Slashdot Mirror


Traditional PC Sales Continue To Slide (zdnet.com)

Sales of traditional PCs continue to decline, although the overall PC market is likely to grow slightly next year. From a report: Traditional PC shipments are forecast to drop by nearly eight percent this year, and another 4.4 percent in 2018, predicts analyst firm Gartner. Which means that, by 2019, 16 million fewer traditional PCs and notebooks will be sold than were shipped this year. However, much of this will be offset by the rise in spending on high-end notebooks like Microsoft's Surface and Apple's MacBook, so that the overall PC market will by 2019 be at pretty much the same level it was last year. Tablets -- defined by Gartner as basic and utility ultramobile devices -- will also decline over the period to 2019.

36 of 223 comments (clear)

  1. Builders vs Buyers by Anonymous Coward · · Score: 3, Interesting

    Based on the traffic at places like Microcenter I'd offer this: more people have the basic skills now to build a computer on their own rather then buy one. I haven't met a single person in over 10 years that bought a computer, everyone built their own. Corporations are buying laptops for telecommuters and staff versus bulky PCs that are easier to transport, deploy, and use less power for the workloads they deal with.

    1. Re:Builders vs Buyers by Anonymous Coward · · Score: 5, Insightful

      I used to build my own computers to try to wring every last ounce of performance out of well selected components; nowadays the technology is so fast it just isn't worth it. I'll buy mass market commodity machines for dirt cheap the run circles around even the most ambitious builds I used to do.

      It's a dying art.

    2. Re:Builders vs Buyers by ZorinLynx · · Score: 5, Interesting

      Yeah, it's sad but true.

      Even with gaming systems, unless you're the sort who wants a flashy LED-lit clear-side case that looks like a spaceship, you can build a top-performing system just by buying a mass market machine and selecting the video card you want to put into it.

      The enthusiast market is still there, but it's mostly for people who want their computer to look like it belongs on the set of Star Trek. I (sadly?) outgrew that once I was out of my 20s.

    3. Re:Builders vs Buyers by oic0 · · Score: 2

      Buying is more cost effective if you want a workstations or web browsing PC. If you want something fast you have to build yourself or you get gouged. Especially if you want a decent video card. They won't just sell you fast guts in whatever you want. They try to make it like buying a car. You have to go on up to the expensive model before you can get a real V8, etc..

    4. Re:Builders vs Buyers by sycodon · · Score: 2

      Just picked up an iCore 5, 12 gigs, 2 terabtes HD...$197

      Great for porn, web, and most games older than about 4 years ago. It's runs SQL Server 2012, VS 2012, and some other IDEs environments just fine.

      Why would I pay $1k for anything else?

      --
      When Fascism comes to America, it will call itself Anti-Fascism, and tell you to give up your guns.
    5. Re:Builders vs Buyers by avandesande · · Score: 2

      it's the stuff that comes out of a demon when you stab him

      --
      love is just extroverted narcissism
  2. longer lifetime by Anonymous Coward · · Score: 5, Insightful

    I think people are keeping their machines for longer and longer as time goes on.

    1. Re:longer lifetime by Headw1nd · · Score: 4, Insightful

      This is most likely the case. Twenty years ago, a six year old computer was basically worthless. Now unless it's for high-end gaming, a six year old home computer is fine.

    2. Re:longer lifetime by ZorinLynx · · Score: 5, Insightful

      Systems are not becoming obsolete as quickly as they used to.

      I remember how back in 1995, a computer from 1991 was considered slow as hell and could hardly run any current software.

      Now in 2017, a computer from 2013 is still perfectly usable and fast. The rate of performance increase has slowed to an utter crawl. The biggest advancements in recent years have been reduced power consumption and increasing density in solid state storage, and the latter can be an upgrade to your old machine.

    3. Re:longer lifetime by darkain · · Score: 4, Insightful

      It isn't so much that performance has not continued to increase. It is that we've hit the point of diminishing returns for 99% of applications. a single-core 1GHz processor will run Microsoft Word about the same as a 10+ core 3+GHz CPU. And with even low end budget GPUs nowadays offering hardware decoding of 4k h.264, the rest of the computational power of the CPU and GPU isn't really meaningful for the majority of consumers.

      Gamers, content producers, and scientific researchers are really the only fields left to push the boundaries of computational power.

    4. Re:longer lifetime by jwhyche · · Score: 4, Interesting

      It isn't "most like" the case. It is the case. Fifteen years ago I was changing out processors or memory every 6 months. Now I design my personal workstations with a 5 year lifespan. I have a friend that did the same thing. Now he has a 5th generation i7 that is almost 3 years old. He said he has no intention of upgrading anything till something fails or some radical advancement in technology comes out.

      I'm in the same boat. I have a i7-6700K which is a 6th generation processor. Intel just announced they are about to release the 8th generation i7. From looking at what is being release I see no real performance gain over my 6th generation.

      My linux server is a centos 6 running on a AMD 8350. As a server it has different requirements than my workstation. Its my file/plex server. It's been doing that role since 2013. It is entering its 5th year in that role and I see no reason to update it for the next 3 or 4 years.

      As for your mundane computer users, the PC used to be a oddity. Now they treat a PC more like an appliance than a oddity. Its more like a refrigerator than a computer. You don't upgrade your refrigerator every year or so. This is even if they have a PC. Most now have laptops and treat them the same way.

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
    5. Re:longer lifetime by jwhyche · · Score: 4, Interesting

      You can blame that on Intel and their sandbagging

      No you can't. I doubt that Intel is sandbagging anything. If they where then AMD would be running away with clockspeeds. But if you look at AMD processors you will see they are running at the same clock speeds as Intel.

      The reason processors are not going any faster is we have hit the limit of what can be done with the current technology. Over the next few years we might see a increase in a few hundred mhz here and there. But there will never be another leap of 2 or 3 ghz again. Not with what we have.

      To go any faster it will take a radical shift in the basic process of comptuer. Quantum processors maybe.

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
    6. Re:longer lifetime by jwhyche · · Score: 2

      You are not the only one that has expressed a interest in this. It started doing that a few months ago after I sent a email to have some issues fixed with my account. I have sent a few more emails to point that out but never got a reply on them.

      So, I figure if the powers that be seem to be fine with me posting at a +3 why should I give a fuck?

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
    7. Re:longer lifetime by Solandri · · Score: 2

      GHz for GHz, Kaby Lake (Jan 2017 desktop release) is only about 20% faster than Sandy Bridge (Jan 2011 desktop release). 20% improvement in 6 years. I'm still telling people with Sandy Bridge systems not to bother upgrading. Unless you want more cores (i3 to i5 or i7), some of the newer features (like USB-C support), or want lower power consumption, there's no reason to stop using a Sandy Bridge system.

      Clock speed has also been relatively static. 3.6-3.9 GHz in 2011 to 4.4-4.5 GHz in 2017. A 19% increase. Combine the two and you get an underwhelming 42% improvement in processor speed in 6 years.

      Even as recently as the late 1990s, by the time a system was 6 years old, it was far beyond obsolete. For example, November 1995 saw the release of the 200 MHz Pentium Pro. If you skip ahead 6 years, you find the 2.0 GHz Pentium 4 released in August 2001. A 900% increase just in clock speed in 6 years.

      PC sales are lackluster because there's no reason to upgrade at anywhere close to the rate we had to upgrade in the past.

    8. Re:longer lifetime by KozmoStevnNaut · · Score: 2

      My desktop is from 2010, and it still keeps up quite. To be fair, it's a 3.2GHz 6-core Phenom II with 16GB RAM, so reasonably high-end for the time, but I honestly think I could keep using it just fine for at least another 3-4 years, before noticing any shortcomings. The graphics card (GTX460) is its weakest point by far, once I upgraded to an SSD.

      --
      Eat the rich.
    9. Re:longer lifetime by Opportunist · · Score: 2

      20 years ago, a 2 year old computer was basically something you toss at your poor friend so he can play some old games while you bought the bleedin' edge computer that allowed you to at least run the new stuff at decent framerates.

      --
      We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
    10. Re:longer lifetime by jwhyche · · Score: 2

      GHz for GHz, Kaby Lake (Jan 2017 desktop release) is only about 20% faster than Sandy Bridge (Jan 2011 desktop release). 20% improvement in 6 years. I'm still telling people with Sandy Bridge systems not to bother upgrading. Unless you want more cores (i3 to i5 or i7), some of the newer features (like USB-C support), or want lower power consumption, there's no reason to stop using a Sandy Bridge system

      This is the reason I upgraded from my old AMD 8150. It wasn't because of processor performance. I upgraded because the features I anted, USB-C and decent m.2 support, where not available on AMD at the time. Large jumps in PC performance are not because of processors but due to better I/O devices like PCI-e M.2 modules.

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
    11. Re:longer lifetime by jwhyche · · Score: 2

      I don't believe we will be seeing a shortage of uses for dual core cpu's for a very long time. Of course I think I said the same thing about single cores too, so take that with a gain of salt. I would say that most things that a PC needs to do can be done quite nicely with a dual core CPU..

      I think the real performance increases for the next 10 years will come in storage. One of the reasons I upgraded was to get a m.2 module for my workstation. There is still plenty of room for performance increases in that area.

      Memory still lags behind processor performance too. DDR4 is fast but there is still plenty of room there too.

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
    12. Re:longer lifetime by jwhyche · · Score: 2

      I would go farther and say even most extreme users don't need more than 4 core or 8 threads. I'm considered a power user by most who know me. I'm also a creator in second life. It's not unheard of for me to have 2 firestorm viewers up and running on one desktop, blender open on another, photoshop on the third, and all the while there is a VM running in the background where my shoutcast station is running to provide a stream to the sims I manage. Doing all this and my processor rarely clocks more than 40% across all 4 cores.

      Lots of people don't realize how powerful modern desktops really are.

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
    13. Re:longer lifetime by jwhyche · · Score: 2

      That has really only been true with the bulldozer core from AMD. Most of the time AMD only lagged behind intel by a few percent, not enough to pay an extra hundred or more bucks for. The bulldozer cores are where AMD fumbled the ball, but even then the processors where still more than enough for most people. Hell, I bought two of them.

      I think a lot of it depends on the software too. I've noticed that when running windows on similar class processors, one AMD and one intel, the AMD system always seems to be more sluggish than the intel. I've noticed nothing like this on linux systems. In-fact my 8350 sometimes out performs xeon class machines on some tasks.

      Of course nothing here is new. Most of us already know that gaming engines are compiled and optimized for the most common processor out there. An that is a quad core intel based system.

      --
      I read at +2. If your post doesn't reach that level I will not see or respond to it.
  3. Pretty simple since 2005... by alexandre · · Score: 2

    Buy 4x the amount of ram people buy, be good for 10 years.

  4. NVMe and M.2 ports will likely boost PC sales by RobinH · · Score: 4, Informative

    I opted to just install an SSD rather than upgrade our PCs at home, and definitely got a few extra years out of them. However, the SSDs are maxing out the data rate of the SATA ports, and now they're coming out the NVMe drives that are 4 times faster than SSDs but you need an M.2 port (as I understand it a direct connection into your PCI bus). For these you typically need a new motherboard. So whereas the upgrade from an HDD to an SSD was very simple and easy, taking the next step means a new motherboard, so if you were already delaying buying a new PC because you did the SSD thing, you're almost certainly going to buy a new PC next time. There's a lot more incentive now.

    --
    "I have never let my schooling interfere with my education." - Mark Twain
    1. Re:NVMe and M.2 ports will likely boost PC sales by Kjella · · Score: 3, Interesting

      Nah, you can just get a $15 adapter and put it in any PCIe 4x slot. By the way, if you really want crazy performance get a X399 motherboard with PCIe bifurcation and load it up with quad NVMe cards and you can go nuts with 28GB/s in an 8-way RAID 0 configuration. Not that you'd really notice at consumer queue depths.

      --
      Live today, because you never know what tomorrow brings
  5. Re:Windows 7 support is needed by viperidaenz · · Score: 2

    also... "We need Microsoft to invest a crap-ton of money to further enhance their old software, for no extra money, to support it after the time they said they would when you bought it."

    Good one.

  6. Re:This trend will destroy Firefox. by Bert64 · · Score: 5, Insightful

    The decline of desktops is more to do with average users realising that a complex computer complete with maintenance requirements and malware risks is not the best choice for someone who just wants to read facebook. These people are better off with an ipad, and they are also the sort of people who will just use whatever browser the machine came with not realising anything else exists.

    --
    http://spamdecoy.net - free throwaway anonymous email - avoid spam!
  7. Old / refurb PCs are good enough by Stormwatch · · Score: 2

    Why would you buy a new machine? You can get a $60 refurbished C2D that will meet a regular user's needs comfortably. Or you can get a $150 i5, just add a $150 video card, et voila, a solid gaming rig for the same price as a current console. Some people may want something more high-end, but that's a bit of a niche.

  8. Re:The x86 PC and security. by jwhyche · · Score: 2

    People never liked the x86 PC

    No. Back in the day a few hackers and engineers didn't like the x86 PC. I wasn't a big fan of it back then ether. I liked the 68K line.

    Then and now most people didn't give a shit what processor their PC was running as long as it did the job. I'm willing to bet that most of the people out there don't know what processor their phone is running. I bet even less of them realize that their smart phone is a really just a mini computer they can carry.

    Now today most hackers and engineers embrace the x86 PC because it is the best game in town. Sure the ARM processors are doing a good job and have the phone/tablet in the box, but the world runs on x86 processors.

    --
    I read at +2. If your post doesn't reach that level I will not see or respond to it.
  9. Re:Windows 7 support is needed by Stormwatch · · Score: 2

    And Windows 10 is a glitchy mess with a butt-ugly interface. Is it a surprise that people cling to the old alternative that actually works better?

    Oh, and Windows is a desktop operating system. No one gives a fuck about those retarded mobile apps.

  10. Not exactly by rsilvergun · · Score: 2, Insightful

    Dell & HP often skip the full sized PCI-E slot for graphics entirely. When they do include it they've been known to use boards that can't deliver enough power on the slot. Finally their power supplies often lack the extra plug needed for most video cards. Asus & Acer are a little better, but it's not a sure bet.

    The major manufacturers all sell 'gaming' pcs and they'll be damned if you're going to buy an i5 equipped PC on sale for $400, stuff a $200 graphics card in it and get 95% of the performance of their $1200 gaming rigs. They figured that out trick out in the late 90s/early 2000s.

    --
    Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
  11. Re:The x86 PC and security. by drinkypoo · · Score: 2

    People never liked the x86 PC

    No. Back in the day a few hackers and engineers didn't like the x86 PC. I wasn't a big fan of it back then ether. I liked the 68K line.
    Then and now most people didn't give a shit what processor their PC was running as long as it did the job.

    I don't think they mentioned x86 because they thought people didn't like the x86. I think they mentioned it because all those other things were also personal computers, and they wanted to differentiate. It was true, too; until about the Windows 95 era, a PC was considered by many if not most people both inside and outside of the industry to be the lame but affordable option. If you were in the industry, you were comparing it to "real" computers like ones from Sun, or even other machines from IBM; if you weren't, you compared it to a Mac, or even an Amiga or Atari (both of which offered much more functionality for less money.) Remember, while the dominant word processor on the PC was still WordPerfect 5.1 for DOS, everyone else was using WYSIWYG with onscreen graphics and scalable fonts. People would drop right back into DOS out of Windows 3.1 to run WP, making Windows a glorified task launcher.

    Anyway, until it sprouted a PCI bus, the x86 PC was a pathetic joke to almost everyone. Before that there were only occasional efforts to build "real" machines around x86 chips, all of which were economic failures. There was an 8-way 486 which would run SunOS, for example, and Sun's own 386-based i86pc. There was the brief flirtation with the EISA bus, and the less brief but more restricted to IBM MCA bus, but these were both crap without even the style of other personal computer buses like Zorro (Amiga) or NuBus (Macintosh) which offered autoconfiguration without floppy disks. VLB was also garbage; in theory you could use two bus-mastering, DMA-transferring devices on the same PC, but in practice this was usually problematic and required extensive troubleshooting to make work if it would in fact work at all.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  12. Re: The x86 PC and security. by Miamicanes · · Score: 2

    I was a die-hard m68k fan... but in retrospect, that's because back in the early 90s (right before DOS4GW), writing programs in anything besides realmode assembly was damn-near impossible.

    I remember how I discovered (sometime around 1992 when I was in college) that every x86 from the 386dx onwards HAD orthogonal registers & could do flat addressing... then went in literal circles for MONTHS trying to find anything that resembled documentation or development tools.

    To this day, I have no idea whether published books about the topic even EXISTED circa 1992, or whether MASM circa '92 could actually DO protected mode. I basically gave up after someone on Usenet disillusioned me... basically, protected-mode (circa '92) meant no BIOS calls (DOS extenders didn't quite exist yet). No BIOS calls meant you couldn't even output characters to the screen without knowing more about the inner workings of a VGA card's CRTC than any 18 year old could have hoped to know at that point (the first meaty books explaining videocard programming weren't readily available yet... they technically existed, but if you didn't already KNOW their titles & author names, your likelihood of discovering them was basically "nil" unless you were rich enough to blindly order expensive books from B. Dalton's or Waldenbooks sight-unseen based on their titles alone... and wait 6-8 weeks for them to arrive. Information-wise, it really WAS the Dark Ages compared to now.

    Pre-Google, finding stuff about esoteric subjects was *hard* -- even as an undergrad at a major research university that HAD internet access... you'd post to Usenet, but if you missed seeing a reply, it was *gone* forever (as far as you were concerned) a few weeks later (from what I recall, the university's admins purged Usenet posts after ~1-4 weeks... 1 week for alt.binaries.*, 4 weeks for comp.sys.*). The library had computerized indices, of course... but you still had to try and FIND the bound journal. Half the time, it was missing. The other half of the time, it was in limbo (the journals for that year pulled from the shelf for hard-binding, but not actually BOUND yet). And if you weren't a professor or grad student, the library staff had zero interest in tracking missing resources down for you.

    Come to think about it, the "good old days" actually sucked pretty badly.

  13. Re: The x86 PC and security. by jwhyche · · Score: 2

    I would agree that it was around windows 2000 where the PC as we know it was born. That was when most Linux people I know abandoned Linux and moved to windows.

    --
    I read at +2. If your post doesn't reach that level I will not see or respond to it.
  14. Re:Windows 7 support is needed by Opportunist · · Score: 2

    Fine, charge me a (sensible!) annual fee to continue using Win7 and we're good.

    --
    We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
  15. Re: The x86 PC and security. by jwhyche · · Score: 2

    When I first saw windows xp I thought "what is this crap?" The interface reminded me of AmigaOS 1.3. After about a month I had figured out how to disable all that theme crap and had XP looking like Windows 2000. Once you got rid of that XP wasn't really that bad of a OS.

    --
    I read at +2. If your post doesn't reach that level I will not see or respond to it.
  16. Re: The x86 PC and security. by jwhyche · · Score: 2

    We romanticize them because when you solved a significant problem you were a goddamned hero. It's much calmer and safer to live in an age which does not require heroism, but compared to nostalgia it seems boring.

    Can't argue with you here. Back in the "dark ages" I remember how happy I was to have set up a usenet and email on my Amiga 500 and it actually worked the first time. Hero time.

    We tend to forget the times when we almost burn down our apartment trying to resurrect a toaster oven we pulled out of the dumpster. Just saying....

    --
    I read at +2. If your post doesn't reach that level I will not see or respond to it.
  17. Re:PS4 CPU is two Athlon 5150s by jwhyche · · Score: 2

    While you may have a point the discussion is geared toward gaming PC's.

    --
    I read at +2. If your post doesn't reach that level I will not see or respond to it.