Dell using Proprietary RAM Modules

[Kisai]

https://www.pcworld.com/article/693366/dell-defends-its-controversial-new-laptop-memory.html

Quote

In an interview with PCWorld, however, both the person who designed and patented the CAMM standard, as well as the product manager of the first Dell Precision laptop to feature it, assured us the intent of the new memory module standard is to head-off looming bandwidth ceilings in the current SO-DIMM designs. Dell’s CAMM, in fact, could increase performance, improve reliability, aid user upgrades, and eventually lower costs too, they said.

 

Also

Quote

In fact, Dell points out, it’s not even “proprietary” on its own laptops. The first Precision workstations that come with CAMM will also eventually be offered with conventional SO-DIMMs using an interposer. Mano Gialusis, product manager for Precision workstations, said the interposer option goes into the same CAMM mount, too.

Yeah sure Dell.

 

I guess just wait and see what happens, my guess is that CAMM won't be adopted by other vendors and they will instead all make their own proprietary module boards the same way.

[hishnash]

14 minutes ago, Kisai said:

I guess just wait and see what happens, my guess is that CAMM won't be adopted by other vendors and they will instead all make their own proprietary module boards the same way.

Many vendors for laptops will go with soldered solutions as even this CAMM solution cant support LPDDR5 or LPDDR5x and if your building a laptop that has an APU your going to what that power savings for the bandwidth the only laptops that fall into the CAMM are the large workstation laptops were users do not expect battery life. Power savings of LPDDR5 are even higher the higher your capacity. The reason vendors have been soldering chips in the last fews years as been LPDDR4, with intel e-Cores and more ARM machines on the market (and even some AMD cpus) do you really want your memory to draw more power than 2 cpu cores on full burn? 

 

The reason dell made CAMM and did not solder is the cpus they are using (HX) do not support LPDDR5. 

[leadeater]

2 hours ago, hishnash said:

Many vendors for laptops will go with soldered solutions as even this CAMM solution cant support LPDDR5 or LPDDR5x

There is legit no reason why this couldn't use LPDDR5 nor LPDDR5x for that matter either. The reason they are doing this is so they can offer ECC.

 

The current/future Dell laptop lineup that uses LPDDR5 supports maximum officially of 32GB non-ECC. Current/future Dell laptops that will use CAMM support 64GB non-ECC and ECC.

 

2 hours ago, hishnash said:

Many vendors for laptops will go with soldered solutions as even this CAMM solution cant support LPDDR5 or LPDDR5x

HX supports both, the IMC doesn't change just from chip binning and microcode. The above is why they are doing CAMM.

[darknessblade]

More info regarding the DELL CAMM module

 

Dell defending their E-waste generator

https://www.pcworld.com/article/693366/dell-defends-its-controversial-new-laptop-memory.html

 

A review of the Laptop featuring the CAMM module

https://wccftech.com/dell-precision-7770-7670-workstation-laptops-official-intel-alder-lake-hx-16-core-cpus-camm-128-gb-ddr5-memory/

 

Dells supposed fix: AKA a sodimm adapter

 

My thought

 

If they created a adapter, why did they not add Sodim adapters to the Dell laptops in the first place?

 

Are these adapters also FREE for anyone wanting to use sodimm instead, or do you have to pay a DELL premium price for it.

[GhostRoadieBL]

So I asked around the office, not a huge sample size but a mid/low double digit number of dedicated laptop workstation users.

Not a single one cares at all about the thickness of the laptop. About a third run inch thick "gaming" notebooks, some of us run thinner workstations and full towers/rack based modules and a few run tough-tablets for field work and workstation laptops for crunching data.

Between performance, reliability, repairability, port selection, noise and size noone cares about size or lightness. It's nice to have but when you are hauling around a 5+lbs laptop in a backpack and it's 1.5inch or 5/8inch thick it doesn't make any difference to the user.

 

Most actually want thicker laptops so they fit better in your hand (mostly the tough tablet users)

 

Dell needs to start listening to professionals more when it comes to product designs, just because it's pretty doesn't mean it works.

[hishnash]

2 hours ago, James Evens said:

Open standard protected by patents?

 

most if not all industry standards are protected by patents. Typically when a standard is made this puts a price cap on the license fee and requires the vendor to license it. You will be hard pressed to find a standard that does not have a ton of patents around it.

[hishnash]

2 hours ago, James Evens said:

That's not what I call open standard. No need to push it into the public domain but at least make it free to use.

General paying for standards is a very big problem (e.g. ISO standards). This is even a bigger issue then scientific paper being locked behind paywalls.

I would not be surprised if the current SODIM standard has patents attached to it that you need to pay for.  And there is no way your making a DDR5 controller without paying at leat 5 to 100 people for related patents to ensure your not being sued, or you have a cross licensing agreement wit them so they can use your patents and you use theres to save you both from the nightmare of the legal hell hole of patent law... every cpu out there will have 1000s of bits that other people have patents that could be interpreted as bing breached so its a complete f-up world of hell if your in the hardware industry. 

[hishnash]

9 minutes ago, James Evens said:

Not only hardware industry. 

Patents are now at the point where they do more harm then good.

absolutely!  

[leadeater]

7 hours ago, James Evens said:

Not only hardware industry. 

Patents are now at the point where they do more harm then good.

I'm afraid I own the patent on that opinion, pay up.

 

7 hours ago, hishnash said:

I would not be surprised if the current SODIM standard has patents attached to it that you need to pay for.

There's a patent of the physical SO-DIMM connector/slot used on most laptops

https://patents.google.com/patent/US20150171535

 

[Kisai]

9 hours ago, James Evens said:

Not only hardware industry. 

Patents are now at the point where they do more harm then good.

Not true, just the "driving innovation" part tends to be "looking for ways to subvert the patent" rather than improve it. Especially in software.

 

Realistically, software, algorithms, and api's should not be protected by patents or copyrights. Only the license (which patents and copyrights shouldn't apply to) is enforceable, and only on the software-as-a-whole. eg Nobody sells Windows 3.1 or DOS 6.22, I've purchased this software at some point in the past, but finding the serial numbers or original diskettes is nothing short of impossible. I should be permitted to use backup copies (regardless of how I acquired it,) or re-implement the DOS and Windows operating systems without being challenged by Microsoft, because Microsoft does not actively sell licences for it, and does not support it.

 

As soon as a software or hardware vendor "no longer supports" something, it should be treated the same as public domain/CC0, in that while you can not strip the branding off it and sell it as your own, or sell it alone to others, if something you wrote relies on DOS or Windows 3.1, you should be able to package DOS and Windows 3.1 to run your software as a virtual hard drive to run on a virtual machine.

 

The reason copyrights really should not apply to software is because software today, and really since the 90's, has been a series of libraries of various licenses, much in the same way music today is created from a series of sound and instrument libraries. Music has a special carve out which requires two licenses (one for the musical "program", and one for the lyrics) to be clear to publish. So should software. No software package out there is competent enough to follow all the software library EULA's that they incorporate, and frequently you can find the licenses for zlib, libpng and other common file formats in both open and closed software buried under some license menu.

 

Hence patents and copyrights should not apply to software, at all. Copyright should apply to only the exact published version, and if the copyright holder no longer sells and/or supports it, it should be PD/CC0. Look at all the freemium P2W stuff that shuts down after 6 months after raking in a few million dollars. Compare that to magazines that at least the physical magazine can still be found in a library in a few months, but the actual newsstand copies are gone after a month.

 

Patents for hardware on the other hand, tend to always be about undermining an already established product. Sure I'm sure the SODIMM slot for DDR5 is different from DDR4 and DDR3 because of the necessity of making it idiot-proof, but these are not "new" inventions, they are the same SODIMM slot you've seen since 1998

https://allpinouts.org/pinouts/connectors/memory/so-dimm-dram-144-pin/

 

DDR2 https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/modules/sodimm/htf8c32_64_128x64hd.pdf

DDR3 https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/modules/sodimm/jtf16c256_512_1gx64hz.pdf

DDR4 https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/modules/sodimm/ddr4/atf8c512x64hz.pdf

DDR5 https://media-www.micron.com/-/media/client/global/documents/products/data-sheet/modules/sodimm/ddr5/ddr5_sodimm_core.pdf

 

The number of pins and position of the cutout changes, but otherwise the mechanical aspects have not changed. Voltage changes are relative to the chips. There is even less difference between DDR2 and DDR3 shapes than there are DDR and DDR2, and DDR3 and DDR4. 

 

Other than the key pin, these boards are all the same size within 0.3mm, with those holes being the same and the side notches being in the same place and size. For the most part these are all mechanical tolerances.

 

So of course there will be a patent on the same boards and slots where they moved the key pin, but otherwise identical. Do these really deserve a "new" patent when just the number of pins change? No, I don't think so, but business processes dictate you patent everything, no matter how stupid and obvious it is.

 

 

 

[IPD]

On 4/27/2022 at 2:24 PM, GhostRoadieBL said:

So I asked around the office, not a huge sample size but a mid/low double digit number of dedicated laptop workstation users.

Not a single one cares at all about the thickness of the laptop. About a third run inch thick "gaming" notebooks, some of us run thinner workstations and full towers/rack based modules and a few run tough-tablets for field work and workstation laptops for crunching data.

Between performance, reliability, repairability, port selection, noise and size noone cares about size or lightness. It's nice to have but when you are hauling around a 5+lbs laptop in a backpack and it's 1.5inch or 5/8inch thick it doesn't make any difference to the user.

 

Most actually want thicker laptops so they fit better in your hand (mostly the tough tablet users)

 

Dell needs to start listening to professionals more when it comes to product designs, just because it's pretty doesn't mean it works.

Been saying that for years.  I want a thicker gaming laptop that won't overheat and doesn't sound like an XF-84 Thunderscreech.

[TetraSky]

In a laptop? ... At that point just solder it on the mainboard like Apple does if you're going to make proprietary junk.