ARM unveils power-optimized 2GHz Cortex-A15 quad-core chip

[Leeky]

ARM Holdings, the UK-based company whose chip designs power the majority of the smartphone and tablet products on the market, has announced on their official blog a new quad-core chip aimed directly at notebooks and network-based hardware.

The new ARM Cortex-A15 MPCore chip is a hard macro variant quad-core clocked at up to 2.0GHz, built using the TSMC 28-nanometer HPM process and offers similar power consumption ratings to their existing Cortex-A9 products. It also includes their NEON and Floating Point Unit technology, EEC for L1 and L2 RAM, as well as the ability to address more than 4GB of RAM and support for virtualization, which makes it suitable for a wide variety of applications.

What's unusual is that ARM has uniquely released it as a hard macro version, which the company says offers the best compromise between performance and power drain as well as allowing for it to hit the market much faster. Normally their chips can be tweaked to the customer's exact specifications, but ARM maintains this will ultimately increase the time it takes to reach store shelves.

"For SoC designers looking to make a trade-off between the flexibility offered by the traditional RTL-based SoC development strategy and a rapid time to market, with ensured, benchmarked power, performance and area, an ARM hard macro implementation is an ideal, cost-effective solution," said Jim Nicholas, ARM's vice president of processor division marketing. "This new Cortex-A15 hard macro is an important addition to our portfolio and will enable a wider array of partners to leverage the outstanding capabilities of the Cortex-A15 processor."

ARM believes the new chip will be ideal for powering notebooks and even more interestingly, thin but fast network-based hardware solutions. The set hardware parameters could potentially decrease the time it takes to see them arrive in new products, which could include smartphones as well as tablets.

There is no exact timeframe on availability yet, but previous announcements regarding ARM Cortex-A15 hardware suggest the new quad-core chip should reach products shipping later this year, or early 2013. More information is expected to follow at the IEEE symposium in Japan later today.

Permalink to story.

https://www.techspot.com/news/48248-arm-unveils-power-optimized-2ghz-cortex-a15-quad-core-chip.html

 

[Guest]

I want it in my server, laptop, and tablet now, where battery time and cost of continious use
are important. My PC can be my power hungry performance option.

 

[EEatGDL]

I don't think you pretty understand about architectures, you can't simply put an ARM processor in a laptop because it has similar specs to an x86 processor. ARM differs in design, set of instructions, and other things that simply makes it totally uncompatible with Windows OS for example different to Windows 8 RT. Is like if you wanted a PC with microcontollers just because you can use them with normal batteries due their extremely low energy consumption, even if they are for example 16-bit microcontrollers.

But maybe you can try laptops with ultra-low voltage processors (~8 W) anyway.

 

[Guest]

EEatGDL, voltages are typically given in volts, not watts.

 

[Zeromus]

That was the most horrible explanation about the quirks of CPU architectures.

 

[Zeromus]

I don't think you pretty understand about architectures, you can't simply put an ARM processor in a laptop because it has similar specs to an x86 processor. ARM differs in design, set of instructions, and other things that simply makes it totally uncompatible with Windows OS for example different to Windows 8 RT. Is like if you wanted a PC with microcontollers just because you can use them with normal batteries due their extremely low energy consumption, even if they are for example 16-bit microcontrollers.

But maybe you can try laptops with ultra-low voltage processors (~8 W) anyway.

This is the most horrible explanation I mean, not the article. Doesn't the upcoming exynos implement A15?

 

[VitalyT]

Power-optimized, and not a word on the consumption - what kind of article is that?

 

[SNGX1275]

I don't think you pretty understand about architectures, you can't simply put an ARM processor in a laptop because it has similar specs to an x86 processor. ARM differs in design, set of instructions, and other things that simply makes it totally uncompatible with Windows OS for example different to Windows 8 RT.

But ARM performance is getting to be pretty good. Can't be a whole lot different than the Via C6 processors that a few of us ran XP on. Win 8 is pretty lightweight and has an ARM version, but unfortunately for us normal people Microsoft doesn't appear to be releasing a standalone ARM version.

 

[EEatGDL]

EEatGDL, voltages are typically given in volts, not watts.

Hehe, I mentioned the watts for the consumption point mentioned by the first guest; not an example of the voltage they work with, I think it was a bad place to write it though [just next to voltage], I agree.

And for the others I wasn't really explaining anything, I didn't talk about Carpinelli, Harvard architecture, CISC, etc. neither about assembly lenguage differences between architectures, etc. etc. etc. RISC (ARM) is a totally different design philosophy from CISC (x86), that's why in the transition from the Power PC to Intel, Apple was presenting a binary translator from one architecture to another [you can find it in a keynote uploaded in Youtube were they -Jobs- show the binary translator using Wolfram Mathematica]. I know the C51 assembly language and I'm sure it has a lot of different instructions, maybe some similar with the x86 architecture, but a lot less and different without knowing a single x86 instruction for certain, I can ask my uncle that works on intel about some differences between maybe C51-MMX or C51-SSE2 to illustrate you more.

I'm half way my career, but I can stand pretty much for what I said. I never said ARM performance is bad, I simply said instructions targeting x86 won't work on an ARM processor. All of that derived from the first commentary about power consumption-architecture-performance; nothing more, nothing less.

 

[EEatGDL]

And even from Power PC to Intel, both were CISC, now imagine between RISC to CISC, or the other way.

 

[SNGX1275]

PowerPC was RISC. Now the 68k processors, which Apple used before PowerPC were CISC. So they went CISC -> RISC -> CISC all while keeping the OS/Program compatibility. System 7 might have worked on PowerPC, I forget, but I know OS 8 did, and if you had OS 8 and an app for 68k it ran. Then with OS X 10.4 and 10.5 it ran on Intel and PowerPC chips, again with program compatibility between them (with specific exceptions like new devs writing only for Intel).

 

[EEatGDL]

Thanks SNGX for the acclaration, I thought PowerPC was just a comercial name for the Motorola 68K, because I knew Motorola was behind PowerPC, but not that the PowerPC was actually a totally different approach. I just happily ended the semester and finally have spare time.

 

[Guest]

I don't think you pretty understand about architectures, you can't simply put an ARM processor in a laptop because it has similar specs to an x86 processor. ARM differs in design, set of instructions, and other things that simply makes it totally uncompatible with Windows OS for example different to Windows 8 RT. Is like if you wanted a PC with microcontollers just because you can use them with normal batteries due their extremely low energy consumption, even if they are for example 16-bit microcontrollers.
But maybe you can try laptops with ultra-low voltage processors (~8 W) anyway.

Use a Linux desktop like Ubuntu or Fedora. These full desktop OSes will run perfectly on this chip.
I run Ubuntu on a Tegra 2 laptop, which is only limited by its memory (512MB). With 4GB of memory, this chip is very suitable for laptops.