x86 with an ARM core

x86 with an ARM core

This seems interesting, would this processor have the potential to run RISC OS?

If so then a RISC OS / Windows / Linux multiboot device could be possible. (If we ever get partition support)

Call me stupid – but what exactly is the point of this? RISC OS is extremely tied to the ARM for obvious reasons. Windows on ARM devices is….perhaps only Windows by name. And Linux exists in builds for either processor.

I am wondering if the only real reason we still suffer the abomination that is x86 is because of the very widespread deployment of Windows? This isn’t to say that ARM is better, there’s a damn fast and quite capable core instead most modern x86 silicon. How about we see that without the baggage that is the x86 instruction set? [though I would imagine it is highly “quirky” due to what it does…shame]

But – serious question – what’s the purpose of this dual-architecture chip?

It’s not a x86+ARM processor. It’s a x86 processor, with a cryptoengine, and this cryptoengine is an ARM core. That’s a bit different :)

The article mentions virtualized Android support for Windows systems. (Providing it isn’t a different instruction set to what we need) this should mean that Virtual Risc should be able to be modified to use the native core, meaning it really would be a virtual machine, rather than an emulator, with consequent efficiency gains.

But the question is whether it would be designed with enough flexibility for the ARM core to act as a main processor, while the x86 is shutdown, or at least on a very low power mode.

Now I fail to see a reason for having an ARM core which can only act as a co-processor.

In the mainstream world, I have yet to see a desktop GUI OS that runs anywhere near satisfactorily on an ARM system.

Raspbian really was very poor, and was scarcely much better than Debian on an old P2 from the last century, and certainly nowhere near as flexible.

I would think the aim of such a chip would be to allow a single device to run an OS such as Android on batteries on a small screen and a full OS using the x86 when docked to a big screen, while retaining the mobile Apps as widgets.

But it would be nice to have a single device that ran RISC OS or Windows or Linux or Android, as you liked. (RISC OS on Arm, Win/Lin on x86)

The article mentions virtualized Android support for Windows systems. (Providing it isn’t a different instruction set to what we need) this should mean that Virtual Risc should be able to be modified to use the native core, meaning it really would be a virtual machine, rather than an emulator, with consequent efficiency gains.

No. Bluestack is a x86 Android device emulator. The Register love to put totally different things in the same sentence and let people understand something that is not true. The Register is an IT tabloid.

Example:

AMD has two new, low-power system-on-chip (SoC) designs for laptops and tablets ready to roll that include an ARM processor core built into the die to handle security and virtualized Android support for Windows systems.

This is not an ARM processor core “to handle security AND virtualized Android support” but “1/ an ARM processor core built into the die to handle security” and “2/ virtualized Android support for Windows systems”.

Now I fail to see a reason for having an ARM core which can only act as a co-processor.

Because TrustZone is an extension for ARM cores. In fact this coprocessor is made just to be a cryptoengine, as some earlier ARM were made just to be sound chips (for Creative, for example).