Tuesday, April 27th 2021

ATP Launches PCIe Gen3 x4 NVMe SSDs in M.2 Type 1620 HSBGA Package

ATP Electronics, the global leader in specialized storage and memory solutions, has announced the launch of its tiniest NVMe flash storage offering: the N700 Series PCIe Gen3 x4 NVMe solid state drives (SSDs, which are available as M.2 Type 1620 heatsink ball-grid array (HSBGA) package. Complying with M.2 specifications, the M.2 Type 1620 HSBGA measures just 16 (L) x 20 (W) x 1.6 (H) mm, supporting high-speed PCIe 3.0 interface x4 lanes and NVMe protocol to deliver up to 32 Gb/s bandwidth at 8 Gb/s per lane. The soldered-down design makes them vibration-proof, while the 291-ball packaging takes up minimal space within tightly confined systems.

For customers who prefer a removable and field-replaceable design, ATP can accommodate the HSBGA onto an M.2 2230 module with the same firmware and NAND configuration. Both variants are suitable for thin and light systems in embedded, industrial and mobile applications. N700 Series SSDs are built with 3D triple-level cell (TLC) configured as pseudo single-level cell (pSLC) NAND flash. By storing only one bit per cell, they increase the reliability and endurance of the NAND flash memory, while benefiting from the lower cost compared with native SLC, due to the higher cell density.
These SSDs are outfitted with a heatsink on top, which complements system airflow to enhance heat dissipation and keep the BGA SSD cool while offering 2X-3X better sustainable performance.

The optimized power consumption of just 5 mW during Power State 4 (Sleep Mode) translates to big power savings.

BGA SSD technology allows the NAND flash and controller to be integrated into one package that is lightweight yet offers powerhouse performance and ample capacities. N700 Series SSDs come in 40/80/160 GB and are packed with advanced features to meet the ultra-portability and reliability requirements of ultra-compact Internet of Things (IoT) devices and embedded systems. They provide high-speed reliable storage in harsh environments such as in transportation, aerospace, smart factories, mining operations, steel fabrication and more.

N700 Series SSDs are available in both C-Temp (0°C to 70°C) and I-Temp (-40°C to 85°C) ratings. They all undergo rigorous tests, such as Rapid Diagnostic Testing (RDT), to screen out weak blocks and ensure reliability and robustness even in extreme operating conditions.

N700 Series Key Features
  • pSLC Mode. Configured to store only one bit per cell to increase endurance and reliability, offering 2X-3X sustainable performance.
  • Stable Performance. The ATP Optimized Thermal Throttling firmware (FW) will maintain the "Steady State" condition to avoid huge performance drops that will adversely impact the system, thus optimizing best performance for application requests and enhancing overall sustained performance.
  • Optimized Power Consumption. Consuming low power at only 5 mW under Power State 4 (Sleep Mode), the ATP NVMe HSBGA SSDs deliver huge power savings.
  • DRAM-Less Configuration. Host Memory Buffer (HMB) support helps these DRAM-less SSDs to improve performance by obtaining DRAM resources as cache, thus overcoming the limited memory capacity within the storage and optimizing I/O performance without requiring the SSD to bring up its own DRAM
  • Vibration-Resistant Storage. ATP N700 Series SSDs in M.2 Type 1620 are soldered down, making them vibration-resistant and able to withstand rigorous shaking.
  • Better Thermal Dissipation. The heatsink effectively transfers heat to cool the device and keep the performance at optimal levels.
  • Optional Security Features
    • HW Write Protect
    • HW Quick Erase
    • HW Secure Erase
    • SW Data Sanitization (AFSSI-5020)
    • AES-256 Encryption
    • TCG Opal 2.0
For more information, visit this page.
Add your own comment

12 Comments on ATP Launches PCIe Gen3 x4 NVMe SSDs in M.2 Type 1620 HSBGA Package

#1
thesmokingman
That's pretty cool. I could see these takeover nuc type boxes.
Posted on Reply
#2
Valantar
Wait, m.2? I though m.2 was an AIC form factor. This is a BGA chip. Did I miss something?
Posted on Reply
#3
TheLostSwede
Valantar
Wait, m.2? I though m.2 was an AIC form factor. This is a BGA chip. Did I miss something?
There's a BGA form factor as well, several companies make M.2 1216 soldered down WiFi modules as an example.

Toshiba launched a similar SSD back in 2017.
www.businesswire.com/news/home/20170802006627/en/Toshiba-Memory-Corporation-Unveils-Single-Package-NVMeTM-Client-SSD-Utilizing-64-Layer-3D-Flash-Memory
And Samsung in 2016.
www.anandtech.com/show/10166/samsung-demos-its-first-bga-ssd

Posted on Reply
#4
Valantar
TheLostSwede
There's a BGA form factor as well, several companies make M.2 1216 soldered down WiFi modules as an example.

Toshiba launched a similar SSD back in 2017.
www.businesswire.com/news/home/20170802006627/en/Toshiba-Memory-Corporation-Unveils-Single-Package-NVMeTM-Client-SSD-Utilizing-64-Layer-3D-Flash-Memory
And Samsung in 2016.
www.anandtech.com/show/10166/samsung-demos-its-first-bga-ssd


I've seen those BGA SSDs before, but I don't think I've noticed them being called m.2 before. The AnandTech article you linked even explicitly differentiates BGA SSDs from m.2 SSDs, though I do see the Toshiba press release calls it m.2 1620 BGA (even if they manage to mistype it as M.2). Still, it makes no sense whatsoever to me. It used to be confusing enough that m.2 could mean both SATA and NVMe SSDs, and that not all drives worked in all slots. Now we for some reason also need to call soldered BGA SSDs m.2? Why? What does anyone stand to gain from that? Why not give it a separate name? Are all standards bodies these days run solely by marketing executives terrified of not being able to check the latest buzzword off on their spec sheets? I mean, if "m.2 SSD" can mean NVMe or SATA and soldered or slot-mounted, the term is effectively meaningless.
Posted on Reply
#5
TheLostSwede
What is gained from any standard? Yes, it could've been called something else, but take a notebook maker as an example, now they know that they can buy WiFi modules or "compact" SSDs from a range of sources and they know they'll all be compatible. In the past, there were a bunch of different non standardised solution that required multiple solder masks on the PCB if you wanted to be able to swap between them. As such, this is a pretty sensible thing.
Posted on Reply
#6
Valantar
TheLostSwede
What is gained from any standard? Yes, it could've been called something else, but take a notebook maker as an example, now they know that they can buy WiFi modules or "compact" SSDs from a range of sources and they know they'll all be compatible. In the past, there were a bunch of different non standardised solution that required multiple solder masks on the PCB if you wanted to be able to swap between them. As such, this is a pretty sensible thing.
What is gained from a standard is ... standardization. I.e. things being similar or the same within the bounds of the standard. The broader those bounds, the less useful the standard. In this case, m.2 now encompasses such a broad range of incompatible form factors that it renders itself meaningless. "Oh, my laptop has an m.2 SSD, guess I can upgrade it!" "Nope, sorry, it's a soldered m.2 SSD. No can do."

Your example doesn't really apply: the people choosing parts for a notebook are engineers. They really should know the relevant standards. It really shouldn't be a problem for them if soldered parts had a different name than slot-mounted ones. Quite the opposite, I'd say, it would make their jobs easier. The different mounting requires different trace layouts and different engineering solutions anyhow, so why use the same name? What you're asking for is a standard, which is quite different from it needing to be the same standard as something else after all.

Fixing this would be as simple as giving the different form factors different names. m.3 was already tried by Samsung (kinda shadily through adopting a name that similar to m.2 without consulting with the relevant standards body, but still), and there are a bunch of others as well. The point of the name is to signal similarity and interoperability, no? Why not call the soldered ones s.2? Or b.2, for BGA? I mean, the only limit is the imagination. Instead, they've chosen to dilute the standard through making it encompass a broad range of incompatible devices. Which really ruins the point of having a standard in the first place.
Posted on Reply
#7
TheLostSwede
Of course it applies. Have you even looked at how many different types of non standardised WiFi modules there are on the market? You can't even use what appears to be the kind of module from another company, as they did a different pin-out, for no good reason.
So having a strict standard for these things, beyond just size, is key to being able to offer different SKUs of a product, or to being able to order an alternative solution when your main provider lets you down.

One of the reasons everything is m.2 is because it's all PCIe plus some other interfaces specified by the m.2 standard, not just PCIe.

I did mention that they could've called it something else, but again, this is not a standard made for consumers as you seem to think, but rather for companies that build systems and devices.
Maybe you should join Intel or one of the other big companies that develop and name these standards? That would make the world a better place for everyone else, as you'd be able to influence the naming of so many things.
Posted on Reply
#8
Valantar
TheLostSwede
Of course it applies. Have you even looked at how many different types of non standardised WiFi modules there are on the market? You can't even use what appears to be the kind of module from another company, as they did a different pin-out, for no good reason.
So having a strict standard for these things, beyond just size, is key to being able to offer different SKUs of a product, or to being able to order an alternative solution when your main provider lets you down.

One of the reasons everything is m.2 is because it's all PCIe plus some other interfaces specified by the m.2 standard, not just PCIe.

I did mention that they could've called it something else, but again, this is not a standard made for consumers as you seem to think, but rather for companies that build systems and devices.
Maybe you should join Intel or one of the other big companies that develop and name these standards? That would make the world a better place for everyone else, as you'd be able to influence the naming of so many things.
So we're not allowed to complain when standards bodies make stupid decisions unless we're willing to devote our lives to changing it? Yeah, sorry, I'm not following you into that cesspit of bad-faith arguments.

Regardless of whether it's a consumer or component manufacturer oriented standard, regardless of wether it's PCIe alone or PCIe+other options (miniPCIe also had provisions for USB and SATA, so I don't quite see the issue there), separate names for separate and fundamentally incompatible classes of devices makes sense. Plain and simple. You're drawing up a straw man here presenting it as if the alternative is no standard at all, rather than properly named standards. I have never at any point argued for the former, and presenting that as the main option beyond "let's call everything m.2" is, yet again, making a bad-faith argument. Please stop. It doesn't matter whether that was the situation previously - that is in no way an argument against the value of criticizing an overbroad standard. Yes, an overbroad standard is better than no standard at all, but several more specific standards would again be better than an overbroad one. I mean, in this case the issue would be solved by a simple naming structure change, as the more specific standards already exist! They've just chosen a weird common naming scheme that supposedly denominates a set of AIC form factors (with potentially different signalling standards), yet at the same also includes fundamentally different form factors. A line ought to have been drawn somewhere here, as there are too many layers of incompatibilities involved.
Posted on Reply
#9
TheLostSwede
Eh? It was a suggestion on how you could improve the situation. Of course you're allowed to complain about it, but tell me, what will it change?

Mini PCIe never had solder down modules or BGA options though, did it?

I never said that, please read my first reply again, where almost the first thing I said was that they could've called it something else.

My personal opinion is that a standard with a bad name is still better than no standard at all.
Look at TransFlash, which is still widely used by various unlicensed Chinese companies in lieu of microSD which was the name the standard got in the end.
Even m.2 was originally known as Next Generation Form Factor (NGFF), but I guess that wouldn't have made sense for solder down things either.
Posted on Reply
#10
Valantar
TheLostSwede
My personal opinion is that a standard with a bad name is still better than no standard at all.
That's the issue here. You're presenting this as a black-and-white choice between two options (which it should be plenty obvious isn't the case - there can be however many standards one wants, after all), and through presenting that as a counterargument to what I was saying, you're strongly implying that I'm arguing that no standard at all is better than a bad standard. Which it should be plenty clear that I never did.

As for making
TheLostSwede
a suggestion on how you could improve the situation
that's another way of sayinng you're individualizing responsibility for a systemic problem while minimizing the value of raising a complaint outside of the system. Intentionally or not, that's a bad-faith arguing tactic that does nothing beyond trying to make your opponent look either irrational or as not trying hard enough. Effectively saying "if you care so much, why don't you..." is not a rational, on-topic argument in any discussion. The same really goes for asking "what will change" - I mean, it's pretty obvious: likely nothing. But potentially, a discussion is started, and it might be brought to the attention of someone involved. It's not likely, but that's not an argument against bringing up problems. You, on the other hand, are using bad-faith arguing tactics to shut down even the potential of debate, which ... why? What's the point? What's the potential harm in discussing this?
Posted on Reply
#11
TheLostSwede
Dude, you seriously woke up on the wrong side of the bed, as you're reading so much into my comments that I didn't say, but whatever.
Posted on Reply
#12
Valantar
TheLostSwede
Dude, you seriously woke up on the wrong side of the bed, as you're reading so much into my comments that I didn't say, but whatever.
Or maybe you should choose your rhetorical devices with a tad more care? You're typically one of the more sensible people around these forums in my experience, so I was rather baffled by your tone and approach here. The sarcasm and dismissiveness was ... just weird.
Posted on Reply
Add your own comment