ASIO Goes Native!

At the Snapdragon conference, Steinberg demonstrated Cubase running on Windows for Arm, using the native ASIO drivers for a low‑latency electric guitar performance.

Windows‑based musicians have long been asking for better support at the operating system level — and it seems to be very much on its way, with native low‑latency ASIO drivers and full support for MIDI 2.0. We reached out to Microsoft to find out more.

At the Qualcomm Snapdragon conference in Hawaii in October, Microsoft surprised us all by announcing a new version of Windows for Arm processors with built‑in ASIO drivers for USB 2 audio interfaces. I say surprise, because it marks a big shift from the way Windows has handled audio in the past. The flexibility of Apple’s Core Audio has always been an attractive feature, not least due to its native latency levels, and it’s a feature that lots of Windows‑based musicians have been requesting for many years. At present, this new feature is exclusive to this version of Windows for Arm processors but, importantly, Microsoft say they plan soon to embed USB ASIO support in the more familiar desktop versions of their OS — so it’s potentially very big and promising news for the music‑making community.

The new driver was developed in partnership with Yamaha, the parent company of Steinberg (who developed the ASIO driver protocol) as well as a manufacturer of various USB audio devices and musical instruments, and at the conference Steinberg demonstrated Cubase running on a Snapdragon laptop. They used the native USB ASIO drivers with one of their interfaces to process a live guitar signal with great low‑latency performance. Seeing this sort of performance (it’s on YouTube if you’re interested) on a native hardware driver in a Windows environment was certainly refreshing.

Several notable developments then followed. Shortly after the presentation, a supporting blog post appeared from Microsoft Principal Software Engineer Pete Brown, in which he discussed the announcements and expanded on the new audio developments, which included not only the native ASIO drivers, but also support for MIDI 2.0. Notably, interface manufacturers Focusrite announced their backing for native Arm drivers. Then, a preview build of Cubase for Arm was released a few days after the event, CockOS launched a preview build of Reaper, and Reason announced that a preview of their DAW would be forthcoming.

A number of companies announced that they’d be developing their popular software and hardware audio products for Arm‑based Windows platforms — and more are sure to follow suit.

But what drove this development? Why did Microsoft focus first on Arm processors, and what might it all eventually mean for those using Windows for music and audio? To help me answer such questions, I caught up with Pete Brown, and our interview forms the second half of this article. First, though, I thought it would be helpful to offer some background on the Arm chips supported by this new OS, and to explain why we might be about to see more of them appearing in ‘serious’ music‑making Windows machines.

From Little Acorns...

The Acorn Archimedes: the first widely available commercial computer to feature an ARM processor.Most of us probably know that the M‑series CPUs now used throughout Apple’s product range are based on the chip technology of Arm Holdings PLC, usually known simply as Arm. Apple’s new processors have an enviable reputation for both power and efficiency, but the Arm name (ARM until a 2017 rebranding) and their processors actually go back several decades. The underlying ‘silicon’ concept dates back to the 1960s, but the acronym ARM was first used in 1983 by Acorn Computers, when it stood for Acorn RISC Machine. Conceived to handle simplified code that can be ‘pipelined’ in a relatively efficient manner, RISC (Reduced Instruction Set Computer) chips are a competing technology to that used in the Intel and AMD CPUs (known as CISC, or Complex Instruction Set Computer chips) that we’ve become accustomed to using in laptop and desktop computers.

Acorn’s first RISC chip appeared in 1987 in their Archimedes desktop computer, which some readers might remember as a more powerful replacement for the BBC Micro in UK schools. But in 1990, Acorn, Apple and VLSI Technology (now NXP Semiconductors) teamed up in a joint venture called Advanced RISC Machines (ARM), which was later listed on the stock exchange and became the Arm we know today. Along with VLSI, Acorn developed the early ARM cores, but it was Apple who first used one of these chips in a general‑market product: their Newton PDA announced in 1992 featured an ARM6 chip. The Newton struggled to establish itself in the face of competition from the Palm Pilot, and its commercial failure led ARM to embrace a different business model: rather than create end‑user products, they’d license their IP to production partners, who would make their own chips based on ARM’s RISC technology.

Apple’s move from Intel to Arm‑based CPUs may seem relatively recent, but along with Acorn they were one of ARM’s founding companies and began selling their first ARM‑powered device, the Newton, way back in 1993.

Starting with the Nokia 6110 GSM phone, which appeared in 1997, their chips found favour in the mobile devices sector, and today they are reportedly found in 99 percent of smartphones. Of course, we’re more interested in their potential for ‘serious’ machines that might be used for music making, so it’s perhaps of greater interest that, in 2012, alongside Windows 8, Microsoft brought out the 32‑bit Windows RT OS for ARM‑based mobile devices. This was used for the Surface tablets and a small number of third‑party devices, all built around Nvidia Tegra 32‑bit ARM processors. Ultimately, the platform’s inability to run popular PC software meant that these devices were largely passed over for more traditional x86‑based laptops.

Then, in 2016, Microsoft launched a 64‑bit ARM version of Windows 10 for Qualcomm’s Snapdragon devices. Hardware shipped over the following years from Microsoft, ASUS and HP, amongst others, but the platform failed to capture users’ attention from the more powerful PC laptops and desktops already on the market and, again, it wasn’t regarded as a commercial success.

Meanwhile, Apple had been successfully using their own custom ARM chips in various devices, including iPads and iPhones, since 2009. But in June 2020 they signalled that they’d be moving from Intel to Arm‑based processors for their more powerful desktop and laptop machines. Their M1 chip first appeared in a product at the end of 2020, and the M‑series chips are now found throughout their product range. Importantly, these Apple chips have enjoyed a good reputation for both power and efficiency, and the products have proved commercially successful.

Right Here, Right Now

This brings us to the present day. As you’ve probably gleaned, in the past it was often software incompatibility that limited the user appeal of older Windows‑based ARM devices. But the emulation layer has matured over the years, and there’s also been a steady increase in PC RISC chip performance. Having the flexibility to run existing applications through the emulation layer should help the platform to expand, and with greater uptake the incentive for developers to compile software specifically for Arm should increase too.

With the barriers to wider adoption falling away, various parties (including silicon behemoths AMD and Nvidia) have announced their entry into the Arm computer market, and new hardware is on the way. Small wonder, then, that Microsoft seem keen not only to support the newest Snapdragon X Elite‑equipped laptop models, but also to deliver the sort of accessible, more streamlined experience that should make the platform attractive for typical computer users.

With ample power and superb battery life, Snapdragon X Elite‑equipped laptops now have the potential to be great portable music‑making machines.

Pete Brown & The ASIO...