Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Tuesday May 30 2017, @05:28AM   Printer-friendly
from the More-Moore? dept.

ARM has announced two new CPU cores, the Cortex-A75 and Cortex-A55. According to ARM, the A75 increases performance by around 22% over the A73 at the same level of power consumption. It can also scale to use more power per core (1-2 W rather than 0.75 W) which could slightly improve the performance of ARM laptops and tablets.

The smaller core, the Cortex-A55, increases performance by around 18% compared to the Cortex-A53, but also increases power consumption by 3%. Thus, power efficiency is about 14-15% better than the A53.

ARM's successor to big.LITTLE, DynamIQ, allows for up to 8 cores of any size (which for now means either the A75 or A55) inside of a single cluster. This means that a configuration including 1x Cortex-A75 and 7x Cortex-A55 cores would be possible, or even optimal according to ARM.

ARM also announced its Mali-G72 GPU, an incremental upgrade to the Mali-G71:

ARM says that the Mali-G72 will see a 25 percent boost to energy efficiency compared with the G71, meaning that SoC designers will have more power to play with to boost performance or increase battery life.

Similarly, the G72 offers 20 percent better performance density, meaning that manufacturers can pack more GPU cores into the same die area as before, giving further potential for a performance boost without an increase in cost. Previously ARM was targeting 16 to 20 Mali-G71 cores as the optimum for mobile, and expects to see the number push closer to the 32 shader core maximum supported by the G72 this time around.


Original Submission

Related Stories

ARM's DynamIQ Introduces Variable Core-Configuration Clusters 20 comments

ARM will replace the big.LITTLE cluster design with a new one that allows up to 8 CPU cores per cluster, different types of cores within a cluster, and anywhere from one to many (unlimited?) clusters:

The first stage of DynamIQ is a larger cluster paradigm - which means up to eight cores per cluster. But in a twist, there can be a variable core design within a cluster. Those eight cores could be different cores entirely, from different ARM Cortex-A families in different configurations.

Many questions come up here, such as how the cache hierarchy will allow threads to migrate between cores within a cluster (perhaps similar to how threads migrate between clusters on big.Little today), even when cores have different cache arrangements. ARM did not yet go into that level of detail, however we were told that more information will be provided in the coming months.

Each variable core-configuration cluster will be a part of a new fabric, with uses additional power saving modes and aims to provide much lower latency. The underlying design also allows each core to be controlled independently for voltage and frequency, as well as sleep states. Based on the slide diagrams, various other IP blocks, such as accelerators, should be able to be plugged into this fabric and benefit from that low latency. ARM quoted elements such as safety critical automotive decisions can benefit from this.

A tri-cluster smartphone design using 2 high-end cores, 2 mid-level cores, and 4 low-power cores could be replaced by one that uses all three types of core in the same single cluster. The advantage of that approach remains to be seen.

More about ARM big.LITTLE.


Original Submission

Alibaba Claims its RISC-V Core is Faster than ARM Cortex-A73 8 comments

Alibaba Reports Their XT910 RISC-V Core To Be Faster Than An Arm Cortex-A73

A few weeks back Alibaba announced the "XT910" as the fastest RISC-V processor featuring 16 cores and clock speeds up to 2.5GHz while being manufactured on a 12nm node. This by far beats most RISC-V hardware currently available and now at this week's Hot Chips conference the Chinese company is reporting that the XT910 is faster than an Arm Cortex-A73.

Alibaba confirmed the XT910 as a TSMC 12nm FinFET design with clock speeds between 2.0GHz and 2.5GHz for this RISC-V 64-bit processor supporting the RISC-V 0.7.1 Vector Extension.

Benchmarks posted by Alibaba's T-Head organization put the XT910 faster than an Arm Cortex-A73 found within the Kirin 970 SoC in areas across automotive, consumer, networking, and telecom spaces.

Cortex-A73 (launched in 2016) reportedly has less floating point IPC than its Cortex-A72 predecessor (found in the Raspberry Pi 4B). ARM's latest performance-oriented cores are the Cortex-A78 and Cortex-X1.

Previously: Alibaba Announces a 16-Core RISC-V CPU

Related: ARM Cortex A73 Unveiled
ARM Cortex-A75, Cortex-A55, and Mali-G72 Announced
ARM Announces Cortex-A78 and Cortex-X1


Original Submission

Arm Announces New Armv9 Cores: The Cortex-X2, Cortex-A710, and Cortex-A510 4 comments

Arm Announces Mobile Armv9 CPU Microarchitectures: Cortex-X2, Cortex-A710 & Cortex-A510

It's that time of the year again, and after last month's unveiling of Arm's newest infrastructure Neoverse V1 and Neoverse N2 CPU IPs, it's now time to cover the client and mobile side of things. This year, things Arm is shaking things up quite a bit more than usual as we're seeing new[sic] three new generation microarchitectures for mobile and client: The flagship Cortex-X2 core, a new A78 successor in the form of the Cortex-A710, and for the first time in years, a brand-new little core with the new Cortex-A510. The three new CPUs form a new trio of Armv9 compatible designs that aim to mark a larger architectural/ISA shift that comes very seldomly in the industry.

Alongside the new CPU cores, we're also seeing a new L3 and cluster design with the DSU-110, and Arm is also making a big upgrade in its interconnect IP with the new cache coherent CI-700 mesh network and NI-700 network-on-chip IPs.

The Cortex-X2, A710 and A510 follow up on last year's X1, A78 and A55. For the new Cortex-X2 and A710 in particular, these are direct microarchitectural successors to their predecessors. These parts, while iterating on generational improvements in IPC and efficiency, also incorporate brand-new architectural features in the form of Armv9 and new extensions such as SVE2.

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 1, Interesting) by Anonymous Coward on Tuesday May 30 2017, @06:17AM (8 children)

    by Anonymous Coward on Tuesday May 30 2017, @06:17AM (#517494)

    Are these designs tied to a particular fab tech? How can they link to specific power usage numbers otherwise?

    • (Score: 2) by c0lo on Tuesday May 30 2017, @06:40AM (5 children)

      by c0lo (156) Subscriber Badge on Tuesday May 30 2017, @06:40AM (#517498) Journal

      How can they link to specific power usage numbers otherwise?

      You've heard of engineering?
      You know, since hundred years back, the guys have been able to predict the durability of a bridge without building it, especially when they are able to use CAD/simulation software.
      Granted, sometimes they predicted wrong, especially on high winds [wikipedia.org]; but most of the time they managed to come quite close.

      Quite crazy these guys, I have to tell yeah... the good ones can't be swayed by markedroids and MBA-s (and may need to be fired). Here's even an example of predicted failure [wikipedia.org].

      --
      https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
      • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @06:53AM (2 children)

        by Anonymous Coward on Tuesday May 30 2017, @06:53AM (#517505)

        Anyone else who's a not drunk idiot?

        • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @06:58AM

          by Anonymous Coward on Tuesday May 30 2017, @06:58AM (#517506)

          I is sober as a fox and I only have one question, are these ARM all pimped out with them NEON lights?

        • (Score: 1) by AssCork on Tuesday May 30 2017, @01:22PM

          by AssCork (6255) on Tuesday May 30 2017, @01:22PM (#517609) Journal

          He's not drunk! (He's just speaking in cursive)

          --
          Just popped-out of a tight spot. Came out mostly clean, too.
      • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @08:55AM (1 child)

        by Anonymous Coward on Tuesday May 30 2017, @08:55AM (#517545)
        • (Score: 2) by c0lo on Tuesday May 30 2017, @09:06PM

          by c0lo (156) Subscriber Badge on Tuesday May 30 2017, @09:06PM (#517884) Journal

          Tacoma bridge is busted. Clearly it was a engineering/design defect.

          --
          https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
    • (Score: 2) by takyon on Tuesday May 30 2017, @08:55AM

      by takyon (881) <reversethis-{gro ... s} {ta} {noykat}> on Tuesday May 30 2017, @08:55AM (#517546) Journal

      The power usage numbers are definitely subject to change. They are from ARM after all.

      Some cores are made identical to ARM's designs, others are customized:

      https://en.wikipedia.org/wiki/Comparison_of_ARMv8-A_cores [wikipedia.org]

      For example, Qualcomm Snapdragon 835 uses customized Cortex-A73 cores. Fab = Samsung’s 10nm "10LPE" FinFET process. The latest Exynos SoCs from Samsung use a combination of custom cores and what looks like normal Cortex-A53 cores.

      Here's a fun one: Apple’s A9 SoC Is Dual Sourced From Samsung & TSMC [anandtech.com]. Meaning you could order your iPhone 6s and get a chip with slightly different performance and efficiency. Luck of the draw.

      There's also GlobalFoundries, and Intel may even make ARM chips for third parties [arstechnica.com] in the future.

      --
      [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
    • (Score: 2) by TheRaven on Tuesday May 30 2017, @09:36AM

      by TheRaven (270) on Tuesday May 30 2017, @09:36AM (#517560) Journal
      Not sure why this is off-topic. The ARM data sheets always list the process that they've used for specific power numbers. These numbers are, according to TFA, from the same process that they quote power numbers for for the A53, though I can't find that in 10 seconds of looking.
      --
      sudo mod me up
  • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @08:47AM (1 child)

    by Anonymous Coward on Tuesday May 30 2017, @08:47AM (#517542)

    Moore. Gordon Moore.

    • (Score: 3, Funny) by bob_super on Tuesday May 30 2017, @05:38PM

      by bob_super (1357) on Tuesday May 30 2017, @05:38PM (#517787)

      "I expect you to shrink, mister Moore"

  • (Score: 1, Informative) by Anonymous Coward on Tuesday May 30 2017, @11:46AM (4 children)

    by Anonymous Coward on Tuesday May 30 2017, @11:46AM (#517592)

    I mean, at the driver level. ARM is not the way forward for FOSS computing.

    • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @01:41PM

      by Anonymous Coward on Tuesday May 30 2017, @01:41PM (#517622)

      screw arm and mali. we need risk or similar.

    • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @08:15PM (2 children)

      by Anonymous Coward on Tuesday May 30 2017, @08:15PM (#517870)

      Enjoy the reading http://libv.livejournal.com/27461.html [livejournal.com] ARM is against FOSS even if it would mean simpler support and more sales. And when reverse engineering was tried, many sticks were pushed into wheels by others in the comunity. We could be using LIMA and TAMIL for some years.

      Radeon FOSS driver start is also full of sticks. http://libv.livejournal.com/27799.html [livejournal.com]

      Companies love to spoil what they get.

      • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @09:08PM (1 child)

        by Anonymous Coward on Tuesday May 30 2017, @09:08PM (#517886)

        > ARM is against FOSS even if it would mean simpler support and more sales.

        You base this on one single blog post that actually has a lot of bias.
        Though at least that blog post reference the video of an interview that explains things a bit differently.
        Anyway as stated it's wrong, ARM certainly contributes back to several FOSS projects, even more so if you include contributions by Linaro.
        Admittedly you can probably consider them all driven by pure self-interest, but still it proves your statement as-is certainly as untrue.
        Also the kernel part of the driver to my knowledge is OpenSource (GPLv2), which would make ARM GPUs more OpenSource from the vendor side than NVidia desktop GPUs for example...

        • (Score: 0) by Anonymous Coward on Tuesday May 30 2017, @09:54PM

          by Anonymous Coward on Tuesday May 30 2017, @09:54PM (#517910)

          What driver to use Mali GPUs? One that links a binary blob you get by means of the commercial SDK? Pretty much the same trick NVidia uses with the kernel and keeps the rest closed. The most they seem to provide is open 2D. Like if this was the 90s.

          If they were barely friendly, they would not declare they don't want a FOSS full driver for the Mali and newer. Friendly would mean sending docs or hints when issues appear. They must be making money hand over fist with the SDK to be so negative about other efforts.

(1)