Announcement

Collapse
No announcement yet.

Trying The Turbo Boost Max 3.0 Patches On Linux 4.9 With A Core i7 6800K Broadwell-E

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #11
    Originally posted by schmidtbag View Post
    AMD is irrelevant. Intel needs to keep their sales up and it's pretty hard to convince people to buy a new product when you have maybe a 5% performance increase over your last generation.
    The thing is, they don't do this for the desktop, but for laptops. When you're power constrained, it becomes that much more important to squeeze every drop of performance you can. And I'll tell you, my 15W CPU is as snappy as they come.
    But on the desktop, you're spot on. I went from a 2500k to a 6600k and the best gains I've seen are better I/O (e.g. USB 3.1 and M.2 ports).

    Comment


    • #12
      Originally posted by bug77 View Post
      The thing is, they don't do this for the desktop, but for laptops. When you're power constrained, it becomes that much more important to squeeze every drop of performance you can. And I'll tell you, my 15W CPU is as snappy as they come.
      But remember - if you're using TB, your CPU isn't actually 15W, which was part of my point.

      The purpose of a laptop is portability. In order for a PC to be portable, you need to be able to comfortably carry it and it should have a usable battery. When you have a CPU that exceeds the advertised wattage, you either need a bigger heatsink and/or a better battery, both of which makes the laptop bigger, heavier, and more expensive. If you don't want that, then the CPU either can't effectively be turboed (in which case you aren't squeezing in extra performance), or, you will lose battery life quicker (which lowers the practicality of having a laptop).

      This is why the original iteration of Turbo Boost was great - you didn't have to make any sacrifices or compromises. You got the performance where you needed it and you got what you paid for.


      Anyway, if you need to be doing some serious number crunching, you shouldn't be doing it on a laptop. The i3 is, in my opinion, the best laptop CPU you can get. Its performance is sufficient and predictable for everyday needs (which in turn makes battery life more predictable). The features it lacks that the i5 or i7 have are often unobtainable in a laptop, or at least impractical. Laptop i3s are comparable to desktop i3s in terms of performance and features; many i5 and i7 models pale in comparison to their desktop counterparts. Mobile i3s are significantly cheaper than i5s or i7s while not being that much slower. The GPU, admittedly, is a bit weak.
      Obviously there are exceptions, but for the average person, an i3 is all they need.

      Comment


      • #13
        Originally posted by schmidtbag View Post
        But remember - if you're using TB, your CPU isn't actually 15W, which was part of my point.
        If you want to take thing literally, the CPU isn't rated at 15W most of the time, because it powers down. Who cares if it bursts above the limit when it is safe to do so? That 15W is what you should expect it to need in the long run.

        Comment


        • #14
          Michael, do you see /proc/sys/kernel/sched_itmt_enabled = 1? If you don't see this entry in the sysctl table, then the firmware of your machine may not support exposing the ITMT feature via the ACPI CPPC objects so ITMT is not actually enabled. If you can send a dump of the dmesg to Srinivas or myself, we can take a closer look. Our emails are in the ITMT patches we've posted.

          Comment


          • #15
            Originally posted by schmidtbag View Post
            AMD is irrelevant. Intel needs to keep their sales up and it's pretty hard to convince people to buy a new product when you have maybe a 5% performance increase over your last generation.
            That's obvious nonsense. The GPU portion of AMD APU's blows Intel's out of the water and for gamers that need a laptop, AMD is is obviously the better choice for the same amount dollars.

            I just wish OEM's would ship AMD laptops with batteries that have more cells. Basically, if you buy an AMD laptop, you need to buy a third party battery. That part sucks ass and is pretty stupid on the part of OEM's.

            Comment


            • #16
              EDIT: It's obvious why it happens, Intel clearly paid OEM's to ship AMD laptops with smaller baterries in an attempt to make it look like AMD gets lesser battery life. But the real truuth is they simply get stuck with smaller batteries.

              Comment


              • #17
                Originally posted by bug77 View Post
                If you want to take thing literally, the CPU isn't rated at 15W most of the time, because it powers down. Who cares if it bursts above the limit when it is safe to do so? That 15W is what you should expect it to need in the long run.
                By that logic, why aren't turbo boost speeds higher? Why not let CPUs go as high as reasonably possible (obviously while retaining reliability and stability) and variably decrease clock speeds so it always operates in a "safe" range? Even without TB, the CPU still needs to throttle performance if it gets too hot.

                It's all completely arbitrary. Getting a 3.4GHz CPU that can turbo to 3.7 is nothing more than a marketing gimmick. The CPU can effortlessly go faster (assuming ideal environmental conditions) but if that were the case there would be no reason to have different grades of performance (per class) and therefore Intel would have a harder time making money.

                Comment


                • #18
                  Originally posted by duby229 View Post
                  That's obvious nonsense. The GPU portion of AMD APU's blows Intel's out of the water and for gamers that need a laptop, AMD is is obviously the better choice for the same amount dollars.
                  I had a feeling my comment was going to trigger someone - I realized my poor phrasing after I already posted it. I meant AMD is irrelevant to this particular discussion. I am an AMD fan myself, but in my perspective AMD's level of competition has very little to do with Intel's design with TB.

                  Comment


                  • #19
                    Originally posted by schmidtbag View Post
                    By that logic, why aren't turbo boost speeds higher? Why not let CPUs go as high as reasonably possible (obviously while retaining reliability and stability) and variably decrease clock speeds so it always operates in a "safe" range? Even without TB, the CPU still needs to throttle performance if it gets too hot.
                    Probably because they aren't (tested to be) safe?

                    Originally posted by schmidtbag View Post
                    It's all completely arbitrary. Getting a 3.4GHz CPU that can turbo to 3.7 is nothing more than a marketing gimmick. The CPU can effortlessly go faster (assuming ideal environmental conditions) but if that were the case there would be no reason to have different grades of performance (per class) and therefore Intel would have a harder time making money.
                    It's not a gimmick, my 6600k can actually sustain those turbo speeds. If I'm not using all cores at once. It's actually super-handy if you want to compile something that takes 15 minutes with a compiler that doesn't know how to multi-thread. Yeah, I know, most users don't do that, for for those that do, it's a valuable feature even if you think otherwise (and if may very well look gimmicky to you, that's also a legit pov).

                    Comment


                    • #20
                      Originally posted by bug77 View Post
                      Probably because they aren't (tested to be) safe?
                      Intel is well aware of the capabilities of their products. When you consider they have some CPUs with turbo speeds above 4GHz, you know there's plenty of room to spare on cheaper models.
                      It's not a gimmick, my 6600k can actually sustain those turbo speeds. If I'm not using all cores at once. It's actually super-handy if you want to compile something that takes 15 minutes with a compiler that doesn't know how to multi-thread. Yeah, I know, most users don't do that, for for those that do, it's a valuable feature even if you think otherwise (and if may very well look gimmicky to you, that's also a legit pov).
                      And good for you (not being sarcastic) - but I can assure you that not everyone can sustain those turbo speeds on the stock heatsink.

                      To clarify, I have no problem whatsoever of a CPU being pushed hard. My gripe is the reason behind TB 2.0. There is no reason why your 6600k couldn't just be called a 3.9GHz part and just slow itself down when it gets too hot. But if that were to happen, Intel would get sued or be a mockery, because the product isn't running at it's advertised speed. By using TB as an excuse, Intel gets to ship a product with a lower normal clock speed, advertise a lower TDP, and appears to outperform it's predecessors (based on that speed), while giving you a cheaper heatsink. And then people act like it's something special when they get to achieve the turbo speeds, which they should have always readily had available to them from the beginning.

                      In other words, you're willingly paying for a product where it's completely ok if it performs below it's actual capabilities, whereas if it's default speed was 3.9GHz with the same heatsink and same price point, it would be considered totally unacceptable if it started to slow down due to heat. The end result is exactly the same, but one looks like a good thing - that's how it's a gimmick. Again, that's great you don't have that issue, but you're not everyone.

                      Comment

                      Working...
                      X