Comments Locked

42 Comments

Back to Article

  • milkywayer - Tuesday, December 31, 2019 - link

    Keep it up Anand tech!

    One request. Add we enter the 3rd decade, any chance of adding the Edit comment and upvote features? :p I am a web dev so lemme know if you want to implemented for free! A gift for AT.
  • shabby - Tuesday, December 31, 2019 - link

    And downvotes to hide trolls while you're at it 👍
  • wenart - Tuesday, December 31, 2019 - link

    Like!
  • The True Morbus - Tuesday, December 31, 2019 - link

    Not the 3rd decade yet, that's in a year's time.
  • ChitoManure - Wednesday, January 1, 2020 - link

    2000-2009 =first decade
    2019-2019= 2nd decade
    2020-2029= 3rd decade
  • sheepdestroyer - Wednesday, January 1, 2020 - link

    You missed the point and are off by one: the very first decade did not start on year 0 as there is no such thing, but on year 1. So Morbus is right saying that the third decade of this millenary only starts in one more year.
  • sheepdestroyer - Wednesday, January 1, 2020 - link

    Ofc, for the same reason, the first year of the 3rd millenary is 2001.
  • zmatt - Friday, January 3, 2020 - link

    Fortunately I'm not a computer so I dont have a seizure and fall drooling on the floor when I encounter imperfect counting systems. Decades are generally agreed to start on 0 and nobody cares that there wasn't a year in between 1 BC and 1 AD.
  • ChitoManure - Sunday, January 5, 2020 - link

    My bad
  • ChitoManure - Sunday, January 5, 2020 - link

    My first year in college starts in ZERO year
  • mkozakewich - Sunday, January 5, 2020 - link

    You're both missing it! The first decade was 1997 (April, to be exact) to 2006. So we're already well into the third decade.
  • stryfe - Thursday, January 9, 2020 - link

    https://xkcd.com/2249/
  • s.yu - Wednesday, January 1, 2020 - link

    I looked it up and learned something new, Anno Domini starts at year 1, and before that, 1BC.
  • vgray35@hotmail.com - Thursday, January 2, 2020 - link

    That is only true if you use the BC/AD nomenclature. Most of us doing calculations require a zero year in calendric computations. The Astronomical Julian Period system now uses the Common Era (CE) system with a zero year and both negative and positive years, rather than the simplistic BC/AD nomenclature.The Romans and the middle ages were confused about usage of zero generally, and there is no need today to think like those of the past. So we have -1CE/ 0 CE/ 1 CE. This new system also defines BCE when it is desired to use traditional BC years. Today we have advanced by adopting zero as part of our counting system, and CE nomenclature better serves everybody, and using this new system 2000 CE is viewed as the start of the millennium. I suggest you join the rest of us in the modern world in realizing 0 CE = 0 AD and the year before is -1 CE instead of 2 BC. Time to let go of the past. When dealing with Calendrical science the lack of a zero year matters greatly.
  • vortmax2 - Thursday, January 2, 2020 - link

    We all know the real reason for the switch to CE and it has nothing to do with zeros.
  • vgray35@hotmail.com - Thursday, January 2, 2020 - link

    That reason was invented by bigots, with no understanding of the rudiments of the system. Anybody can see the efficacy of the system as regards protracted computations,except those with a fabricated agenda. Astronomers have a real reason for using it, quite apart from any imagined "real reason" for its invention. Your adage "We all know the real reason" is a misconception without foundation - regurgitating slurs originating from bigots with little understanding of the facts. The common Era remains rooted in a referential standard origin, centered around the approximate birth of Christ, if that is what you were referring to, and as I said it so remains without apology. So 2000 CE remains as the first year of the current millennium, while 2001 AD is an ancient quirk that arose from generations past that still used Roman numerals - nothing more! That simplistic view was easily fixed with the common Era system, which provided us with less complicated calendrical
    computations. Ordinary folk without need of such simpler computations, I can see might be gullible to such "We all know the real reason" rhetoric!
  • philehidiot - Friday, January 3, 2020 - link

    Who knew you could have arguments about bigoted people and whether zero exists? I have a headache.

    And for us ignorant folk... what is this "real reason"? Or is it like the name of The Scottish Play?
  • BurntMyBacon - Tuesday, January 7, 2020 - link

    @vgray35@hotmail.com: "Anybody can see the efficacy of the system as regards protracted computations,except those with a fabricated agenda."

    So should I be depressed that it took everybody 2000+ years to come up with a solution or should I assume that everybody had a fabricated agenda for 2000+ years. ;')

    I joke. I know this was largely a byproduct of the Roman number system (that lacked a zero) that was dominant at the time. Still it is disappointing that it took so long to fix.
  • DanNeely - Tuesday, December 31, 2019 - link

    At some point in the future Anandtech will be migrating from its homebrew platform to Future's shared CMS. I wouldn't expect any major functional changes to the site until then.
  • NICOXIS - Thursday, January 2, 2020 - link

    this! the commenting section is one of the most important aspects of this site, stop neglecting it!
  • chrysrobyn - Thursday, January 2, 2020 - link

    And a spam report button... the bots love to sneak into old articles
  • Drumsticks - Tuesday, December 31, 2019 - link

    Really enjoying these year in review articles. Keep it up!
  • ikjadoon - Tuesday, December 31, 2019 - link

    I'm a little confused by recent reporting on 128L-class NAND and that SK Hynix is ahead of Samsung. In August, Anandtech wrote about Samsung already *releasing* 136L NAND in 2019:

    "Later this year Samsung intends to release 512 Gb 136-layer V-NAND devices that will be used for other drives as well as eUFS storage solutions."
    https://www.anandtech.com/show/14721/samsung-unvei...

    SK Hynix, from its November article and again here, are *announcing* for retail consumers at CES 2020. It seems as if Samsung has actually released first (to ODM/OEMs first). How is NAND achievement being measured here? By the date when the product is released to paying OEM/ODM clients or it being announced to the consumers at tech conferences?

    If it's about "shipping in volume to clients", I'd say Samsung has beaten SK Hynix. But I think the wording might not be accurate here. Most plausibly, Samsung missed their 136L launch target, but that didn't get reported: maybe?
  • Billy Tallis - Tuesday, December 31, 2019 - link

    Samsung's 136 layer number is probably not counting the same thing as everyone else. They're probably including the dummy layers to inflate the count. At ISSCC in February 2019 they were presenting about NAND with "more than 110 layers". They're obviously not going to give us an exact count of active layers (which is what everyone else counts) until they're back in the lead, and we'll have to wait for TechInsights to do a teardown before we know how close Samsung got to 128L with their 6th-gen V-NAND.

    Also, "shipping in volume" can mean pretty much anything. Retail availability of a full SSD product line is less ambiguous.
  • Diogene7 - Tuesday, December 31, 2019 - link

    I am very disappointed that in 2019, there is still not much Storage Class Memory (SCM) products available on the market for consumers (except Intel Optane products).

    In my mind, it is clear that SCM would bring new benefits (lower latency, persistent memory) and new software opportunities (byte addressability to data) that have the potential to be disruptive and signifiantly improve end-user experience.

    I wish soooo much that a manufacturer finally invest heavily in the manufacturing of SCM that have high endurance (ex: Nantero Carbon Nanotubes NRAM, MRAM) to finally realise the dream of normally-off computing, with also low latency memory : I would pay a premium for this but I guess the issue is, that the cost (for the added benefit) of such incumbent technology would be too high compare to existing technologies...
  • name99 - Thursday, January 2, 2020 - link

    Things change at first slowly, then very fast. Think smartphones.
    SCM requires
    - changes to the core (instructions to order the write out of cache lines)
    - changes to the caches to propagate those instructions
    - changes to the memory controller to support the new memory class
    - changes to the OS to make best use of the new functionality
    - actual SCM available at a price that makes it interesting to a variety of would-be users

    All of these take time...
    We’re only at about year 2 of a 5 year process.
  • ikjadoon - Tuesday, December 31, 2019 - link

    A great article. I've been loving this "look back" series: it's great to put it all into perspective.

    I totally forgot about Toshiba's XFMEXPRESS design. I'm not sure where it lands, especially if they've still not sent it to an open standards body like PCI-SIG (which felt like the natural fit).
  • realbabilu - Tuesday, December 31, 2019 - link

    I really like your whole drive fill test. That test show the real speed and cache of the controller.
    Made me wonder what storage the data read coming from to fill the fastest ssd
  • jabber - Tuesday, December 31, 2019 - link

    Very happy with the 1TB Corsair 510 I got two days ago. It pumps some data!
  • jabber - Tuesday, December 31, 2019 - link

    As for Optane, Intel needs to stop thinking it's something special and sell it cheap like the rest before the world moves on...which it might have already...
  • milkywayer - Tuesday, December 31, 2019 - link

    Intel has a habit of milking the cow to death. Ala the 4 core cpus they've been flooding the market with for a 7 years until real competition arrived and now they're selling $2000 cpus for half the price and planning i3s with HT just because there is no more free lunch. Thank you AMD.
  • Someguyperson - Tuesday, December 31, 2019 - link

    "...the Phison E16. Drives using that controller launched alongside AMD's latest generation of Ryzen processors, but the new SSDs haven't been a clear winner like those CPUs"

    Well, it would sure be nice if you guys reviewed a single Phison E16 drive so you could actually back that statement up with some facts.
  • Billy Tallis - Tuesday, December 31, 2019 - link

    I have an E16, but the review unfortunately won't be out until after CES. It's the Seagate Firecuda 520, and I also got one of the updated E12-based 510s with 96L NAND so I can show exactly what effect the controller upgrade has vs changing to newer, faster NAND. But I'm hardly the only competent SSD reviewer out there.

    If you care about real-world performance rather than setting a high score on CrystalDiskMark, E16 is obviously not a game-changer—especially after you check the prices. At best, it's a hint at what is to come when the PCIe 4.0 ecosystem matures.
  • HaroldM - Tuesday, December 31, 2019 - link

    Western Digital with its development partner 4DS Memory are collaborating with IMEC to bring into production a new storage class memory called "Non-filamentary interface switching ReRAM".

    This unique Non-filamentary SCM solution enables higher density memory to operate cooler, faster, longer and overcome the limitations of traditional data storage and emerging filamentary technologies.

    We will be hearing a lot about this early 2020.
  • Pro-competition - Tuesday, December 31, 2019 - link

    This article states that "Both Intel's Alder Stream and the Micron X100 are likely to use PCIe 4.0 and second-generation 3D XPoint memory, the last generation of storage technology co-developed by Intel and Micron before their breakup."

    However, the Oct 2019 article states that "Since Micron hasn't said anything about a second generation of 3D XPoint memory being ready, the density and costs of the X100 shouldn't be drastically different from Intel's Optane offerings." (https://www.anandtech.com/show/15029/micron-finall...

    Question: Is the X100 likely to be using second-generation 3D XPoint memory or not? If so, the Oct 2019 article should be updated.

    Thanks for clarifying!
  • Billy Tallis - Tuesday, December 31, 2019 - link

    I'm not aware of a definitive statement from either company about the status of second-gen 3D XPoint or whether that's what is going into their recently announced products. But the longer we go without the X100 announcement turning into real availability or even real specs, the more likely it seems that it's using second-gen 3DXP. Since it's already been two months with no further updates, it looks likely to be a few more months before it makes visible progress toward release, and that's long enough to make the use of first-gen 3DXP seem really unlikely (and unwise).
  • HaroldM - Wednesday, January 1, 2020 - link

    @Billy Tallis, Interested to get your take on Non-filamentary interface switching ReRAM.
  • Billy Tallis - Wednesday, January 1, 2020 - link

    For what it's worth, the latest performance numbers leaked from Intel about their second-gen Optane SSD point to roughly a 25% improvement in QD1 random read performance relative to my previous benchmarks of the first-gen P4800X. That would be impressive but not entirely unbelievable improvement if it was a result of using the same 3DXP memory with a newer controller and newer Linux kernel.
  • IntelUser2000 - Wednesday, January 1, 2020 - link

    I think the high end Optane drives need a bigger sequential boost, because its fast enough for random speeds.

    Unfortunately it seems its limited by power consumption, especially with writes. The Micron X100 has a 6-pin power connector, so there's a possibility of exceeding 75W. The 9GB/s stated speed is 4.5x the bandwidth of the 905P drive, and multiplying 18W by 4.5x would equal 80W which coincides with needing an external power connector.

    It's possible Gen 2 media might improve this, but considering how much lower power the lower specced Optane Memory is, maybe the improvement needs to come from the controller side, possibly with new uarch and/or new process.
  • s.yu - Wednesday, January 1, 2020 - link

    I appreciate how that slide shot was perspective corrected, very few media outlets do that.
  • Tomatotech - Wednesday, January 1, 2020 - link

    Two questions:
    - As peak ssd transfer speeds bump against SATA / PCIe ceilings, I was expecting to see more attention to random transfer speeds. In my experience random transfer speeds matter far more than sequential peak speeds. However this doesn’t seem to be happening - random speeds still stumble along at a few dozen MB/sec. Could anyone explain why? Am I overrrating random speeds or are they truly difficult to raise?

    - What is the future of NVME in the customer space? We have SATA I, II, & III and PCIe 1.0 - 4.0 etc. Is there a NVME 2.0 planned? (Article only mentions NVME 1.4 which makes little difference.) Is there a cap to NVME max transfer speeds?
  • Death666Angel - Wednesday, January 1, 2020 - link

    Your second point: you are confusing software protocol with hardware standards. NVME replaced AHCI for flash storage (connected via SATA), but it still uses PCIe as the physical transfer protocol. AHCI is only on version 1.3.

    As for the random speeds: Bench shows current SSDs having peak 60 to 80 MB/s 4kB QD1 random read and 300 to 500 MB/s 4kB QD1 random write numbers for current drives. So not necessarily something I would equate to "a few dozen MB/s". Considering that random accesses are probably smaller in nature, this already seems pretty good and is better than SATA SSDs manage. The Optane performance in random read shows that there seems to be a fundamental drawback to flash storage in that area. Enhancing sequential speeds can be done by throwing more processing cores at the problem, having more NAND flash die in parallel. Getting better random performance has to come from increased single thread throughput, higher flash die transfer rates, which is more difficult.

Log in

Don't have an account? Sign up now