diff --git a/docs/deployment/linux/api.md b/docs/deployment/linux/api.md index 423d6d09..6e99045a 100644 --- a/docs/deployment/linux/api.md +++ b/docs/deployment/linux/api.md @@ -1,8 +1,8 @@ # Deploy SkyEye on Linux - Cloud API -This guide is a step-by-step on how to run SkyEye on a Linux computer or server alongside DCS, TacView and SRS Server, using the OpenAI API for cloud speech recognition. +This guide is a step-by-step on how to run SkyEye on a Linux computer, using the OpenAI API for cloud speech recognition. This can be the same computer running DCS, TacView and SRS Server, or a separate one. -You can also deploy SkyEye on Linux using either [a separate computer using CPU speech recognition](cpu.md) or [the same computer as DCS using GPU speech recognition](gpu.md). +You can also deploy SkyEye on Linux using either [GPU speech recognition](gpu.md) or [a separate computer using CPU speech recognition](cpu.md). ## Getting Help diff --git a/docs/deployment/linux/cpu.md b/docs/deployment/linux/cpu.md index 0c8e8e8a..5046441a 100644 --- a/docs/deployment/linux/cpu.md +++ b/docs/deployment/linux/cpu.md @@ -1,8 +1,8 @@ # Deploy SkyEye on Linux - CPU -This guide is a step-by-step on how to run SkyEye on a Linux computer or server, separate from the computer running DCS, TacView and SRS Server, using the CPU for local speech recognition. This guide is not tied to any particular hosting provider; it works on a spare Linux machine, a rented cloud server, or a container host. If you want a guide tailored to a specific cloud provider, see the [Hetzner Cloud](../cloud-providers/hetzner.md) or [Vultr](../cloud-providers/vultr.md) guides. +This guide is a step-by-step on how to run SkyEye on a Linux computer, separate from the computer running DCS, TacView and SRS Server, using the CPU for local speech recognition. This guide is not tied to any particular hosting provider; it works on either a Linux computer you own, or a rented cloud server. If you want a guide tailored to a specific cloud provider, see the [Hetzner Cloud](../cloud-providers/hetzner.md) or [Vultr](../cloud-providers/vultr.md) guides. -You can also deploy SkyEye on Linux using either [the same computer as DCS using GPU speech recognition](gpu.md) or [cloud API speech recognition](api.md). +You can also deploy SkyEye on Linux using either [GPU speech recognition](gpu.md) or [cloud API speech recognition](api.md). ## Getting Help @@ -10,7 +10,7 @@ See [the admin guide](../../ADMIN.md#getting-help) for how to get help if you ha ## Hardware Requirements -This guide requires a **second** Linux computer or server, separate from the one running DCS, TacView and SRS Server. The computer running SkyEye needs a fast, multithreaded, **dedicated** CPU with support for [AVX2](https://en.wikipedia.org/wiki/Advanced_Vector_Extensions#Advanced_Vector_Extensions_2), 3GB of RAM, and about 2GB of disk space. +This guide requires a **second** Linux computer, separate from the one running DCS, TacView and SRS Server. The computer running SkyEye needs a fast, multithreaded, **dedicated** CPU with support for [AVX2](https://en.wikipedia.org/wiki/Advanced_Vector_Extensions#Advanced_Vector_Extensions_2), 3GB of RAM, and about 2GB of disk space. CPU Series|AVX2 Added In -|- diff --git a/docs/deployment/linux/gpu.md b/docs/deployment/linux/gpu.md index 1249e2b3..c46de10f 100644 --- a/docs/deployment/linux/gpu.md +++ b/docs/deployment/linux/gpu.md @@ -1,8 +1,8 @@ # Deploy SkyEye on Linux - GPU -This guide is a step-by-step on how to run SkyEye on the same Linux computer as DCS, TacView and SRS Server, using the GPU for local speech recognition via the experimental Vulkan build. +This guide is a step-by-step on how to run SkyEye on a Linux computer with a GPU, using local speech recognition via the experimental Vulkan build. This can be the same computer running DCS, TacView and SRS Server, or a separate one. -You can also deploy SkyEye on Linux using either [a separate computer using CPU speech recognition](cpu.md) or [cloud API speech recognition](api.md). +You can also deploy SkyEye on Linux using either [cloud API speech recognition](api.md) or [a separate computer using CPU speech recognition](cpu.md). > ⚠️ The Vulkan (GPU) build of SkyEye is experimental. Performance and speech recognition quality can vary significantly between GPU models — it might run great on one and perform poorly on another. I can only test against the GPU hardware I personally own, so I have no control over and very limited ability to troubleshoot how well the Vulkan build runs on any particular GPU or driver. diff --git a/docs/deployment/macos/gpu.md b/docs/deployment/macos/gpu.md index 49bea2ce..8520966e 100644 --- a/docs/deployment/macos/gpu.md +++ b/docs/deployment/macos/gpu.md @@ -1,8 +1,6 @@ # Deploy SkyEye on macOS - GPU -This guide is a step-by-step on how to run SkyEye on macOS, using the GPU/Neural Engine for local speech recognition. This is the fastest and highest-quality way to run SkyEye, but requires an Apple Silicon Mac and a separate Windows computer for DCS, TacView and SRS Server. - -There is no cloud API or CPU-only guide for macOS: local speech recognition on the GPU/Neural Engine is the only recommended way to run SkyEye on macOS. +This guide is a step-by-step on how to run SkyEye on an Apple Silicon computer, separate from the computer running DCS, TacView and SRS Server, using the Apple Silicon GPU for local speech recognition. ## Getting Help @@ -10,7 +8,9 @@ See [the admin guide](../../ADMIN.md#getting-help) for how to get help if you ha ## Hardware Requirements -You need an Apple Silicon Mac (M-series), such as a Mac Mini or MacBook Air/Pro. SkyEye requires around 3GB of RAM and about 2GB of disk space. Intel Macs are not supported. +You need an Apple Silicon Mac, such as a Mac Mini or MacBook Air/Pro. SkyEye requires around 3GB of RAM and about 2GB of disk space. + +Intel Macs are not supported. ## Set Up DCS, TacView, and SRS