https://cherla.org/tags/gpu/Recent content in GPU on Srikanth CherlaHugoen-USMon, 22 Jun 2026 16:01:29 +0200https://cherla.org/posts/2026/06/deploying-flux.1-kontext-dev-on-cloud-run-gpu/Fri, 19 Jun 2026 00:00:00 +0000https://cherla.org/posts/2026/06/deploying-flux.1-kontext-dev-on-cloud-run-gpu/<style> :root { --bg: #FAFAF7; --bg-card: #FFFFFF; --text: #1A1A18; --text-secondary: #6B6B66; --text-tertiary: #9C9C96; --border: #E5E5DF; --border-light: #F0F0EA; --purple: #7F77DD; --purple-light: #EEEDFE; --purple-dark: #3C3489; --amber: #EF9F27; --amber-light: #FAEEDA; --amber-dark: #633806; --teal: #1D9E75; --teal-light: #E1F5EE; --teal-dark: #085041; --coral: #D85A30; --coral-light: #FAEAE3; --coral-dark: #5C200A; --blue: #378ADD; --callout-bg: #F3F2FE; --callout-text: #2A2458; --font-body: 'Source Serif 4', Georgia, serif; --font-ui: 'DM Sans', system-ui, sans-serif; --font-mono: 'JetBrains Mono', monospace; } @media (prefers-color-scheme: dark) { :root { --bg: #141413; --bg-card: #1E1E1C; --text: #E8E6DD; --text-secondary: #A3A29B; --text-tertiary: #6B6A65; --border: #2E2E2B; --border-light: #252523; --purple-light: #26215C; --amber-light: #412402; --teal-light: #04342C; --coral-light: #3D1A10; --purple-dark: #CECBF6; --amber-dark: #FAC775; --teal-dark: #9FE1CB; --coral-dark: #F5B49A; --callout-bg: #28235A; --callout-text: #D4D0F8; } } .post { max-width: 720px; margin: 0 auto; padding: 4rem 1.5rem 6rem; } .lead { font-size: 19px; color: var(--text-secondary); line-height: 1.75; margin-bottom: 2.5rem; font-family: var(--font-body); } .post h2 { font-family: var(--font-ui); font-size: 14px; font-weight: 600; letter-spacing: 0.06em; text-transform: uppercase; color: var(--text-secondary); margin: 3.5rem 0 1.25rem; padding-bottom: 0.5rem; border-bottom: 1px solid var(--border); } .post h3 { font-family: var(--font-ui); font-size: 16px; font-weight: 600; color: var(--text); margin: 2rem 0 0.75rem; } .post p { margin-bottom: 1.25rem; line-height: 1.75; font-family: var(--font-body); font-size: 17px; } .post strong { font-weight: 600; } .post em { font-style: italic; } .post a { color: var(--purple); } .post code { font-family: var(--font-mono); font-size: 0.82em; background: var(--border-light); padding: 2px 6px; border-radius: 4px; } .post pre { background: var(--bg-card); border: 1px solid var(--border); border-radius: 8px; padding: 1rem 1.25rem; overflow-x: auto; margin: 1.5rem 0; } .post pre code { background: none; padding: 0; font-size: 0.84em; line-height: 1.65; color: var(--text); } .post ul, .post ol { padding-left: 1.5rem; margin-bottom: 1.25rem; } .post li { margin-bottom: 0.4rem; font-family: var(--font-body); font-size: 17px; line-height: 1.7; } .post hr { border: none; border-top: 1px solid var(--border); margin: 3rem 0; } /* Viz containers */ .viz-container { margin: 2rem -0.5rem; padding: 1.5rem; background: var(--bg-card); border: 1px solid var(--border); border-radius: 12px; overflow: hidden; } .viz-caption { font-family: var(--font-ui); font-size: 13px; color: var(--text-tertiary); margin-top: 12px; text-align: center; line-height: 1.5; } .anim-fade-in { opacity: 0; transform: translateY(12px); transition: opacity 0.6s ease, transform 0.6s ease; } .anim-fade-in.visible { opacity: 1; transform: translateY(0); } /* Callouts */ .callout { border-left: 3px solid var(--purple); padding: 1rem 1.25rem; margin: 1.5rem 0; background: var(--callout-bg); color: var(--callout-text); border-radius: 0 8px 8px 0; font-size: 16px; line-height: 1.7; font-family: var(--font-body); } .callout.warn { border-left-color: var(--amber); background: var(--amber-light); color: var(--amber-dark); } .callout.good { border-left-color: var(--teal); background: var(--teal-light); color: var(--teal-dark); } /* Buttons */ .sv-btn { font-family: var(--font-ui); font-size: 13px; padding: 6px 16px; border: 1px solid var(--border); background: var(--bg-card); color: var(--text-secondary); border-radius: 8px; cursor: pointer; transition: all 0.2s; } .sv-btn:hover { background: var(--border-light); } .sv-btn.active { background: var(--purple); color: #fff; border-color: var(--purple); } /* Shape tags */ .shape-tag { display: inline-block; font-family: var(--font-mono); font-size: 12px; background: var(--border-light); border: 1px solid var(--border); padding: 2px 8px; border-radius: 6px; color: var(--text-secondary); margin: 2px; } /* Stepper */ .step-nav { display: flex; gap: 4px; margin-bottom: 16px; flex-wrap: wrap; } .step-title { font-family: var(--font-ui); font-size: 14px; font-weight: 600; margin-bottom: 6px; color: var(--text); } .step-desc { font-family: var(--font-ui); font-size: 13px; color: var(--text-secondary); margin-bottom: 12px; line-height: 1.6; } /* Storage comparison bars */ .metric-table { width: 100%; border-collapse: collapse; } .metric-table th { font-family: var(--font-ui); font-size: 11px; font-weight: 600; letter-spacing: 0.05em; text-transform: uppercase; color: var(--text-tertiary); padding: 0 8px 10px; text-align: left; } .metric-table th:not(:first-child) { text-align: center; } .metric-table td { padding: 6px 8px; vertical-align: middle; border-top: 1px solid var(--border-light); } .metric-row-label { font-family: var(--font-ui); font-size: 13px; color: var(--text-secondary); white-space: nowrap; min-width: 110px; } .bar-cell { position: relative; min-width: 120px; } .bar-track { height: 20px; border-radius: 4px; background: var(--border-light); position: relative; overflow: hidden; } .bar-fill { height: 100%; border-radius: 4px; transition: width 0.8s cubic-bezier(0.4,0,0.2,1); } .bar-val { font-family: var(--font-mono); font-size: 11px; color: var(--text-secondary); margin-top: 3px; display: block; } /* Cold start bars */ .cs-row { display: flex; align-items: center; gap: 12px; margin: 10px 0; } .cs-label { font-family: var(--font-ui); font-size: 13px; color: var(--text-secondary); min-width: 180px; flex-shrink: 0; } .cs-bar-wrap { flex: 1; } .cs-track { height: 22px; border-radius: 5px; background: var(--border-light); position: relative; overflow: visible; } .cs-fill { height: 100%; border-radius: 5px; display: flex; align-items: center; padding-left: 8px; font-family: var(--font-mono); font-size: 11px; color: rgba(255,255,255,0.9); white-space: nowrap; overflow: hidden; transition: width 0.8s cubic-bezier(0.4,0,0.2,1); } .cs-note { font-family: var(--font-ui); font-size: 11px; color: var(--text-tertiary); margin-top: 3px; } /* VRAM stacked bar */ .vram-outer { margin: 16px 0; } .vram-label-row { display: flex; justify-content: space-between; align-items: baseline; margin-bottom: 6px; } .vram-config-label { font-family: var(--font-ui); font-size: 13px; font-weight: 600; color: var(--text); } .vram-total-label { font-family: var(--font-mono); font-size: 12px; color: var(--text-tertiary); } .vram-bar { display: flex; height: 44px; border-radius: 8px; overflow: hidden; position: relative; } .vram-seg { display: flex; align-items: center; justify-content: center; font-family: var(--font-mono); font-size: 10px; color: rgba(255,255,255,0.9); white-space: nowrap; overflow: hidden; transition: all 0.5s ease; } .vram-seg.cpu { background: repeating-linear-gradient(45deg, var(--amber-light) 0px, var(--amber-light) 5px, transparent 5px, transparent 10px); border: 1px dashed var(--amber); color: var(--amber-dark); } .vram-empty { flex: 1; background: var(--border-light); display: flex; align-items: center; justify-content: center; font-family: var(--font-ui); font-size: 11px; color: var(--text-tertiary); } .vram-legend { display: flex; gap: 16px; flex-wrap: wrap; margin-top: 10px; font-family: var(--font-ui); font-size: 12px; color: var(--text-secondary); } .vleg { display: flex; align-items: center; gap: 5px; } .vleg-dot { width: 12px; height: 12px; border-radius: 3px; flex-shrink: 0; } .vleg-dot.dashed { background: var(--amber-light); border: 1.5px dashed var(--amber); } .vram-tab { display: flex; gap: 4px; margin-bottom: 16px; } /* Bug stepper grid */ .grid-viz { display: flex; justify-content: center; align-items: flex-start; gap: 24px; flex-wrap: wrap; } .grid-block { text-align: center; } .grid-label { font-family: var(--font-mono); font-size: 11px; color: var(--text-tertiary); margin-bottom: 6px; } .grid-annotation { font-family: var(--font-ui); font-size: 12px; margin-top: 6px; } /* Detector diagram */ .detector-wrap { display: flex; flex-direction: column; align-items: center; gap: 0; font-family: var(--font-ui); font-size: 13px; } .det-node { border: 1.5px solid var(--border); border-radius: 8px; padding: 8px 20px; text-align: center; background: var(--bg-card); font-size: 13px; } .det-node.decision { background: var(--purple-light); border-color: var(--purple); color: var(--purple-dark); font-weight: 600; } .det-node.bad { background: var(--coral-light); border-color: var(--coral); color: var(--coral-dark); } .det-node.good { background: var(--teal-light); border-color: var(--teal); color: var(--teal-dark); opacity: 0.5; } .det-arrow { width: 1px; height: 20px; background: var(--border); margin: 0 auto; position: relative; } .det-arrow::after { content: ''; position: absolute; bottom: -1px; left: -4px; border-left: 4.5px solid transparent; border-right: 4.5px solid transparent; border-top: 6px solid var(--border); } .det-branches { display: flex; gap: 40px; } .det-branch { display: flex; flex-direction: column; align-items: center; gap: 0; } .det-branch-label { font-family: var(--font-mono); font-size: 11px; color: var(--text-tertiary); margin: 4px 0; } /* Auth diagram */ .auth-diagram { display: flex; flex-direction: column; gap: 0; align-items: stretch; } .auth-layer { border: 1.5px solid var(--border); border-radius: 10px; padding: 12px 16px; background: var(--bg-card); } .auth-layer-title { font-family: var(--font-ui); font-size: 13px; font-weight: 600; color: var(--text); margin-bottom: 3px; } .auth-layer-desc { font-family: var(--font-ui); font-size: 12px; color: var(--text-secondary); line-height: 1.5; } .auth-arrow { text-align: center; padding: 6px 0; color: var(--text-tertiary); font-size: 18px; line-height: 1; } .auth-path-row { display: flex; gap: 12px; margin-top: 16px; flex-wrap: wrap; } .auth-path { flex: 1; min-width: 200px; border: 1px solid var(--border); border-radius: 8px; padding: 10px 14px; background: var(--bg-card); } .auth-path-title { font-family: var(--font-ui); font-size: 12px; font-weight: 600; color: var(--text); margin-bottom: 4px; } .auth-path-desc { font-family: var(--font-ui); font-size: 12px; color: var(--text-secondary); line-height: 1.5; } .auth-path.prod { border-color: var(--teal); } .auth-path.dev { border-color: var(--amber); } /* Token viz */ .token-section-label { font-family: var(--font-ui); font-size: 11px; color: var(--text-tertiary); text-align: center; margin-bottom: 4px; } .token-row { display: flex; gap: 2px; flex-wrap: wrap; justify-content: center; } .tok { width: 18px; height: 24px; border-radius: 3px; flex-shrink: 0; transition: opacity 0.3s; } .tok.both { background: var(--purple); opacity: 0.7; } .tok.t5only { background: var(--amber); opacity: 0.6; } .tok.truncated { background: var(--border); opacity: 0.5; } .tok-legend { display: flex; gap: 16px; justify-content: center; flex-wrap: wrap; font-family: var(--font-ui); font-size: 12px; color: var(--text-secondary); margin-top: 10px; } .tok-dot { width: 12px; height: 12px; border-radius: 3px; display: inline-block; vertical-align: -2px; margin-right: 5px; } .clip-marker { font-family: var(--font-ui); font-size: 11px; color: var(--purple-dark); text-align: center; margin: 6px 0 2px; } .t5-marker { font-family: var(--font-ui); font-size: 11px; color: var(--amber-dark); text-align: center; margin: 6px 0 2px; } </style> <link rel="preconnect" href="https://fonts.googleapis.com"> <link href="https://fonts.googleapis.com/css2?family=Source+Serif+4:ital,wght@0,400;0,600;1,400&family=JetBrains+Mono:wght@400;500&family=DM+Sans:wght@400;500;600&display=swap" rel="stylesheet"> <article class="post"> <p class="lead">Six steps to get a self-hosted image generation service into production on Cloud Run GPU — what I landed on at each step, how I got there, and what each one still costs.</p>