model number.

My favorite, back in the day, was the 8800m GTX which magically became the 9800m GT (sold in the same OEM laptops)... the MXM card was hilariously priced at $150 more for OEMs. I guess changing an 8 to a 9 is worth $150 to NVidia.
 
  • Like
Reactions: N4CR
like this
This seems more like an OEM choice than NVIDIAs as they're the one ordering the parts. Or are we pretending the OEMs have been fooled by NVIDIA because they're just THAT EVIL? Come on, this shit's really grasping at straws now.
 
Last edited:
Ultrabooks have very limited TDP ceilings.

You want thin and light? You get it! You want top-end performance out of your variable-performance parts?

:ROFLMAO:
 
Back
Top