In 1-2 weeks, we'll be releasing an updated platform for 3rd party developers, including individuals and companies like VSL. This will be our first public release, featuring the latest binaries, hardware, and software platforms (i.e., M4 silicon) support. I’ll confirm the exact support we’ll provide, but once it’s live and VSL integrates it, you’ll be able to upgrade on your end.
This marks the start of smoother, more transparent, and frequent updates moving forward. I'll keep you posted about the progress here. I am in the contact with VSL team.
Hi Sasha,
In 1-2 weeks it'll be a YEAR ago for the first update announcing to appear. A year ago Sasha. I guess you'd understand my skepticism. Seeing is believing in this case. Looking forward thought.
In 1-2 weeks, we'll be releasing an updated platform for 3rd party developers, including individuals and companies like VSL. This will be our first public release, featuring the latest binaries, hardware, and software platforms (i.e., M4 silicon) support. I’ll confirm the exact support we’ll provide, but once it’s live and VSL integrates it, you’ll be able to upgrade on your end.
This marks the start of smoother, more transparent, and frequent updates moving forward. I'll keep you posted about the progress here. I am in the contact with VSL team.
Hi Sasha,
In 1-2 weeks it'll be a YEAR ago for the first update announcing to appear. A year ago Sasha. I guess you'd understand my skepticism. Seeing is believing in this case. Looking forward thought.
Regards,
Willem
Hi Willem,
A quick update: we’re just days away from the SDK release, as the team has informed me.
To provide more context: this is the first-ever technology of its kind, created by a small startup. It’s something that can’t be built using existing frameworks like CUDA or OpenCL or SPIR-V, as they won’t allow you to handle multiple software instances with low latency. What we’re doing requires ISA-level technology to be created and maintained, which means essentially writing assembler code tailored for each GPU architecture. On top of that, we’re working with different operating systems, drivers, GPUs, and frameworks. Creating a single binary that covers all of these is nearly impossible—but we’re still managing to do it.
I understand your frustration about unmet expectations, but please know that we’re not just sitting back and relaxing. A one-year release covering Windows, Nvidia, AMD, Mac, and M-Silicon is incredibly fast for a platform launch.
It should become publicly available in a few days. This release includes M4 support and some AMD support (some fixes, I don't know specific details yet). The next step is VSL takes new binary and integrates them back into Vienna Power House. It should not take too much of time. The ongoing GPU AUDIO Platform updates shall be easy to adopt.
Hey everyone, the SDK is out (https://gpu.audio/sdk), the binaries are there, now it's VSL's turn to re-integrate / update and ship the new version of Vienna Power House to you guys.
Hello, I have a Macmini with M2 Pro, 32GB memory, Sonoma 14.6.1 macOS, I installed MIR 3D with Power House and works fine with cpu. Buffer @512. I tested on Logic, Reaper and Luna different sample rates. Started MIR ahead or later but as soon as I enable gpu all DAWs quit.
Is there a fundamental thing I’m missing? Are the Apple M2 chips gpu not supported?
I have a new Nvidia RTX 5080 and it doesn't seem to work with Vienna Power House. On the MIR 3D plugins in Cubase I am shown this error:
Could not create convolver. Error: 2
Secondly, I cannot seem to find any option to choose GPU. I can't find any settings or options for Vienna Power House anywhere. Just the button GPU Power.
I do have integrated graphics in the CPU and I want to be sure that's not being used.
Things I wonder is how/if the Impulse Response Cache Size affects GPU memory and what happens if it's set to lower than the loaded venue. Some statistics would be nice to see to understand how much is being used.
I have a new Nvidia RTX 5080 and it doesn't seem to work with Vienna Power House. On the MIR 3D plugins in Cubase I am shown this error:
Could not create convolver. Error: 2
Secondly, I cannot seem to find any option to choose GPU. I can't find any settings or options for Vienna Power House anywhere. Just the button GPU Power.
I do have integrated graphics in the CPU and I want to be sure that's not being used.
Things I wonder is how/if the Impulse Response Cache Size affects GPU memory and what happens if it's set to lower than the loaded venue. Some statistics would be nice to see to understand how much is being used.
AFAIK powerhouse is just the GPU button though from post #43 it looks like there is (or was) some control over what GPU is being used. I'd be keen to know how this goes because I have a Power House license, but have been holding off on a new NVIDIA card until things get a bit more stable.
