#hardwarechoices — Public Fediverse posts

Debating TinyBox (TinyCorp) vs NVIDIA DGX Station is like choosing between a tactical off-grid cabin and a corporate panic room.
TinyBox: OSS-friendly, sips power, runs whisper quiet—great for AI nerds who want root.
DGX: Raw CUDA firepower, but you’ll need NV cash and maybe a colocation rack.
Depends if you want control… or just raw scale.
#AI #TinyBox #DGX #SelfHosted #OpenSource #HardwareChoices

Planning to locally host an LLM 🤖. Having a lot of internal struggles figuring out the best price curve.

A used NVIDIA Jetson Xavier NX can be acquired for ~$250 💰.
A 4060 w/ 16gb for ~$400 💳.

Not sure I really want to do either 🤔. Waiting for NPU support on Transformer models may be a better approach.

#LLM #LocalHosting #NVIDIA #AI #MachineLearning #NPU #Transformers #TechDilemmas #HardwareChoices

Planning to locally host an LLM 🤖. Having a lot of internal struggles figuring out the best price curve.

A used NVIDIA Jetson Xavier NX can be acquired for ~$250 💰.
A 4060 w/ 16gb for ~$400 💳.

Not sure I really want to do either 🤔. Waiting for NPU support on Transformer models may be a better approach.

#LLM #LocalHosting #NVIDIA #AI #MachineLearning #NPU #Transformers #TechDilemmas #HardwareChoices

Planning to locally host an LLM 🤖. Having a lot of internal struggles figuring out the best price curve.

A used NVIDIA Jetson Xavier NX can be acquired for ~$250 💰.
A 4060 w/ 16gb for ~$400 💳.

Not sure I really want to do either 🤔. Waiting for NPU support on Transformer models may be a better approach.

#LLM #LocalHosting #NVIDIA #AI #MachineLearning #NPU #Transformers #TechDilemmas #HardwareChoices

Planning to locally host an LLM 🤖. Having a lot of internal struggles figuring out the best price curve.

A used NVIDIA Jetson Xavier NX can be acquired for ~$250 💰.
A 4060 w/ 16gb for ~$400 💳.

Not sure I really want to do either 🤔. Waiting for NPU support on Transformer models may be a better approach.

#LLM #LocalHosting #NVIDIA #AI #MachineLearning #NPU #Transformers #TechDilemmas #HardwareChoices

Planning to locally host an LLM 🤖. Having a lot of internal struggles figuring out the best price curve.

A used NVIDIA Jetson Xavier NX can be acquired for ~$250 💰.
A 4060 w/ 16gb for ~$400 💳.

Not sure I really want to do either 🤔. Waiting for NPU support on Transformer models may be a better approach.

#LLM #LocalHosting #NVIDIA #AI #MachineLearning #NPU #Transformers #TechDilemmas #HardwareChoices