@Mary said:
@Sasha-T said:
Hi everyone!
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.
Thanks for understanding!
