DEV Community

#gpu

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
How much GPU do you need to run an LLM locally? The rules of thumb we use
froid1911 profile

How much GPU do you need to run an LLM locally? The rules of thumb we use

#llm #selfhosted #ai #gpu
Comments
4 min read
Clover-35B-A3B on llama.cpp: what "AR mode only" means, and a GGUF that answers in Gs
inferencekeydev profile

Clover-35B-A3B on llama.cpp: what "AR mode only" means, and a GGUF that answers in Gs

#ai #llm #gpu #selfhosted
Comments
4 min read
Your CUDA Code Didn't Change. Your Risk Register Just Did: What Nvidia's CEO Exit Means for Builders
shivaysinghrajput profile

Your CUDA Code Didn't Change. Your Risk Register Just Did: What Nvidia's CEO Exit Means for Builders

#nvidia #ai #gpu #cuda
Comments
4 min read
What are the best SaladCloud alternatives for GPU workloads?
viswaretas_kotra profile

What are the best SaladCloud alternatives for GPU workloads?

#gpu #machinelearning #cloud #ai
Comments
8 min read
How Many Closed Findings Before You Trust a Routing Change
qisuancloud profile

How Many Closed Findings Before You Trust a Routing Change

#llm #ai #gpu #inference
1
Comments
10 min read
How do you let Cursor's agent run code on a cloud GPU?
viswaretas_kotra profile

How do you let Cursor's agent run code on a cloud GPU?

#cursor #mcp #ai #gpu
Comments
8 min read
GPU Paths Honestly: What csperf Can and Cannot Claim Today
aabhinavg profile

GPU Paths Honestly: What csperf Can and Cannot Claim Today

#gpu #hip #opencl #csperf
Comments
4 min read
How can an AI startup get free GPU and cloud credits?
viswaretas_kotra profile

How can an AI startup get free GPU and cloud credits?

#startup #cloud #gpu #machinelearning
Comments
9 min read
Serving Hy-SimulMT live on one 48 GB GPU: the startup spike, the FP8 trap, what to measure
inferencekeydev profile

Serving Hy-SimulMT live on one 48 GB GPU: the startup spike, the FP8 trap, what to measure

#ai #machinelearning #gpu #selfhosted
1
Comments
5 min read
LoRA vs full fine-tuning: when is LoRA good enough?
viswaretas_kotra profile

LoRA vs full fine-tuning: when is LoRA good enough?

#machinelearning #llm #finetuning #gpu
Comments 1
8 min read
Certified kernels, guaranteed at install time
benkelaya profile

Certified kernels, guaranteed at install time

#opensource #machinelearning #python #gpu
Comments
2 min read
A100, H100, H200, B200, B300: which GPU should you pick?
viswaretas_kotra profile

A100, H100, H200, B200, B300: which GPU should you pick?

#gpu #machinelearning #nvidia #ai
Comments
8 min read
Serving Groot-19B-A3B at 256K on one RTX 5090: the memory budget and the thinking trap
inferencekeydev profile

Serving Groot-19B-A3B at 256K on one RTX 5090: the memory budget and the thinking trap

#ai #llm #gpu #selfhosted
Comments
5 min read
Nvidia lost $600B in 4 days, then its CEO resigned
frankchu profile

Nvidia lost $600B in 4 days, then its CEO resigned

#discuss #ai #nvidia #gpu
1
Comments
4 min read
Continuous Batching: Why Your GPU Waits for the Slowest Request
net_satsawat profile

Continuous Batching: Why Your GPU Waits for the Slowest Request

#llm #machinelearning #performance #gpu
Comments
7 min read
đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.