NotebookCHECK - Notebook Forum

Dell's CAMM in 2023 in Samsung and only a year later in Lenovo? 😃

P.S. The news is really good. Laptop and PC manufacturers are switching to more efficient memory with easy replacement*

Now let's move on to things that are much more unpleasant for the consumer:
1. 32GB 5600 module costs $210. Despite the fact that DDR5 5600 costs $97. From the same company Crucial (for example).
2. These modules DO NOT provide any real significant speed advantages.
3. The declared lower consumption with the DO prefix must be proven in practice. In fact, the module consumption will change by 1-5% of the SoC consumption.

What do we actually have? That manufacturers are colluding to switch to a rare and expensive (at retail prices) type of memory under the pretext of "optimization," which does not give anything, but reduces competition by an order of magnitude, because these modules are practically not on sale in any store with rare exceptions and the prices are 2 times higher than now; besides, vile and greedy laptop manufacturers can justify increased prices for their models, where even with DDR4/DDR5 the prices for memory modules were inflated 3-4 times, like the prices for SSDs. Moreover, both memory modules and ssds come with a general warranty as part of the laptop, which is most often 1 year and rarely 2-3 years. Despite the fact that in retail most DDR4/5 memory modules come with a 5-10 year warranty, and 5 summer on SSDs of all leading brands at a much lower price than as part of a laptop.

Do you still need these stupid and super expensive LPCAMM2 modules?

Quote from: NikoB on May 08, 2024, 16:53:11Now let's move on to things that are much more unpleasant for the consumer:
1. 32GB 5600 module costs $210. Despite the fact that DDR5 5600 costs $97. From the same company Crucial (for example).
2. These modules DO NOT provide any real significant speed advantages.
3. The declared lower consumption with the DO prefix must be proven in practice. In fact, the module consumption will change by 1-5% of the SoC consumption.

What do we actually have? That manufacturers are colluding to switch to a rare and expensive (at retail prices) type of memory under the pretext of "optimization," which does not give anything, but reduces competition by an order of magnitude, because these modules are practically not on sale in any store with rare exceptions and the prices are 2 times higher than now; besides, vile and greedy laptop manufacturers can justify increased prices for their models, where even with DDR4/DDR5 the prices for memory modules were inflated 3-4 times, like the prices for SSDs. Moreover, both memory modules and ssds come with a general warranty as part of the laptop, which is most often 1 year and rarely 2-3 years. Despite the fact that in retail most DDR4/5 memory modules come with a 5-10 year warranty, and 5 summer on SSDs of all leading brands at a much lower price than as part of a laptop.

Do you still need these stupid and super expensive LPCAMM2 modules?

chill out, it'll get cheaper with time, and also, chinese factories will be pumping them out in a bit, there's no lock in...any removable , upgradable component is better than none

Quote from: Alexander_ on May 08, 2024, 16:33:36Dell's CAMM in 2023 in Samsung and only a year later in Lenovo? 😃

P.S. The news is really good. Laptop and PC manufacturers are switching to more efficient memory with easy replacement*

That was CAMM1, which was Dell's proprietary thing, after Dell sent it for standardization, CAMM2 was made which is what is in this laptop

Quote from: NikoB on May 08, 2024, 16:53:11Now let's move on to things that are much more unpleasant for the consumer:
1. 32GB 5600 module costs $210. Despite the fact that DDR5 5600 costs $97. From the same company Crucial (for example).
2. These modules DO NOT provide any real significant speed advantages.
3. The declared lower consumption with the DO prefix must be proven in practice. In fact, the module consumption will change by 1-5% of the SoC consumption.

What do we actually have? That manufacturers are colluding to switch to a rare and expensive (at retail prices) type of memory under the pretext of "optimization," which does not give anything, but reduces competition by an order of magnitude, because these modules are practically not on sale in any store with rare exceptions and the prices are 2 times higher than now; besides, vile and greedy laptop manufacturers can justify increased prices for their models, where even with DDR4/DDR5 the prices for memory modules were inflated 3-4 times, like the prices for SSDs. Moreover, both memory modules and ssds come with a general warranty as part of the laptop, which is most often 1 year and rarely 2-3 years. Despite the fact that in retail most DDR4/5 memory modules come with a 5-10 year warranty, and 5 summer on SSDs of all leading brands at a much lower price than as part of a laptop.

Do you still need these stupid and super expensive LPCAMM2 modules?

New stuff costing more is nothing new, DDR5 was also more expensive when it came out. You are paying the upfront R&D and equipment cost until it scales

Weren't you all about increasing memory width? I guess you finally admit that memory width provides not much significant speed advantages

The real benefit of this in my opinion is that manufacturers have 0 excuse now to solder LPDDR

Also, I took a look at current prices of CAMM2 modules

Crucial 32GB LPCAMM2 LPDDR5x-7500 for $175, $157.5 after 10% off coupon
Crucial 64GB LPCAMM2 LPDDR5x-7500 for $330, $297 after 10% off coupon

And this is directly from Crucial which is MSRP, generally when it hits stores it will be cheaper

Quote from: A on May 08, 2024, 20:46:32I guess you finally admit that memory width provides not much significant speed advantages

What kind of wild nonsense is this? Of course it does, otherwise the servers would not have installed HBM3e with a 1024-bit bus, soldering it as close as possible to the SoC to reduce latency.

These new modules do not improve frequencies and delays in any way, but they cost 2 times more (minimum). And yet they are rare, which forces you to take only the expensive ones.

Everything else is marketing nonsense for ordinary idiots.

If you want to increase the memory frequency, solder it right next to the SoC on a motherboard of at least 32GB.

Do you want a fast igpu in a mainstream laptop/PC? - solder 16-32GB HBM3+ with a 512/1024 bit bus directly into the SoC chiplet, getting real 250GB/s-1TB/s, and not the shameful current 50-90.

When there is mass production and competition from dozens of suppliers, prices quickly fall significantly. When there are several, a stable oligopoly develops, which can only be violated by force by anti-monopoly authorities if their officials are not corrupt and not idiots who were accepted, contrary to the laws of meritocracy, as is now customary in the US.

I am a giant fascist who hates good things

Another idiot masquerading as me, but all these idiots are betrayed by only one thing - they are not able to write even 2-3 paragraphs of logically coherent and intelligent text, because they are endlessly stupid compared to me. =)

Quote from: NikoB on May 08, 2024, 16:53:112. These modules DO NOT provide any real significant speed advantages.

They should bring an advantage when using integrated graphics, but I'm not sure how much. I would like to see comparisons of devices with exact same specs, but one with SODIMM-5600, SODIMM-6400, LPCAMM2-7500....

(that comparison is something I would like to see from laptop reviewers...)

(but whatever, I don't really care about now, as it will probably take some years to become more widespread...)

Quote from: Hotz on May 09, 2024, 20:51:56(but whatever, I don't really care about now, as it will probably take some years to become more widespread...)

In fact, every sane consumer should be concerned about this, because... a change in the standard always leads to a deliberate reduction in the range of modules that can be inserted into new models, competition and an increase in prices due to this. This is how it was originally intended.

The main thing that these modules should have provided right away (I emphasize again IMMEDIATELY) is at least a 256-bit bus and a minimum frequency of 7500. There is nothing of this. This is 100% a scam of suckers for money and nothing more.

And yes, I have read the product specifications page for all Intel processors. Yes, it does say "Max # of Memory Channels: 2", but what reason do I have to pay any attention to that. This new memory standard should force the processors to simply accept more memory channels or else it is simply a scam, a ruse to rip more money out of clueless consumers. As the informed buyer I am, I will never buy anything other than SODIMM memory even if it means I have to only buy old hardware going forward.

They should just put LPDDR5X on a SODIMM, why can't they do that? That baffles me the most, it's as if they think we are stupid. The chips are available, and all I hear is bogus explanations such as "signal integrity" and "the chips operate at low voltage so noise can interfere as the traces get longer" and then show me datasheets demonstrating this. Like, do they think I'm stupid?

Quote from: NikoB on May 08, 2024, 16:53:112. These modules DO NOT provide any real significant speed advantages.

I'm quoting you again here, as you are right. Someone posted PCMark10 test results in a forum, which show LPCAMM2-7500 having 7% improvement compared to a 5600-SODIMM in video and 3D usage (the things which need it most porbably). Ugh... what a joke.

Quote from: NikoB on May 10, 2024, 00:43:17This new memory standard should force the processors to simply accept more memory channels or else it is simply a scam, a ruse to rip more money out of clueless consumers. As the informed buyer I am, I will never buy anything other than SODIMM memory even if it means I have to only buy old hardware going forward.

They should just put LPDDR5X on a SODIMM, why can't they do that? That baffles me the most, it's as if they think we are stupid.

That's it probably. It would require more channel, but they won't give us that. What a joke. And yes, they could put LPDDR5x on SODIMM as well.

It is technically impossible to put LP memory on a SO-DIMM.

Lol, almost any comment section under almost any article, and there is Niko producing astonishing amounts of innacurate spam full of fantasies (and conspiracy theories) just to get angry about literally anything.

So, aside from this from Benjamin...

Quote from: Benjamin Herzig on May 10, 2024, 11:07:07It is technically impossible to put LP memory on a SO-DIMM.

...I will just add this very informative article (and video) from iFixit: LPCAMM2 Is Finally Here, and It's a Big Deal (https://www.ifixit.com/News/95078/lpcamm2-memory-is-finally-here)

Quote from: Hotz on May 10, 2024, 09:48:35That's it probably. It would require more channel, but they won't give us that. What a joke. And yes, they could put LPDDR5x on SODIMM as well.

The second trash comment from the troll at 0:43 is not from me. Don't you feel the difference between my comments and the comments of idiots masquerading as me? It is sad...

I have nothing more to add to what I have already written - these modules are an oligopolistic conspiracy to eliminate competition in the market and raise prices significantly. That's what happened. These modules do not provide the consumer with any technical advantages that justify this price level, but he will be forced to buy only such models, because There will soon simply be no other options left due to oligopolistic collusion. It is necessary to distinguish between real technical progress and greedy hidden collusions of producers. I don't see any real progress or step forward in them. If there was a doubling of the bandwidth, that is, at least a 256-bit memory controller even in the lowest-level module, that would be progress. But this is not the case.

Maybe if I scream oligopoly over and over again people will believe me (even though I had absolutely no problems about SO-DIMM being the standard for removable laptop memory before this and that was proposed by JEDEC as well).

Another moron masquerading as me at 20:32. And as always, he's so stupid that he can't even write an adequate paragraph of text.

How come there are so many people pretending to be me? Don't they know that my stupidity is unique and that nobody has an obsession with writing paragraphs of useless fluff and drivel like me?

Well this is just first iteration,problems has to be solved ,the LPDDR6 Memory is expected to offer 17.066 MT/s LPDDR5X= 8.533 MT/s so from a workstation class i wouldn't expect/ buy anything less just to justify the purchase.

Quote from: Alletheya on May 15, 2024, 19:27:04Well this is just first iteration,problems has to be solved ,the LPDDR6 Memory is expected to offer 17.066 MT/s LPDDR5X= 8.533 MT/s so from a workstation class i wouldn't expect/ buy anything less just to justify the purchase.

This is unrealistic given the fact that AMD had problems even with the 7500 memory soldered, at least.

The only thing that matters is increasing the bus bit rate; increasing frequencies is a losing strategy in advance.

For several years now, x86 should have a 256-bit bus in mass models and a 512-bit bus in top gaming and professional series.

And all igpus must come with at least 4GB soldered directly into the SoC gddr6x or hbm3+ chiplet. That's when igpu will really compete with dgpu.

What is the problem and prostration on the part of consumers? They see retail prices going from store to store at wild values (reaching hundreds of dollars for the same model, and the "discounts" are simply ridiculous and shameful, when they suddenly drop even $800-1000 - this is normal, so raise the MSRP to at the beginning for a product that is clearly not worth the money? Is this generally immoral against the backdrop of subsequent huge "discounts" in price), despite "market competition", Especially in third world countries, where competition is even worse even in retail, although This exists in Western countries, but buyers do not see the real cost of the laptop from the manufacturer. Seeing such wild variations in retail prices, a fairly experienced consumer rightly thinks that with such surges in retail prices, he has the right to expect additional improvements to be included in the cost of a product for which he is paying a clearly inflated price. Only by seeing the real cost of a laptop can a buyer correctly assess the manufacturer's ability to integrate more advanced technologies into the product.

Of course, "discounts" clearly indicate that there is no demand, but some still buy at an inflated price. Even though they have the money for it, they clearly do not get what they could at the start of sales if the price were more fair in relation to the quality and capabilities of the product, if the cost of the product was always proportional to the retail price.

When, conditionally, one buyer takes a model for $5,000 at the start of sales, and then with "discounts" another for $3,000, doesn't the first buyer have the right to expect proportionately greater quality and capabilities of the same product than the second buyer already for "3,000" ? And aren't such "discounts" essentially economic fraud? It turns out that the first buyer simply wasted $2,000 without receiving anything in return. There needs to be some balance and decency in this regard from the manufacturers. But of course the real world is somewhat different. Nothing but the thirst for profit and the search for "fools" with money.

I am for the retail price to be limited in the markup above the cost according to a certain formula. This also eliminates unnecessary speculators from the chain and does not allow manufacturers to engage in outright fraud when some buyers cannot correctly assess the real price of a product.

Well, at least look at what pharmaceutical companies are doing in the USA. If this is a "free market", why is even the Biden administration screaming at the top of its lungs that prices are obviously fraudulently inflated?

Or even take grocery chains - one day a buyer comes and buys the same products with a 1.5-2 times markup, and the next day, he or another suddenly sees the price 1.5-2 times lower.
Some kind of unfair lottery that allows retailers to maintain a certain "average price". But how does a person feel who bought products yesterday that were 1.5-2 times more expensive and seeing the same thing tomorrow is just as cheaper, but when he no longer needs them and he simply didn't get to the day (or "discount" hours).

This is an extremely common practice in fact and it is also pure price fraud. But already from the retail side.

Why this is 100% scam - a simple example. The buyer takes sausages as an example. At a price of approximately $15 per package of a certain weight. He sees that the price of these sausages is higher than the price of sausages of other varieties and manufacturers, for example, 1.5 times. If this is a random buyer and/or he is not aware of the vile scheme with temporary "discounts" in retail chains, he has the right to decide that these sausages are clearly of higher quality in composition than other varieties and other manufacturers with the same weight (they are now practically deceiving with labels everything and besides, rarely does anyone read the full composition in small print). What will be the shock of this buyer if the next day he discovers that their price has dropped by 2 times and has become cheaper than other sausages? And then on the third day (relatively speaking) the price increased again to the previous level at which he bought? Does he have the right to decide that the retail chain is engaged in outright price fraud by not allowing him to choose a quality product on an arbitrary day of purchase, based on a fair proportional difference in price in relation to the quality of the product? Obviously this is so from the point of view of rational and common sense thinking and simply basic business decency and requires the intervention of consumer supervisory authorities.

I'm just trying to explain that the expected improvements in the characteristics and quality of the product at an increased price visible to the buyer at the moment may turn out to be a pure bluff. And if in a laptop, for example, this does not have a significant impact on health, then in the case of products these are clearly criminal schemes.

And all this was written to try to explain that the wild differences in retail prices for the same laptop model more than cover the addition of a gddr6x or hbm3 controller directly into the SoC and the desoldering of fast 4GB (at least) memory for the igpu. And about the same with other technologies. For example, AMD or Intel include TB4/USB40 controllers, HDMI2.1 (FRL6) ports, RJ45 2.5Gbps+ port, etc. in the SoC, but the laptop manufacturer does not display them on the case. Despite the fact that the difference in retail prices between different stores completely covers the costs in the cost price.

What am I getting at? It is necessary to legally require the publication of the cost of production of all goods and punish for fraud (audit of the declared cost is a separate interesting topic). Then the consumer himself will assess how adequate the speculative component is in retail or when setting the starting price for a product by the manufacturer.

This is the second way to limit speculative markups - through clear knowledge by the buyer what the real cost of the product is. Then there is no need to limit the speculative markup by law according to a certain formula - buyers will quickly understand where the product clearly has an unfair price.

We are trying to find a way out of the impasse of current "capitalism" and make goods of better quality, more technologically advanced and cheaper, right? Competition with the rise of huge multinational corporations and oligopolistic conspiracies (which have been proven in court many times, but with no real fair punishment and in time for buyers - but it took years) simply does not work.

This means we need a different pricing model through regulation of this by society in relation to business. And this is precisely what will allow us to instantly remove from the market companies that use fraudulent methods to inflate prices, in defiance of decent producers, and then, having made money on poorly versed buyers, begin to engage in corruption of legislators, etc. in a vicious circle, increasing the gap from decent manufacturers and ultimately leading them to bankruptcy, despite a high-quality product and a reasonable markup. Or you have to act like everyone else. But this is a bad, degrading trend from the point of view of society. Well, for example, now they write about something similar with American medicine in general and the medical staff. This is a greedy, immoral business, not an attempt to cure as many people as possible as quickly, efficiently and cost-effectively as possible. And the fact that the majority of doctors in the United States have left private practice is also indicative of the latest statistics - it is easier to work for insurance conglomerates, but not for the patient.

Well, this is the type of initiative that eliminates from the chains unnecessary parasites-speculators and opportunities for fraud with inflated prices by various methods:
variety.com/2024/music/news/ticket-act-passes-house-of-representatives-1236005647/

Isn't this a departure from the principles of capitalism? But this benefits the entire society, right?

Quote from: NikoB on May 10, 2024, 13:55:46

Quote from: Hotz on May 10, 2024, 09:48:35That's it probably. It would require more channel, but they won't give us that. What a joke. And yes, they could put LPDDR5x on SODIMM as well.

The second trash comment from the troll at 0:43 is not from me. Don't you feel the difference between my comments and the comments of idiots masquerading as me? It is sad...

I have nothing more to add to what I have already written - these modules are an oligopolistic conspiracy to eliminate competition in the market and raise prices significantly. That's what happened. These modules do not provide the consumer with any technical advantages that justify this price level, but he will be forced to buy only such models, because There will soon simply be no other options left due to oligopolistic collusion. It is necessary to distinguish between real technical progress and greedy hidden collusions of producers. I don't see any real progress or step forward in them. If there was a doubling of the bandwidth, that is, at least a 256-bit memory controller even in the lowest-level module, that would be progress. But this is not the case.

They are not some oligarch conspiracy to bring up prices, CAMM was originally made by Dell who has nothing to do with the memory business. It was given to JEDEC as a standard and adopted into CAMM2

CAMM2 allows 128bit per module compared to 64bit for SODIMM

And again, the biggest advantage of this is that there is no excuse to solder ram anymore. What do you think carries a bigger premium, buying soldered ram from the manufacturer, or purchasing ram from a 3rd party retailer and reusing ram from older computers? I'll take the CAMM2 over soldered any day

Prices will drop by the day as more start producing them in bulk

PS your BOM idea will never work, because companies will just make shell companies that bill them higher manufacturing costs and it will show up as "services" in their income statement.

Quote from: A on May 17, 2024, 08:11:22CAMM2 allows 128bit per module compared to 64bit for SODIMM

I don't see 256 bit CAMM2s on the market doubling the bandwidth. The first memory controller with a 256-bit bus (to the shame of Intel) will again be available from AMD in Zen5 and only in the top segment, but it should have been in mass-market PCs/laptops a long time ago.

And what is typical, as I predicted, only the top version has support for 8K monitors with UHBR20. =)

It's a shame for the entire x86 industry - it is at least 5-6 years behind user requirements.

Prices are now 2-2.5 times higher - this is 100% not a good deal unless the buyer is an idiot. With a 128-bit bus.

Quote from: A on May 17, 2024, 08:11:22your BOM idea will never work, because companies will just make shell companies that bill them higher manufacturing costs and it will show up as "services" in their income statement.

It will work, but the scammers must be destroyed. Why does civilization need unscrupulous business? They need to be eliminated at the beginning.