Please create an account to participate in the Slashdot moderation system


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×
Intel Upgrades Linux

Intel's Core i5 6500 Shines As a $199 Skylake Processor, Works With Linux ( 119

An anonymous reader writes: Intel has begun releasing more "Skylake" processors that are cheaper than the launch SKUs of the i5-6600K and i7-6700K. One of the new processors that is now widely available is the Core i5 6500 and it costs just $199 USD — that puts it just a few dollars more than the AMD FX-8370 and significantly less than the higher-end Skylake and Haswell CPUs. At least with Ubuntu Linux, the Core i5 6500 is showing competitive performance that for some workloads puts it faster than Core i7 Haswell/Broadwell processors and much faster than any AMD processors. The Intel Skylake CPUs are fully supported under Linux but the caveat is needing the very latest kernel otherwise there's no graphics acceleration or sound support.
This discussion has been archived. No new comments can be posted.

Intel's Core i5 6500 Shines As a $199 Skylake Processor, Works With Linux

Comments Filter:
  • Ubuntu (Score:1, Offtopic)

    I just installed Ubuntu on a friend's XP laptop a couple of days ago, since XP is not safe to connect to the Internet. The laptop is now much faster, and comes with all the functionality that a non-gamer needs.
    The install process went very smoothly, much better than my Win 10 experience.
    • So you installed an Intel Core I5 6500 into the laptop then?
    • by Anonymous Coward

      Welcome to 2010.

  • Meh. (Score:5, Interesting)

    by Moof123 ( 1292134 ) on Thursday October 15, 2015 @02:26PM (#50737541)

    Moore's law is dead. Core count is stagnant. All recent gains seem to be in incremental power savings.

    In this particular case the 6500 is a fixed 3.2 GHz, while the 6600K is 3.5 GHz as shipped. The 6600K can readily be pushed to 4.2 GHz on air, which is partly why you pay the extra K tax to be able to pull those shenanigans, and you get to leave the cheaper part in th edust.

    • by creimer ( 824291 )

      In this particular case the 6500 is a fixed 3.2 GHz

      IIRC, the 6502 in my Commodore 64 was fixed at 1MHz. Nice to see old tech getting a much need speed boost! :P

      • by Anonymous Coward

        The Commodore 64 used a 6510. It was derived from the 6502 design, but had a few hardware feature (tri-state buffer) that the 6502 didn't have.

    • Re:Meh. (Score:4, Insightful)

      by Anonymous Coward on Thursday October 15, 2015 @02:41PM (#50737659)

      Guess what? Until AMD steps up and delivers a killer product at a killer price you will see nothing but stagnation from Intel.

      • Sad but true. You're paying for some pretty pathetic performance.
    • Re:Meh. (Score:5, Insightful)

      by FlyHelicopters ( 1540845 ) on Thursday October 15, 2015 @02:47PM (#50737711)

      Moore's law is dead. Core count is stagnant. All recent gains seem to be in incremental power savings.

      You discard the improvements in power savings like they are nothing.

      Today you can get the same performance as 5 years ago, for 1/3 the power consumption.

      That is a massive improvement. Moore's Law isn't dead, instead of more performance, Intel has focused on using less power to provide the same, or slightly better performance. Give these new chips 130w to play with and they'll blow away the older stuff. But what took 130w 5 years ago now only takes 45w.

      That is a big deal.

      • Re:Meh. (Score:5, Informative)

        by Junta ( 36770 ) on Thursday October 15, 2015 @03:13PM (#50737907)

        Yeah, that's the interesting thing. If you just follow the desktop parts, performance has been modestly improving and power consumption is dipping. Meanwhile on the server side, power consumption has been more steady and the core count has been going drastically up. Sure workloads that don't scale to many cores aren't getting that much of a boost (and that is the state of most desktop platforms), but areas with multithreaded workload and/or consolidated bunch of single threaded workload have continued to benefit from advances.

        • Laptops are where the really big difference has been, IMHO...

          5+ years ago, the idea of having a thin and light notebook that got 6+ hours of battery life while being useful was a fantasy.

          Today, you can get a really useful laptop for a really reasonable price that has really nice power life.

          The reasonable performance you can get in 15 watts today vs. 5, 10, or 15 years ago is astounding...

          • by Hadlock ( 143607 )

            It's nearly impossible to find a Chromebook with less than 6 hours battery life now. I took pause on buying a $150 chromebook because it "only" had 8 hours battery life, while the competing $149 laptop had 10.5 hours. The new Dell XPS 15 laptop boasts 18 hours continuous use and has a power-sucking 4K display. And yet it's about half the volume of my 2001-era Powerbook G4 that was capable of 5 hours on the lowest brightness setting.

      • Power in a server farm matters, as it does for a laptop, but for a home PC it is not nearly my top concern. Crappy integrated graphics are important for a laptop, but. It for a decent gaming machine. So the top two improvements are of near zero value to me who want a good gaming machine and are frustrated by the focus on GPU and power. I want more core and clock speed without the SEVERE increase in price Intel demands for aa 6 or 8 core CPU.

        • I understand your point and what you want.

          I'm simply pointing out that gains have been made, if not in the area you personally care about.

        • by amiga3D ( 567632 )

          I bought a dual I-7 2ghz quad core equipped mac mini 3 years ago. In all reality it's fast enough. It's the first computer I ever felt that way about. It does everything so quick that the only time I have to wait more than a second or two is when I transcode video. That is time consuming but still pretty good. I don't game so the HD 4K graphics is adequate. After 3 years I feel no pressure to upgrade and always in the past a year after buying or building a computer I'd start craving more speed. If I

          • by laffer1 ( 701823 )

            You don't want to upgrade. The new Mac Mini is slower than the quad core. They have no replacement for the mac mini 2012 quad core model yet without spending $2000 on an iMac. I just went through this and couldn't find a single mac that was faster for under $1800.

            That said, many of the new macs have SSD and better graphics but CPU performance is a joke. I decided to put a SSD into my mini instead and now I have most of the performance of a new mini in terms of IO but much faster processor.

            • by amiga3D ( 567632 )

              I got a deal on a couple of 1TB 2.5 drives that rotate at 7200 RPM. I was going to buy an ssd but the mini's original drives weren't that bad so I picked up two new ones for 110 bucks and it's pretty sweet. I've got my video files on an external 4TB drive in a FW800 case so it's pretty sweet. I'll probably stick with it for another 2 or 3 years. I like the look of the new Mac Pro but it's just too pricey for a hobby.

      • is the enemy of electronics. One of the things I'm wondering is if these low power (and low heat) boxes are going to last a lot longer, and eventually create a glut.
    • Moore's law is about transistor density, not the number of cores or their performance.

      We've reached a point where the amount of computational power in the chips that are being made is in excess of what the majority of consumers need so instead of further increasing the number of cores, which is pointless for most workloads or making the individual cores more powerful, the density improvements are being used to make smaller chips that are more power efficient because the most pervasive computers these day
      • Was just about to post this. The biggest gains in computer performance over the past few decades came from increasing clock speeds and eliminating delays. We're pretty much to the limit of silicon clock speeds, and delays are low enough that it's hard to find new ways to reduce them—throwing more transistors at the problem won't help. Moore's Law hasn't stopped, it's just not relevant to CPU performance anymore.

        And you know what? I don't mind. Software developers only optimize until their code is 'fas

      • Moore's law is about transistor density, not the number of cores or their performance.

        Actually, it's about transistor count.

    • by fnj ( 64210 )

      Moore's law is dead. Core count is stagnant. All recent gains seem to be in incremental power savings.

      Moore's Law says nothing about core count. It says nothing about performance. It only speaks of transistor count. How you put those added transistors to use is subject to certain realities. There are real limits to how effectively you can utilize enormous core counts in general purpose CPUs in desktop or mobile use. In servers you have enormously scaled parallel use, and can make better use of them.

      It looks

    • Saving power saves heat and potentially extends lifespan. So it's a big deal; arguably as important or possibly MORE important than faster cycles.

      The ability to shrink size and still increase processing power while also reducing power consumption and heat generation is massive progress.

      I'm wandering if anybody is arguing Moore's Law is off now because it's been exceeded.
  • Will it work with OpenBSD?
  • I want a $50 processor and a $50 motherboard, which I can easily do with the 1150 chipset. For the 1151 chipset, it's $300+. Of course, that doesn't include memory. Motherboards that uses a newer processor with DDR3 memory, or an older processor with DDR4 memory, cost $150. I might have to go with a transitional motherboard in the near future to bridge price gap.
    • Or do like me: Stop buying shit that represents only marginal improvements.
      I'm sitting pretty on my 4.5 GHz i7 from 4 years ago. The only thing I feel like upgrading is my GPU.

      Can we please get back to upping clock speed?

      • by creimer ( 824291 )

        The 4.5GHz is probably as good as its going to get as far as clock speed. If you haven't noticed, processors have multiple cores because the 5GHz barrier can't be broken at an economical price point. In fact, server processors are clocked at a lower clock speed to reduce power consumption and run far more multiple cores than consumer counterpart. []

      • by jonwil ( 467024 )

        The only reason I want to upgrade this Core 2 Duo is because Visual Studio takes forever to compile/link large projects.

        Put an SSD and a new GPU in it a while back and its been going great on gaming ever since (hopefully it will be good enough to handle Fallout 4 when that hits)

        • by creimer ( 824291 )
          The last time I rebuilt my gaming PC was for Windows Vista in 2007. Since then I've added a SSD, changed out the CPU twice, and went through three different video cards. I'm planning to switch out the AMD 690 (DDR2 RAM) motherboard for a AMD 990 (DDR3 RAM) motherboard to get a few more years before I rebuild again.
  • by avandesande ( 143899 ) on Thursday October 15, 2015 @02:44PM (#50737681) Journal

    I have a I7 950 that is pretty old. It is still more than fast enough for my needs but would consider replacing it for something that uses less power. I would be curious to know how much it uses under the same testing conditions.

    BTW the i5 6500 avg 40 watts.... nice!

    • by slaker ( 53818 )

      That old i7 is in benchmark terms competitive with current i3 CPUs. I don't think of current i3s as slow and I don't think of five year old i7s as slow either. I upgraded to a 5960k last December because I actually do enough video encoding to keep it fed, but if I'm not stealing Blu-Rays there's no subjective difference from that monster to a the i7-980 it replaced.

      • Your stating that a dual core i3 processor out performs a quad core i7 processor? How fucking slow is the i7?
        • by FlyHelicopters ( 1540845 ) on Thursday October 15, 2015 @05:44PM (#50739203)

          It depends on what you're doing...

          I have a i7-920 on the test bench, it is still plenty fast for most anything you'd do with your average desktop computer, but the newer chips are indeed faster.

          For example, comparing the i7-920 against the i3-6320, you'll find the 2.66 GHz i7 actually slower in some tasks than the modern i3 chip.

          The i3 runs at 3.9 GHz, this is more than 45% faster, and it doesn't even including the IPC improvements across that many generations.

          Now, on some very specific tasks, the old i7 might be faster thanks to its triple channel memory and its 4 true cores and 8 threads.

          But those situations are very specific. The dual core, quad thread i3 is enough for a lot of things and even for those where it isn't, the faster clock speed combined with the higher IPC makes up a lot of the difference.

        • Pretty slow. []

      • The point was that I don't care about the performance at this point- just power usage. I think there are many people in the same boat as me that would upgrade their system just to save power.

  • If you're on haswell, don't bother. Unless you're doing it to increase the girth and length of your e-peen.
  • It's much faster in a few tests, a little faster in quite a lot more, slower in some and much slower in others. How is that "much faster"?

    • by gweihir ( 88907 )

      You must not have gotten Intel's payment from the "honesty in journalism" fund. After you have, of course everything Intel is much faster!

      • Christ alive! I got modded troll! Can you believe that?

        Well, due to your (entirely judtified, I might add) paranoia about intel, probably not.

        It's there in the graphs. Yeah, the AMD loses some badly, but so does the intel one.

        PS I'm enjoying my new FX9590s :)

  • by hairyfeet ( 841228 ) <> on Thursday October 15, 2015 @03:22PM (#50737991) Journal

    That hasn't been the sweet spot for the FX chips in like...well ever.

    Us that buy AMD chips are not buying the FX-8370, the bang for the buck just isn't there, if you want an octocore the sweet spot is the FX-8320, either the FX-8320 [] or the FX-8320E [] if you want to lower the power a tad, which is what I personally went for. The really hot spot right now is the FX-6300 [] which is a great gaming chip. A 3.5Ghz/4.1Ghz turbo chip with 6 cores for $99? Its a kick ass buy at that price.

    As for the 6500? Its a good chip, although I don't know why they are calling them "Linux friendly" when its Intel that have been coming up with all the nasty DRM chips going all the way back to HDCP and Palladium, but strictly based on the CPU? Its a good chip. I just have to wonder how many chips both Intel and AMD are gonna move when a 7 year old C2Q or Phenom II has no problem running pretty much every piece of software out there including games, both companies have built such badass chips for years that most users have piles of cycles left over.

    • That was my choice as well. The FX-8320E is a great CPU at a great price - Intel offerings still kill it in performance-per-core, but it is awesome for heavily multithreaded applications. You can build a system with a MSI board and 16GB RAM for under $350.

    • The Intel Core i5 6500 (no HyperThreading) appears to be faster than AMD's FX-6300 [] in single-process benchmarks by 40-100%, yet the multi-process benchmarks don't show as much of a pronounced difference: 10-40%. Granted the scores/graphs are all unmarked, which limits their usefulness in trying to make an informed decision, when you don't know the actual time-frame, nor if the FX-6300 was overclocked (which it can be, and the i5 can't). As you stated, TFA compares it to a similarly priced AMD processor, as
      • The FX-6300 was also a drop-in replacement for the Phenom II, on an AM3+ motherboard, which made it that much easier to justify.
      • Re: (Score:2, Troll)

        by hairyfeet ( 841228 )

        Just remember the cardinal rule which is if you don't know which compiler they used the test is worthless as Intel to this day uses the Cripple code in their compiler and pays benchmark companies to use it with predictable results. You let ME write the conditions on the compiler I can make a Sempron from 03 beat the latest i7, don't make the test worth a fuck.

        If you want to see how the chip REALLY stacks up without cripple code? Just look at these Linux benchmarks [] and surprise! Remove the cripple code and

      • Look at the link I provided, most benchmarks are compiled with the Intel cripple compiler which makes their results worthless. If its compiled with ICC it will scan for CPUID and if it doesn't return an Intel CPUID? It will send ALL math code through the X87 codepath, a path that hasn't been mainstream since 1995. No SSE, no AVX, none of the modern math processing is touched. How big of a difference does that make? Well if you take a Via CPU (the only one that allows softmodding of CPUID) and change its CPU

        • Depends on program actually and what version of ICC. What you're describing was old ICC behavior before they were sued. My own tests compiling Povray [] with ICC 13 I've found that the ICC dispatch code does work properly on my FX 8350. In other words, compile with AVX dispatch support and the program will choose the AVX code path, which gives the same performance as compiling with only AVX support (the dispatched program also runs on my PhenomII 1090T, but the AVX exclusive program doesn't). But that doesn't

    • It's Linux friendly because the GPU drivers are open source and don't suck, unlike the AMD GPU drivers. Although if you don't do any 3D related work, any of the open source drivers for all the GPUs are good enough.

  • by future assassin ( 639396 ) on Thursday October 15, 2015 @03:39PM (#50738141) Homepage

    as this price will be brutal if Intel gets the market to itself. This could be Intel trying to get a lower price going for when Zen comes out, if they get a proper price point they could hold off Zen in price vs power.

    • The latest rumour puts Zen release date in Q4 2016.
      Between then and now there are no new CPU products planned, so they have to survive on Kavari / Carrizo rebadges til then.

  • by rsilvergun ( 571051 ) on Thursday October 15, 2015 @04:03PM (#50738347)
    Not trolling. I just have a really hard time justifying the purchase. I'm running an old Athlon 6000 X2 and it rocks, but my friends with 8350s either had stability problems ( no overclocking, Asus or Gigabyte board) or just plain couldn't run their games. What I really want to know is why aren't AMD processor prices in free fall? Who's buying them that keeps the price near an i5?
    • Weird. I've put together several boxes using the FX-8320E and 8350 and had zero stability issues.

      Pricing is a concern though. The 8320 line used to be an excellent option in the bang for the buck department, but these new Intel offerings really close the gap.

    • by gweihir ( 88907 )

      I have both a gaming-box and a long-running Linux server on the FX8350. Absolutely no problems.

  • from here on out I will not be buying products from companies that support TPP, which pretty much subjugates entire nations to the will of corporations. Intel, Apple, Microsoft, Facebook, they'll support this awful radical anti-democratic treaty, which has so little to do with trade.
  • I have a third gen i5-3570k. I only use it for gaming (fps, open world third person games mostly). Should i upgrade to this 199$? Another one? Or wait? I have a nvidia 970 gtx.

UNIX is hot. It's more than hot. It's steaming. It's quicksilver lightning with a laserbeam kicker. -- Michael Jay Tucker