News, tips, advice, support for Windows, Office, PCs & more. Tech help. No bull. We're community supported by donations from our Plus Members, and proud of it
Home icon Home icon Home icon Email icon RSS icon
  • Test-driving Intel’s Optane in a new PC

    Posted on Tracey Capen Comment on the AskWoody Lounge

    Home Forums AskWoody blog Test-driving Intel’s Optane in a new PC

    This topic contains 17 replies, has 7 voices, and was last updated by  rc primak 1 month, 4 weeks ago.

    • Author
      Posts
    • #1829832 Reply

      Tracey Capen
      AskWoody MVP

      Can Optane really deliver on Intel’s claim of “SSD-like responsiveness from a conventional spinning-platter drive?” A brand-new PC with a factory-inst
      [See the full post at: Test-driving Intel’s Optane in a new PC]

    • #1830606 Reply

      mn–
      AskWoody Lounger

      Well, Optane is sort of weird.

      You’d want to use it as a mostly-read cache for frequently fetched content, given the benchmark results at https://www.phoronix.com/scan.php?page=article&item=intel-optane-16gb&num=2 … for write-constrained applications a traditional high-end SSD (with lots of chips to write in parallel), suitably overprovisioned of course, is better for the price.

      (Oh and on Linux it’s just another block device that you can configure any way you choose.)

      What I find particularly annoying, is that Intel RST on Windows doesn’t ALLOW you to use a standard / non-Optane SSD as a cache device in some cases now, even when you know in advance that your workload will be write-heavy… used to be allowed in previous-generation systems, built one last year that worked fine… wasn’t an option to configure it so with this year’s models any more.

      2 users thanked author for this post.
      • #1834829 Reply

        Ascaris
        AskWoody_MVP

        What I find particularly annoying, is that Intel RST on Windows doesn’t ALLOW you to use a standard / non-Optane SSD as a cache device in some cases now, even when you know in advance that your workload will be write-heavy… used to be allowed in previous-generation systems, built one last year that worked fine

        Annoying, but not particularly surprising.  About as surprising as Google, an ad company, moving to limit ad-blocking in their browser!

        I am not sure what the difference is between using a standard NVMe x4 SSD and this special Intel Optane SSD are supposed to be (other than your report that Intel disabled the feature on non-Intel drives in their latest software).  According to ZDNet, Intel claims Optane’s sequential read speeds are “seven times faster” than than those of competing NVMe SSDs,” (Emphasis mine).   The article then states that “Intel promises sequential read speeds up to 2500MB/s and sequential writes of 2000MB/s,” and goes on to compare Optane to a series of SATA 3 SSDs, where (of course) it wins handily.

        If you want a more apples-to-apples comparison, the Samsung 970 Evo is good for 3500MB/s on read and 2500MB/s on write… unless I’ve made an error in thinking that’s one of the competing NVMe SSDs Intel was referring to, that’s not what I would call “seven times faster.”  The Intel is seven-tenths as fast on sequential reads… maybe they slipped up and misplaced the decimal!

        Group "L" (KDE Neon User Edition 5.16.4).

        • #1835173 Reply

          Ascaris
          AskWoody_MVP

          I think I may have a part of what Intel was getting at.  Maybe.

          The reported sequential speeds in the Samsung 970 Evo are possible because of internal caching, particularly with regard to writes, and once that cache (buffer) is exhausted in a sustained write, performance drops precipitously, down to the much slower native speed of the TLC NAND.  The Pro series doesn’t do that, I guess, with its NAND able to run at the rated speed without limit.  I’m guessing that the Pro series is not in the group that Intel considers Optane’s competition.

          Still, Optane still just seems to be a pretty standard NVMe SSD with a marketing campaign attached, at least in terms of hardware. They keep calling it “memory,” which while technically correct, is only further confusing consumers who already have a hard time differentiating between primary and secondary storage.

          Group "L" (KDE Neon User Edition 5.16.4).

          2 users thanked author for this post.
          • #1840883 Reply

            rc primak
            AskWoody_MVP

            They keep calling it “memory,” which while technically correct, is only further confusing consumers who already have a hard time differentiating between primary and secondary storage.

            This follows the trend where the storage of many devices which have eMMC storage is listed as “memory”. Users think it’s RAM, but it is entirely separate.

            So is that tablet really 16 GB RAM, or is it 16 GB onboard storage (and 2 GB system RAM)? It can be hard to tell from the advertising sometimes.

             

            -- rc primak

        • #1840818 Reply

          rc primak
          AskWoody_MVP

          About as surprising as Google, an ad company, moving to limit ad-blocking in their browser!

          Um… Google is not limiting ad blocking in their Chrome and Chromium browsers. What they are doing is to limit the length of the lists upon which blacklisting extensions now rely. And this is a good thing for browser stability and performance. Those excessively long block lists slow down and destabilize Chrome, and add to its reputation for being a memory hog or having “memory leaks”.

          There is also a Google API (webRequest API in Manifest V3) which is now being restricted to paying Enterprise users.

          Non-list extensions like NoScript are not being limited in any way. Abine Blur is sort of a hybrid, so I don’t know what to expect with it. Ghostery is another hybrid, and may need to be redesigned. None of these extensions is being banned or limited, last I read, from Google or from the extension information Home Pages and Blogs.

          If your block list is more than 150,000 entries, you are seriously damaging Chrome’s performance and stability, and wasting RAM and Internet bandwidth.

          See:

          Google to restrict modern ad blocking Chrome extensions to enterprise users

          https://9to5google.com/2019/05/29/chrome-ad-blocking-enterprise-manifest-v3/

          Google says Chrome isn’t killing ad blockers
          It’s just trying to make extensions safer.

          https://www.cnet.com/news/google-says-chrome-isnt-killing-ad-blockers-its-making-them-safer/

          So yeah, changes are in the air. But ad blocking is still alive and well on all Chromium based web browsers, including the upcoming versions of Edge, Opera and Vivaldi. It will just have to trim its weight.

          -- rc primak

          • #1843489 Reply

            Ascaris
            AskWoody_MVP

            Um… Google is not limiting ad blocking in their Chrome and Chromium browsers. What they are doing is to limit the length of the lists upon which blacklisting extensions now rely.

            Google is blocking the bit of the WebRequest API that allows most adblockers to work.  They propose to replace it with DeclarativeNetRequest, a much less powerful API.  The new API would come with a predetermined limit for the number of rules it will allow to be enforced, but that’s far from being its only limitation.

            The proposed blocking method that would have to be used is not compatible with most ad- and script-blockers, with the only major one that would still work being Adblock Plus, the old, heavyweight addon that was outclassed by much lighter, faster challengers like uBlock Origin years ago.  ABP is also the only one of the major blockers that advertisers can pay to have it allow their ads to go through it by default.

            Google is limiting the number of adblocking rules that it will allow, and they’re preventing most popular adblockers (and certainly all of the good ones) from working at all.  If that’s not “limiting adblockers,” then what is?

            And this is a good thing for browser stability and performance. Those excessively long block lists slow down and destabilize Chrome, and add to its reputation for being a memory hog or having “memory leaks”.

            If this is all about protecting user choice like Google claims, they should let the user make the choice.  If someone is concerned about slowing down or destabilizing the browser, he really should use a lightweight adblocker like uBlock Origin rather than the old, heavyweight Adblock Plus… but with Google’s changes, you can’t.  To be that lightweight on resources, it relies on the API that Google’s reserving for its enterprise customers (who, presumably, have no need for the alleged stability or privacy “benefits” of getting rid of that API).

            Non-list extensions like NoScript are not being limited in any way.

            Indeed, since NoScript runs on Firefox only. <g>  Chrome users have had to do without this addon, but other offerings like Ghostery, Privacy Badger, ScriptSafe, and uMatrix have filled in the gap.  Unfortunately, none of them will work with the changes Google has proposed.  The article you linked quotes Ghostery president Jeremy Tillman:

            “This would basically mean that Google is destroying ad blocking and privacy protection as we know it…”

            The EFF and uBlock Origin/uMatrix author Raymond Hill have made similar comments.

            If your block list is more than 150,000 entries, you are seriously damaging Chrome’s performance and stability, and wasting RAM and Internet bandwidth.

            The limit was 30,000 not that long ago, which would have disallowed even half of the most popular adblocking list, Easylist, from loading.  Google’s bowed to pressure on the rule count, but the Webrequest API needed by all those addons listed above will still be crippled under the new plan.  The low rule count was only one of the problems with the new API.

            Google can make any claim they want, but if the ad giant claims that preventing most ad- and script-blockers from working is for any reason other than to promote pushing ads, I would recommend taking it with a 50-pound water-softener size block of salt.  Google listed adblockers as a threat to their profits in their last quarterly report, and here we have the most efficient and effective blockers being broken by deliberate action by Google.  I believe that’s for the benefit of the users about as much as I think Microsoft’s WaaS is for the benefit of the users.  If it was, there would be no exceptions for enterprise customers, since they would want to benefit from the goodness too, if there was any.

            Group "L" (KDE Neon User Edition 5.16.4).

            3 users thanked author for this post.
            • #1850288 Reply

              rc primak
              AskWoody_MVP

              If what you are saying is true, Chrome and all its derivatives (including Edge and Opera) would be rendered useless for a significant minority of users. Firefox would still be useful, if I read all of this correctly.

              This is all Google’s loss as far as I’m concerned. Losing Chrome is only a problem for me if Firefox would not work on some future Chromebook purchase I may make. Right now, Fedora (30) Linux won’t work with the CHRX install script, and there are no Base Images from which to redirect the script. So I now use Fedora 30 without Chrome OS on that Chromebook. (The Anaconda Installer and its Blivet partitioning don’t allow compatibility with the way Chrome OS boots.)

              I would think that if a large number of ad blockers and script blockers (and some security add-ons) are being broken, the developers of these add-ons would pressure Google to modify what they are doing. Several big antivirus companies are among those whose products would be compromised if all of this is true and if it actually goes into effect in a future release of the Chrome Browser.

              If Google chooses to ignore all this pressure, Mozilla and Firefox would benefit, to Google’s loss. It remains to be seen if the ad revenue Google would lose in such a mass-exodus would offset the amount of revenue they would recapture from users who are too lazy to switch browsers. I’m guessing Google will lose a lot of revenue, but not enough to make them change their course.

              The real losers would be end users who don’t switch from the Windows default (Edge) to Firefox.

              Personally, I don’t think Google will follow through with all of their threats. Ad blocking is not dead. If for no other reason than that Firefox continues to allow it, and web sites can’t and won’t go to all the trouble they would have to, just to block all Firefox users. Doing so (blocking Firefox) would almost certainly lead to lawsuits, which would cost site operators much more than all their lost ad revenues.

              Some sites have in the past tried to block Firefox, and they got nowhere. The future would reflect the past in this regard. But more paywalls are certainly in all of our futures. And more dirty tricks like stealth cryptomining, which I currently block in all my browsers.

              -- rc primak

              • This reply was modified 1 month, 4 weeks ago by  rc primak.
              1 user thanked author for this post.
    • #1835511 Reply

      doriel
      AskWoody Lounger

      I dont know all of your experience, but since Intel is using NVMe SSD (Optiplex 7060+), for example Acronis True Image doesnt recognize these drives, when you doot from DLC USB – no full backup possible 🙁

      Another issue I had was during TYPICAL installation of windows” Windows 10 AiO build 1803 (x86 and x64 architecture) doesnt recognize these drives either and installation cannot be completed. I tried to download manually drivers from Samsung, but this is so complicated and even after few hours of trying differen versions, I had success with:

      BIOS mode (no UEFI)
      SecureBoot – Disabled
      LegacyBoot – Enabled

      Obviously, I am really bad at this, or I am missing something important. Did someone had same issue?

      Thanks and good luck.
      Doriel.

      I have not failed. I've just found 10,000 ways that won't work.
      --- Thomas A. Edison

      2 users thanked author for this post.
    • #1836110 Reply

      wavy
      AskWoody Plus

      Lance sorry to hear you are not satisfied with your new laptop, I would remind you of the price performance ratio for laptops is high and of course the ability to upgrade is negligible w/ a lt.
      The optane usage at present would seem to be Intel trying to market a nice technology whose time has not come. Just to pricey for the base memory of a consumer product. Do you have a return window??

      🍻

      Just because you don't know where you are going doesn't mean any road will get you there.
      • #1840854 Reply

        rc primak
        AskWoody_MVP

        Minor point — the article is by Fred Langa, not Lance Whitney.

        -- rc primak

    • #1836223 Reply

      Cybertooth
      AskWoody Lounger

      When “building” my newest computer on HP’s website, this Optane thing was offered. However, it was available only for Windows PCs and the plan was to put Linux on the new computer, so instead of Optane I selected a bigger SSD for the build.

      Honestly, I didn’t understand the point of Optane back then and I still don’t understand it: just get more RAM or a bigger/faster SSD, make it NVMe if the hardware allows it.

       

      2 users thanked author for this post.
      • #1838043 Reply

        mn–
        AskWoody Lounger

        And I really thought originally that Optane would’ve been the modern equivalent to the old Solid Data Excellerator.

        Those used to be able to read OR write at approximately max link speed. (Which, at the time, was U320 SCSI or 2G FC, IIRC.)

        (Oh well, soliddata.com seems to be down nowadays, but there’s at least an old piece of news still up at https://www.itworldcanada.com/article/solid-data-excellerator/35110 and…)

        1 user thanked author for this post.
        • #1838456 Reply

          Cybertooth
          AskWoody Lounger

          That article is from 2000. With that in mind, check out the following excerpt:

          Dubbed the Excellerator 100, 800 and 1000 Ultra SCSI, these drives use solid state disk technology to speed access to data.

          Wow, SSDs in the year 2000? That’s amazing, I had no idea they’d been around that long.

          And now for the other quote:

          The Model 100 is a 1 7/8-inch high Ultra SSCI rack-mountable drive that holds from .5G byte to more than 2G bytes of data.

          2GB capacity. How things have changed since then!

           

          • #1839955 Reply

            mn–
            AskWoody Lounger

            Wow, SSDs in the year 2000? That’s amazing, I had no idea they’d been around that long.

            Heh, actually, they’ve been around longer… it’s just, they weren’t mainstream products. There were two main lines of development, one for ruggedness (no moving parts) and the other for performance.

            2GB capacity. How things have changed since then!

            And rackmount chassis, for that single 2GB unit already. The 1000 was… hm, standard 19″ rack for sure, I think a 3U chassis… been a while.

            This is the type that was built out of DRAM modules, and then additional support parts to make it seem like a disk and not lose data during a power outage.

            I really did hope after the launch hype that the Optane would’ve been usable for that performance niche nowadays.

            1 user thanked author for this post.
    • #1840877 Reply

      rc primak
      AskWoody_MVP

      My decision to go with a small form factor desktop “nano-PC” (my Intel NUC) made a larger SSD the logical choice. Compared with Optane, bigger SSDs are more expensive, but more flexible. My SSD is the M.2 form factor.  At about $100 per TB, this type of SSD is not a bad choice for a lot of folks buying a new PC or laptop.

      As posted earlier, Linux does not use Optane, due to not having suitable driver software, among other differences from Windows.

      In a laptop or 2 in 1 I would be looking for Linux dual-boot capabilities, so Optane would not fit my use scenario.

      As we will probably see in the third installment, Optane also may not be the best solution for virtual machine environments.

      -- rc primak

      1 user thanked author for this post.
      • #1841654 Reply

        mn–
        AskWoody Lounger

        As posted earlier, Linux does not use Optane, due to not having suitable driver software, among other differences from Windows.

        Well… it’d be more accurate to say that Linux doesn’t use the Intel special mode for Optane, so it isn’t good for a dual-boot system.

        In a Linux-only system you can use the block device just fine as a lvmcache, zfs L2ARC, or whatever. Even as your boot disk if you really want to.

        1 user thanked author for this post.
        • #1842939 Reply

          rc primak
          AskWoody_MVP

          Not a subtle difference granted, but this still makes Optane not suitable for dual-booting with Linux, as you say.

          -- rc primak

    Please follow the -Lounge Rules- no personal attacks, no swearing, and politics/religion are relegated to the Rants forum.

    Reply To: Test-driving Intel’s Optane in a new PC

    You can use BBCodes to format your content.
    Your account can't use Advanced BBCodes, they will be stripped before saving.