Announcement

Collapse
No announcement yet.

Intel Broadwell CPUs Should Be Out In Time For Christmas

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

  • Intel Broadwell CPUs Should Be Out In Time For Christmas

    Phoronix: Intel Broadwell CPUs Should Be Out In Time For Christmas

    We're anxious to test out Intel's next-generation Broadwell processors but it doesn't look like they'll be out in time for "back to school" shopping but should certainly arrive in time for the holidays...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    I think the mobile chips will come out in time for X-Mas, but I still have strong doubts desktop chips will make it out the door in 2014. Honestly, since Intel has no competition from AMD in mid to upper desktop segments and is doing just fine in the lower desktop segments, there's really no reason for them to rush Broadwell desktop chips to the market. On the mobile front, the die shrink should further help with power consumption as Intel tries to muscle in on the ARM-dominated space. Intel needs to get these out the door as quickly as possible, so I figure all their efforts will be concentrated on getting the mobile chips out the door first.

    Comment


    • #3
      Desktop is scheduled for mid-2015. This year only Broadwell-Y and maybe Broadwell-U (2+2).

      Comment


      • #4
        I'm holding out for 2016. Skylake should bring HEVC/VP9 and all sorts of other new features that will be useful along with new DDR and PCIe revisions.

        Comment


        • #5
          I remember when this was predicted like a year ago or more, and many didn't want to admit that Intel will fall 6 months behind schedule. This is shrinking the gap between Intel and other chip makers in terms of process technology a lot. When the others start having FinFET, the gap will be really small (less than half a node).

          While some like Anandtech may like to claim that the "x86 myth was busted", the truth is Intel Atom can barely hold out on its own with 22nm vs 28nm, and Trigate, which gives the advantage of a full node shrink basically, against ARM CPUs. Intel has a node and a half advantage and it barely competes in CPU, while in GPU, it's usually 2x or more behind in performance. When the process technology difference will be all but wiped out, it will be more obvious that x86 is indeed more bloated.

          As for PC/laptop users, I don't even know why you're getting so excited. Haswell barely brought any improvements in performance overall, and Broadwell will bring even fewer. Intel is focusing 90 percent on power consumption, not on increasing overall performance (except for GPUs a bit perhaps). We'll probably start seeing Core i5 1.1 Ghz CPUs in laptops next year, and everyone will wrongly assume that those chips are "faster" than the current Haswell ones, or than the previous IVB ones.

          Comment


          • #6
            Originally posted by Krysto View Post
            I remember when this was predicted like a year ago or more, and many didn't want to admit that Intel will fall 6 months behind schedule. This is shrinking the gap between Intel and other chip makers in terms of process technology a lot. When the others start having FinFET, the gap will be really small (less than half a node).

            The gap is increasing not decreasing. No 20nm GPUs from AMD and Nvidia this year. No mobile 20nm SoC from TSMC officially announced yet. We are looking for a ~3 years timeframe from 28nm to 20nm for TSMC. Intel will launch Broadwell in September-October which is a ~2.5 years timeframe between 22nm and 14nm. Since years people keep saying the gap is shrinking, till now it's the opposite. Every node the same talk and every time Intel and TSMC proved them wrong.


            Originally posted by Krysto View Post
            While some like Anandtech may like to claim that the "x86 myth was busted"
            It is busted of course.

            Comment

            Working...
            X