guilt

Comparison of Interconnect Technologies: Hardware, Latency, Lanes, Overhead, Throughput, Vendor, and OS Support

Interconnect	Hardware Support	Latency	Number of Lanes	Protocol Overhead	True Unidirectional Throughput Per Lane	True Unidirectional Throughput (Overall)	True Bidirectional Throughput	Vendor Support	OS Support
NVLink	NVIDIA GPUs (Pascal, Volta, Hopper, Blackwell), NVIDIA CPUs (e.g., Grace)	Sub-µs (0.1–0.5 µs, intra-node P2P)	72 (18 ports × 4 lanes, NVLink 5.0)	~2.5–3.5% (128b/130b, headers)	6.09 GB/s (48.72 Gbps)	438.75 GB/s	877.5 GB/s	NVIDIA	Linux (primary), Windows (limited via WSL)
InfiniBand	CPUs, NVIDIA/AMD/Intel GPUs (via GPUDirect RDMA), NICs (e.g., Mellanox)	~1–5 µs (point-to-point)	32 (8 ports × 4 lanes, N

guilt

try:
    import cupy as cp
except ImportError:
    import numpy as cp

try:
    from rich.progress import Progress, BarColumn, TextColumn, TimeRemainingColumn, TimeElapsedColumn
except ImportError:
    Progress = None

guilt

#!/usr/bin/env sh

# Variable for the script description
_ARGPARSE_SCRIPT_DESCRIPTION=""

# Use this one wierd trick to shorten your CLI Boolean Flags
_ARGPARSE_FLAG_BEHAVIOUR="Add this flag to set the argument to 'true' (don't use '--flag true')"

# Declare an associative array for argument properties
_ARGPARSE_ARG_VALUES=()

guilt

LLAMA Benchmark Comparisons

LLAMA.cpp Version

b3051

Model

TinyLlama 1.1B Chat

guilt

Raspberry Pi Setup Steps

1. Make the Box Ready. Either Keep the Box or Attach the Racking Unit
2. Write Image, note down Hostname
3. Install MicroSD Card
4. Connect PoE and Reboot

Raspberry Pi Imaging Manual

1. Use rapios-lite

guilt

%define		_snapid H8ZpNgIoPyvmkgxOWw5MSzsXK1wRZiHn
%define		_snaprev 18

Name:           authy
Version:        2.2.3
Release:        1%{?dist}
Summary:        Two factor authentication desktop application
ExclusiveArch:  x86_64
License:        Proprietary
URL:            https://authy.com/

guilt

# Custom Aliases

alias dir = ls
alias del = rm
alias ren = mv
alias move = mv
alias rmdir = rm
alias rd = rmdir
alias open = start

guilt

// Demo of unroll.h
#include "unroll.h"

int main()
{
	int v;

	U_unroll(v, 17, {
		printf("Up Unroll\n");
	});

guilt

import os

from accelerate import Accelerator
from diffusers import StableCascadeDecoderPipeline, StableCascadePriorPipeline
from slugify import slugify
import torch

priorModelId = "stabilityai/stable-cascade-prior"
modelId = "stabilityai/stable-cascade"
dataType = torch.bfloat16 #Use torch.float32 alternatively

guilt

ROCm Setup Steps

ROCm Docker Image

1. Install Docker and ensure you can run docker ps correctly, add yourself to the docker group if necessary.
2. Run build-rocm.sh builds a ROCm docker image for your Linux System. It is configured to use the render group configured in your Linux distribution and ensure that /dev/kfd and /dev/dri are writeable by the render group users. Ensure you are added to the render group if necessary.
3. Run launch-rocm.sh if you wish to only use ROCm with the docker image you built.