[{"data":1,"prerenderedAt":813},["ShallowReactive",2],{"blog-post-nl-\u002Fblog\u002Fbuilding-the-engine-for-the-ai-race-the-4-month-path-to-nvidia-gb300-nvl72-power":3,"blog-posts-sidebar-nl":341},{"id":4,"title":5,"body":6,"categories":313,"date":327,"description":328,"extension":329,"heading":330,"image":331,"meta":332,"navigation":333,"originalUrl":334,"path":335,"seo":336,"slug":337,"stem":338,"updated":339,"__hash__":340},"blogNl\u002Fblog\u002Fbuilding-the-engine-for-the-ai-race-the-4-month-path-to-nvidia-gb300-nvl72-power.md","NVIDIA GB300 NVL72: modulair datacenter in vier maanden",{"type":7,"value":8,"toc":302},"minimark",[9,13,16,24,27,30,36,39,44,71,78,84,87,90,95,115,118,123,129,132,137,142,173,178,192,198,201,221,227,230,235,285,290,296,299],[10,11,12],"p",{},"Bij AI-infrastructuur is snelheid een essentieel concurrentieel voordeel. Van de trainingssnelheid van modellen tot inferentielatency: elke milliseconde telt. Maar nog vóór de eerste workload draait, bepaalt één cruciale factor vaak het succes of falen: time-to-market.",[10,14,15],{},"Traditionele bouwprojecten nemen doorgaans 18 tot 24 maanden in beslag. In de huidige AI-cyclus is dat simpelweg te traag. We hebben onze modulaire aanpak zo verfijnd dat we binnen vier maanden operationeel kunnen zijn, met een gebruiksklare faciliteit die specifiek is ontworpen voor de NVIDIA GB300 NVL72.",[17,18,20],"h3",{"id":19},"de-uitdaging-uitzonderlijk-krachtige-hardware-huisvesten",[21,22,23],"strong",{},"De uitdaging: uitzonderlijk krachtige hardware huisvesten",[10,25,26],{},"De NVIDIA GB300 NVL72 integreert 72 NVIDIA Blackwell Ultra GPU's en 36 Arm-gebaseerde NVIDIA Grace CPU's in een volledig vloeistofgekoelde, rack-scale architectuur.",[10,28,29],{},"Dat verandert de eisen aan datacenterinfrastructuur fundamenteel. We voeden niet zomaar servers, maar een cluster met een extreem hoge vermogensdichtheid waarin elk rack bij normale belasting ongeveer 120 kW verbruikt en tijdens intensieve workloads mogelijk een piek van 150 kW bereikt.",[17,31,33],{"id":32},"architectuuroverzicht-precisie-op-schaal",[21,34,35],{},"Architectuuroverzicht: precisie op schaal",[10,37,38],{},"Om die densiteit aan te kunnen, gebruikt onze faciliteit een lay-out die is geoptimaliseerd voor de specifieke footprint van de GB300.",[10,40,41],{},[21,42,43],{},"Onze faciliteit biedt plaats aan:",[45,46,47,54],"ul",{},[48,49,50,53],"li",{},[21,51,52],{},"16 GB300 NVL72-racks van 600 mm"," voor de primaire compute-infrastructuur",[48,55,56,59,60],{},[21,57,58],{},"12 extra racks van 800 mm",", verdeeld als:\n",[45,61,62,65,68],{},[48,63,64],{},"5 racks voor netwerkinfrastructuur",[48,66,67],{},"5 racks voor beheer- en opslagsystemen",[48,69,70],{},"2 racks voor toekomstige uitbreiding",[10,72,73],{},[74,75],"img",{"alt":76,"src":77},"","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fsingle_unit_invert-1.webp",[17,79,81],{"id":80},"stroomdistributie-de-4-makes-3-architectuur",[21,82,83],{},"Stroomdistributie: de 4-makes-3-architectuur",[10,85,86],{},"De extreme vermogensdichtheid van GB300 NVL72-systemen, 150 kW piek per rack, vereist een energiedistributiestrategie op bedrijfsniveau.",[10,88,89],{},"We implementeerden de 4-makes-3-busbarconfiguratie uit het NVIDIA-referentieontwerp. Vier onafhankelijke busbars van 1 MW leveren stroom volgens een redundant $N=3+1$-model.",[10,91,92],{},[21,93,94],{},"Deze architectuur levert:",[45,96,97,103,109],{},[48,98,99,102],{},[21,100,101],{},"Volledig gelijktijdig onderhoud:"," elke busbar kan voor onderhoud offline worden gehaald zonder de dienstverlening te onderbreken.",[48,104,105,108],{},[21,106,107],{},"Echte redundantie:"," drie busbars kunnen samen de volledige operationele belasting dragen.",[48,110,111,114],{},[21,112,113],{},"Totaal vermogensbereik:"," ongeveer 2,8 MW (variabel op basis van opslagconfiguratievereisten).",[10,116,117],{},"Daarnaast beschikt de GB300 NVL72 over geavanceerde voedingseenheden met energieopslag. Die kunnen de piekvraag op het elektriciteitsnet met maximaal 30% verlagen door vermogenspieken van AI-workloads af te vlakken.",[10,119,120],{},[74,121],{"alt":76,"src":122},"\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002F4-makes-3_trans-1.webp",[17,124,126],{"id":125},"geavanceerd-thermisch-beheer-hybride-koeling",[21,127,128],{},"Geavanceerd thermisch beheer: hybride koeling",[10,130,131],{},"Direct Liquid Cooling kan tot 80% van de warmte van een SU afvoeren. Voor de resterende warmte zijn andere afvoermethoden nodig. Daarom ontwikkelden we een hybride thermisch systeem dat geoptimaliseerd is voor rekenefficiëntie en modulariteit.",[10,133,134],{},[74,135],{"alt":76,"src":136},"\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fmechanicalcooling_diagram_trans.webp",[10,138,139],{},[21,140,141],{},"Primaire koelingstrategie",[45,143,144,160],{},[48,145,146,149],{},[21,147,148],{},"GB300 NVL72-computerracks:",[45,150,151,154,157],{},[48,152,153],{},"80% thermische belasting beheerd via Direct Liquid Cooling (DLC).",[48,155,156],{},"20% restwarmte opgevangen via achterdeurwarmtewisselaars (RDHx) of CRAH-units.",[48,158,159],{},"DLC voert warmte rechtstreeks af via cold plates op de warmteproducerende componenten, aangesloten op leidingen met koelvloeistof.",[48,161,162,165],{},[21,163,164],{},"Hulpinfrastructuur (netwerk\u002Fbeheer\u002Fopslag):",[45,166,167,170],{},[48,168,169],{},"Standaard volledig gekoeld via RDHx.",[48,171,172],{},"Een vlot migratiepad naar DLC wanneer de thermische vereisten toenemen.",[10,174,175],{},[21,176,177],{},"De kritieke lussen",[45,179,180,186],{},[48,181,182,185],{},[21,183,184],{},"Secundaire lus (CDU's):"," in-rack coolant distribution units bewaken continu debiet, temperatuur en drukverschillen en bieden uitgebreide lekdetectie.",[48,187,188,191],{},[21,189,190],{},"Primaire lus (warmteafvoer):"," het primaire thermische systeem bestaat uit drie chillers\u002Fdry coolers van 1.125 kW in een $N=2+1$-configuratie. Zo vermijden we het waterverbruik van traditionele koelmethoden.",[17,193,195],{"id":194},"structureel-en-milieuontwerp",[21,196,197],{},"Structureel en milieuontwerp",[10,199,200],{},"De kern-IT-omgeving maakt gebruik van open stalen frameconstructiesystemen, waardoor een veilige en gecontroleerde omgeving wordt gecreëerd met:",[45,202,203,209,215],{},[48,204,205,208],{},[21,206,207],{},"Brandbestrijding:"," systemen op basis van argon, een schoon, niet-geleidend blusmiddel dat zuurstof verdringt.",[48,210,211,214],{},[21,212,213],{},"Luchtzuivering:"," HEPA-filtratie met behoud van ISO Klasse 8 cleanroom-normen.",[48,216,217,220],{},[21,218,219],{},"Toegangscontrole:"," Biometrische authenticatie en mantrap-configuraties.",[17,222,224],{"id":223},"schaalbaarheid-ontworpen-voor-groei",[21,225,226],{},"Schaalbaarheid: ontworpen voor groei",[10,228,229],{},"Een van de meest cruciale voordelen van onze modulaire aanpak is naadloze schaalbaarheid die precies aansluit bij de referentiearchitectuur van NVIDIA. Onze modulaire datacenters kunnen worden ingezet om deze gestandaardiseerde schaalvergrotingen te ondersteunen:",[10,231,232],{},[21,233,234],{},"Deploymentconfiguraties:",[45,236,237,243,249,255,261,267,273,279],{},[48,238,239,242],{},[21,240,241],{},"1 SU"," (16 racks, 1.152 GPU's): basisdeployment van één schaalbare unit",[48,244,245,248],{},[21,246,247],{},"2 SU"," (32 racks, 2.304 GPU's): eerste productieschaal",[48,250,251,254],{},[21,252,253],{},"4 SU"," (64 racks, 4.608 GPU's): cluster op ondernemingsschaal",[48,256,257,260],{},[21,258,259],{},"7 SU"," (112 racks, 8.064 GPU's): grootschalige trainingsfaciliteit",[48,262,263,266],{},[21,264,265],{},"8 SU"," (128 racks, 9.216 GPU's): AI-fabriek met hoge densiteit",[48,268,269,272],{},[21,270,271],{},"9 SU"," (144 racks, 10.368 GPU's): AI-fabriek met hoge densiteit",[48,274,275,278],{},[21,276,277],{},"18 SU"," (288 racks, 20.736 GPU's): grootschalige deployment",[48,280,281,284],{},[21,282,283],{},"36 SU"," (576 racks, 41.472 GPU's): supercomputercluster op frontier-schaal",[10,286,287],{},[74,288],{"alt":76,"src":289},"\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002F7su_cluster.webp",[17,291,293],{"id":292},"conclusie-engineering-voor-morgen",[21,294,295],{},"Conclusie: Engineering voor morgen",[10,297,298],{},"De GB300 NVL72 is een rack-scale unit van ongeveer 140 kW en levert 1,1 exaFLOPS aan FP4-rekenkracht. Die compute-densiteit verandert de eisen aan datacenters fundamenteel.",[10,300,301],{},"Onze deploymentcyclus van vier maanden brengt geavanceerde thermische engineering, redundante stroomarchitecturen en een doelgericht structureel ontwerp samen. De toekomst van AI-infrastructuur is modulair, redundant en geoptimaliseerd voor extreme densiteit, en kan bovendien binnen zestien weken operationeel zijn.",{"title":76,"searchDepth":303,"depth":303,"links":304},2,[305,307,308,309,310,311,312],{"id":19,"depth":306,"text":23},3,{"id":32,"depth":306,"text":35},{"id":80,"depth":306,"text":83},{"id":125,"depth":306,"text":128},{"id":194,"depth":306,"text":197},{"id":223,"depth":306,"text":226},{"id":292,"depth":306,"text":295},[314,315,316,317,318,319,320,321,322,323,324,325,326],"Alle","Modulaire DC","Niet gecategoriseerd","150 kW-rack","AI-infrastructuur","DLC","Hoge dichtheid","Vloeistofkoeling","Modulair datacenter","NVIDIA Blackwell Ultra","NVIDIA GB300","NVL72","Snelle implementatie","2025-11-20T14:10:19","Ontdek een modulair datacenterontwerp voor NVIDIA GB300 NVL72, met redundante voeding, hybride koeling en een bouwplanning van vier maanden.","md","De motor voor de AI-race bouwen: in vier maanden klaar voor NVIDIA GB300 NVL72","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fsuperpodmodflexfrontimage.webp",{},true,"https:\u002F\u002Feliovp.com\u002Fbuilding-the-engine-for-the-ai-race-the-4-month-path-to-nvidia-gb300-nvl72-power\u002F","\u002Fblog\u002Fbuilding-the-engine-for-the-ai-race-the-4-month-path-to-nvidia-gb300-nvl72-power",{"title":5,"description":328},"building-the-engine-for-the-ai-race-the-4-month-path-to-nvidia-gb300-nvl72-power","blog\u002Fbuilding-the-engine-for-the-ai-race-the-4-month-path-to-nvidia-gb300-nvl72-power",null,"rLyfv2YiltzKnpPdtjmaeiE-LkT-FCudW9o_X2EeQ7Y",[342,358,368,380,391,404,413,425,457,467,489,507,526,545,547,564,576,592,607,619,628,636,651,663,674,685,695,708,718,731,742,752,763,772,784,795,804],{"path":343,"title":344,"description":345,"date":346,"slug":347,"image":348,"originalUrl":349,"categories":350},"\u002Fblog\u002Fpaiton-qwen38-mxfp4-dflash2-r9700","Qwen3.8: 400,7 tok\u002Fs op één R9700 | Paiton","Qwen3.8 op één R9700: 400,7 tok\u002Fs met ROCm 10 en vLLM 0.29, plus publieke 200K\u002F220K-chatprofielen. Benchmarks, beperkingen en startopdrachten.","2026-09-16T07:30:00Z","paiton-qwen38-mxfp4-dflash2-r9700","\u002Fasset\u002Fimages\u002Fblog\u002Fpaiton-qwen38-mxfp4\u002Fupdate-2026-09-19\u002Fhero.webp","https:\u002F\u002Feliovp.com\u002Fblog\u002Fpaiton-qwen38-mxfp4-dflash2-r9700",[351,352,353,354,355,356,357],"Paiton","AMD Radeon","Lokale AI","AI-inferentie","Inferentie-optimalisatie","Grote taalmodellen","vLLM",{"path":359,"title":360,"description":361,"date":362,"slug":363,"image":364,"originalUrl":365,"categories":366},"\u002Fblog\u002Fpaiton-qwen38-neo-gguf-vllm-r9700","Qwen3.8 GGUF in vLLM: sneller antwoord op één Radeon","Draai de originele NEO CODER MAX GGUF met Paiton in vLLM op een R9700. Bekijk de gemeten responstijden, beeldinvoer en lokale installatie.","2026-09-14T07:30:00Z","paiton-qwen38-neo-gguf-vllm-r9700","\u002Fasset\u002Fimages\u002Fblog\u002Fpaiton-neo-gguf\u002F00-hero-neo-gguf-r9700.webp","https:\u002F\u002Feliovp.com\u002Fblog\u002Fpaiton-qwen38-neo-gguf-vllm-r9700",[351,352,353,367,357],"GGUF",{"path":369,"title":370,"description":371,"date":372,"slug":373,"image":374,"originalUrl":375,"categories":376},"\u002Fblog\u002Fpaiton-minimax-h3-radeon-ai-pro-r9700","MiniMax H3 op Radeon: 15 seconden video met stereogeluid","Paiton genereert lokaal 15 seconden MiniMax H3-video met stereogeluid op één Radeon AI PRO R9700 in 5 min 33 s, met 16,7% minder wachttijd dan stock.","2026-09-09T07:30:00Z","paiton-minimax-h3-radeon-ai-pro-r9700","\u002Fasset\u002Fimages\u002Fblog\u002Fpaiton-minimax-h3\u002F00-featured-minimax-h3-r9700.webp","https:\u002F\u002Feliovp.com\u002Fblog\u002Fpaiton-minimax-h3-radeon-ai-pro-r9700",[351,352,353,377,378,379],"Videogeneratie","MiniMax H3","ComfyUI",{"path":381,"title":382,"description":383,"date":384,"slug":385,"image":386,"originalUrl":387,"categories":388},"\u002Fblog\u002Fpaiton-flux2-klein-radeon-ai-pro-r9700","Lokale FLUX.2 klein op Radeon AI PRO R9700: sneller beelden genereren met minder VRAM","Paiton genereert FLUX.2 klein-beelden van 1024 × 1024 in 1,054 seconden op een R9700, met 16,2% minder generatietijd en 33,4% minder piekallocatie in Torch.","2026-09-07T09:00:00","paiton-flux2-klein-radeon-ai-pro-r9700","\u002Fasset\u002Fimages\u002Fblog\u002Fpaiton-flux2-klein\u002Ffox-paiton.webp","https:\u002F\u002Feliovp.com\u002Fblog\u002Fpaiton-flux2-klein-radeon-ai-pro-r9700",[351,352,353,389,390,379],"Beeldgeneratie","FLUX",{"path":392,"title":393,"description":394,"date":395,"slug":396,"image":397,"originalUrl":398,"categories":399},"\u002Fblog\u002Fpaiton-ornith15-radeon-ai-pro-r9700","Ornith 1.5 haalt 44,6 tok\u002Fs op één Radeon AI PRO R9700","Paiton draait Ornith 1.5 35B A3B op één Radeon AI PRO R9700 met 44,63 outputtokens per seconde, 27% sneller en met 21,3% lagere gemodelleerde kosten.","2026-09-05T09:00:00","paiton-ornith15-radeon-ai-pro-r9700","\u002Fasset\u002Fimages\u002Fblog\u002Fpaiton-ornith15\u002F00-featured-ornith15-r9700.webp","https:\u002F\u002Feliovp.com\u002Fblog\u002Fpaiton-ornith15-radeon-ai-pro-r9700",[351,400,352,354,401,402,355,356,357,403],"Kunstmatige intelligentie","GPU-prestaties","Inferentielatentie","Kostenefficiëntie",{"path":405,"title":406,"description":407,"date":408,"slug":409,"image":410,"originalUrl":411,"categories":412},"\u002Fblog\u002Fpaiton-qwen38-radeon-ai-pro-r9700","Paiton: 21% meer Qwen3.8-doorvoer op Radeon AI PRO R9700","Paiton draait AMD's Qwen3.8 27B op één Radeon AI PRO R9700 met 39,77 outputtokens per seconde. Dat levert 21% meer throughput en 17,4% lagere gemodelleerde kosten op.","2026-09-04T09:00:00","paiton-qwen38-radeon-ai-pro-r9700","\u002Fasset\u002Fimages\u002Fblog\u002Fpaiton-r9700\u002F00-featured-paiton-r9700.webp","https:\u002F\u002Feliovp.com\u002Fblog\u002Fpaiton-qwen38-radeon-ai-pro-r9700",[351,400,352,354,401,402,355,356,357,403],{"path":414,"title":415,"description":416,"date":417,"slug":418,"image":419,"originalUrl":339,"categories":420},"\u002Fblog\u002Fai-data-center-power-requirements-gpu-per-megawatt","Stroomvereisten voor AI-datacenters: de GPU-per-MW-illusie","Waarom verschillen GPU-aantallen per megawatt? Lees hoe PUE, piekbelasting, opslag, netwerken en koeling de inzetbare AI-capaciteit bepalen.","2026-07-27T23:52:00","ai-data-center-power-requirements-gpu-per-megawatt","\u002Fasset\u002Fimages\u002Fblog\u002Fai-data-center-power-requirements-gpu-per-megawatt\u002Fgpu-per-megawatt-illusion.webp",[314,318,421,422,423,424],"Datacenters","ModFlex","HPC","AMD Helios",{"path":426,"title":427,"description":428,"date":429,"slug":430,"image":431,"originalUrl":432,"categories":433},"\u002Fblog\u002Fpaiton-returns-to-its-diffusion-roots-optimizing-wan2-2-t2v-a14b-on-amd-mi355x","Wan2.2-videogeneratie: Paiton op AMD MI355X","Vergelijk Wan2.2-videogeneratie op AMD MI355X met Paiton en NVIDIA B200 via Diffusers. Lees hoe we diffusiemodellen optimaliseren.","2026-06-10T14:04:04","paiton-returns-to-its-diffusion-roots-optimizing-wan2-2-t2v-a14b-on-amd-mi355x","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fpaitonwan2.webp","https:\u002F\u002Feliovp.com\u002Fpaiton-returns-to-its-diffusion-roots-optimizing-wan2-2-t2v-a14b-on-amd-mi355x\u002F",[314,400,351,434,435,436,437,438,439,440,441,442,443,444,445,446,447,448,449,450,351,451,452,453,454,455,456],"14B","AMD","B200","Benchmarks","Blackwell","Compute","Diffusie","Eliovp","Generatieve AI","GPU","Hardware","Inferentie","Instinct","MI355x","NVIDIA","On-premises","Optimalisatie","Soevereine AI","T2V","Tekst-naar-video","Tuning","Video-generatie","Wan2.2",{"path":458,"title":459,"description":460,"date":461,"slug":462,"image":463,"originalUrl":464,"categories":465},"\u002Fblog\u002Ffrom-the-attic-to-the-front-page-eliovp-recognized-as-a-pioneer-in-chip-optimization-data-center-infrastructure","ElioVP in De Tijd: chipoptimalisatie en datacenters","Lees hoe De Tijd ElioVP belicht, van de oorsprong in chipoptimalisatie tot het werk aan modulaire datacenters en koeling voor hoge vermogensdichtheid.","2026-02-10T20:48:12","from-the-attic-to-the-front-page-eliovp-recognized-as-a-pioneer-in-chip-optimization-data-center-infrastructure","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fphysicalnewspaper.webp","https:\u002F\u002Feliovp.com\u002Ffrom-the-attic-to-the-front-page-eliovp-recognized-as-a-pioneer-in-chip-optimization-data-center-infrastructure\u002F",[314,400,315,316,435,466,315,448],"De Tijd",{"path":468,"title":469,"description":470,"date":471,"slug":472,"image":473,"originalUrl":474,"categories":475},"\u002Fblog\u002Fprivacy-is-geen-it-probleem-meer-het-is-een-strategische-prioriteit","AI en privacy: een strategische prioriteit in de Benelux","Privacyrisico's van generatieve AI, vertrouwen, dataopslag en governance. Waarom bedrijven in de Benelux veilige AI strategisch moeten benaderen.","2026-01-29T13:51:11","privacy-is-geen-it-probleem-meer-het-is-een-strategische-prioriteit","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fheaderimage.webp","https:\u002F\u002Feliovp.com\u002Fprivacy-is-geen-it-probleem-meer-het-is-een-strategische-prioriteit\u002F",[314,400,476,316,477,478,479,480,481,482,483,484,441,485,442,486,487,488],"Trending","AI Act","Antropomorfisme","AVG","Benelux","ChatGPT","Cyberbeveiliging","Databeheer","Gegevensbeveiliging","GDPR","Microsoft Copilot","Privacy","Shadow AI",{"path":490,"title":491,"description":492,"date":493,"slug":494,"image":495,"originalUrl":496,"categories":497},"\u002Fblog\u002Fitsme-bij-ons-is-het-its-not-me-en-dit-is-waarom","Waarom wij itsme niet gebruiken: privacy en soevereiniteit","Waarom ElioVP itsme niet gebruikt: onze afwegingen rond identiteitsmetadata, cloudafhankelijkheid, privacy en datasoevereiniteit.","2025-11-27T09:32:14","itsme-bij-ons-is-het-its-not-me-en-dit-is-waarom","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Ffrontimage.webp","https:\u002F\u002Feliovp.com\u002Fitsme-bij-ons-is-het-its-not-me-en-dit-is-waarom\u002F",[314,498,499,500,482,501,502,503,485,504,505,506,487],"AWS","Belgian Mobile ID","CLOUD Act","Datasoevereiniteit","Digitale identiteit","eIDAS","itsme","Liberty Global","MyGov.be",{"path":508,"title":509,"description":510,"date":511,"slug":512,"image":513,"originalUrl":514,"categories":515},"\u002Fblog\u002Ffield-report-the-reality-of-building-agentic-ai-in-2025","Praktijkrapport: de realiteit van Agentic AI bouwen in 2025","Praktijklessen over lokale AI-agents in 2025 gaan in op workflowontwerp, observability, modeltraining, hallucinaties en GPU-geheugenlimieten.","2025-11-25T14:03:39","field-report-the-reality-of-building-agentic-ai-in-2025","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Ffieldreport.webp","https:\u002F\u002Feliovp.com\u002Ffield-report-the-reality-of-building-agentic-ai-in-2025\u002F",[314,400,516,476,517,518,519,520,521,522,523,524,451,525],"Oplossingen","Agentic AI","AI-techniek","AI-strategie","Autonome agenten","Bedrijfs-AI","Lokale LLM","Modelverfijning","AI op locatie","VRAM-optimalisatie",{"path":527,"title":528,"description":529,"date":530,"slug":531,"image":532,"originalUrl":533,"categories":534},"\u002Fblog\u002Fthe-synthetic-unicorn-bubble","De synthetische unicornzeepbel","Een analyse van investeringsrisico’s bij AI-neoclouds: circulaire financiering, infrastructuurclaims, contractvoorwaarden en due diligence.","2025-11-24T19:22:28","the-synthetic-unicorn-bubble","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fsyntheticunicorn.webp","https:\u002F\u002Feliovp.com\u002Fthe-synthetic-unicorn-bubble\u002F",[314,400,476,535,536,537,538,539,540,541,542,543,544],"AI Infrastructure","AI Neocloud","Circulaire financiering","GPU Cloud","Beleggingsrisico's","Opstartwaardering","Synthetische bubbel","Technische analyse","Vaporware","Durfkapitaal",{"path":335,"title":5,"description":328,"date":327,"slug":337,"image":331,"originalUrl":334,"categories":546},[314,315,316,317,318,319,320,321,322,323,324,325,326],{"path":548,"title":549,"description":550,"date":551,"slug":552,"image":553,"originalUrl":554,"categories":555},"\u002Fblog\u002Fwhy-cuda-translation-wont-unlock-amds-real-potential","CUDA-vertaling versus AMD-gerichte optimalisatie","Waarom CUDA-compatibiliteit niet hetzelfde is als AMD-prestaties: over ROCm, HIP, kerneloptimalisatie en hardwaregerichte afstemming.","2025-11-12T14:48:37","why-cuda-translation-wont-unlock-amds-real-potential","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fchatgpt-image-nov-11-2025-09_16_10-pm-1.webp","https:\u002F\u002Feliovp.com\u002Fwhy-cuda-translation-wont-unlock-amds-real-potential\u002F",[314,400,351,316,556,400,557,558,559,560,561,562,351,563],"AMD MI300X","CUDA-vertaling","FP8","GPU-optimalisatie","High-performance computing","HIP","Kerneltuning","ROCm",{"path":565,"title":566,"description":567,"date":568,"slug":569,"image":570,"originalUrl":571,"categories":572},"\u002Fblog\u002Fpaiton-the-simplest-way-to-supercharge-ai-inference","Paiton: snellere AI-inferentie in uw bestaande stack","Lees hoe Paiton aansluit op bestaande inferentiestacks, met AMD MI300X-benchmarks en vergelijkingen van prestaties per dollar.","2025-11-11T10:31:22","paiton-the-simplest-way-to-supercharge-ai-inference","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fpaiton-powaaah.webp","https:\u002F\u002Feliovp.com\u002Fpaiton-the-simplest-way-to-supercharge-ai-inference\u002F",[314,400,351,354,573,556,403,574,355,562,351,575,357],"AMD Instinct","Hoge throughput","SGLang",{"path":577,"title":578,"description":579,"date":580,"slug":581,"image":582,"originalUrl":583,"categories":584},"\u002Fblog\u002Fstop-overpaying-paiton-mi300x-moe-beats-h200-b200-on-1m-tokens","Paiton MoE-benchmarks: MI300X versus H200 en B200","Vergelijk Qwen3-30B-A3B MoE-benchmarks van MI300X met Paiton, H200 en B200: throughput en kosten per miljoen tokens.","2025-09-26T13:36:18","stop-overpaying-paiton-mi300x-moe-beats-h200-b200-on-1m-tokens","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fhulkvshulkpaitonwins.webp","https:\u002F\u002Feliovp.com\u002Fstop-overpaying-paiton-mi300x-moe-beats-h200-b200-on-1m-tokens\u002F",[314,400,351,585,556,586,355,587,588,589,590,351,591],"AI-benchmarks","Kosten per token","Mixture of Experts","MoE","NVIDIA B200","NVIDIA H200","Qwen3",{"path":593,"title":594,"description":595,"date":596,"slug":597,"image":598,"originalUrl":599,"categories":600},"\u002Fblog\u002Fagentic-ai-but-make-it-local-from-inbox-to-insight-to-action-en","Lokale Agentic AI: van inbox naar actie","Lokale AI-agents zetten e-mails, documenten en beelden om in tickets, rapporten en acties, met modellen op maat van uw gegevens en systemen.","2025-09-16T13:09:00","agentic-ai-but-make-it-local-from-inbox-to-insight-to-action-en","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Ffrontfotoblog.webp","https:\u002F\u002Feliovp.com\u002Fagentic-ai-but-make-it-local-from-inbox-to-insight-to-action-en\u002F",[314,400,516,316,517,601,602,603,604,522,524,451,605,606],"Schadedetectie","Documentverwerking","E-mailautomatisering","Factuurextractie","Ticketautomatisering","Workflowautomatisering",{"path":608,"title":609,"description":610,"date":611,"slug":612,"image":613,"originalUrl":614,"categories":615},"\u002Fblog\u002Fmi300x-fp8-data-parallel-benchmarks-8-64-gpus-h200-left-behind-b200-within-reach","MI300X FP8-benchmarks: GPU-partitionering met Paiton","Bekijk hoe Paiton presteert met Llama 3.1 8B FP8 op gepartitioneerde MI300X-GPU's, vergeleken met NVIDIA H200 en B200.","2025-07-31T13:32:57","mi300x-fp8-data-parallel-benchmarks-8-64-gpus-h200-left-behind-b200-within-reach","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fimage-2-1.webp","https:\u002F\u002Feliovp.com\u002Fmi300x-fp8-data%e2%80%91parallel-benchmarks-8-64-gpus-h200-left-behind-b200-within-reach\u002F",[314,400,351,316,616,435,436,617,618,447,448,351,357],"AI","H200","MI300X",{"path":620,"title":621,"description":622,"date":623,"slug":624,"image":625,"originalUrl":626,"categories":627},"\u002Fblog\u002Fapplicable-ai-for-businesses","Toepasbare AI voor bedrijven","Ontdek hoe ElioVP lokale AI voor bedrijfsprocessen bouwt, met modeltraining op maat en automatische schadedetectie voor de logistiek.","2025-07-09T21:35:30","applicable-ai-for-businesses","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fscherm_afbeelding-2025-07-09-om-23.30.17.webp","https:\u002F\u002Feliovp.com\u002Fapplicable-ai-for-businesses\u002F",[314,400,516,517,601,602,603,604,522,524,451,605,606],{"path":629,"title":630,"description":631,"date":632,"slug":633,"image":76,"originalUrl":634,"categories":635},"\u002Fblog\u002Fintroducing-paitons-free-evaluation-models","Maak kennis met de gratis evaluatiemodellen van Paiton","Test Paiton met gratis evaluatiemodellen voor AMD-GPU's. Vergelijk de prestaties voor tekst, beeldanalyse en beeldgeneratie met uw eigen workloads.","2025-07-07T11:26:13","introducing-paitons-free-evaluation-models","https:\u002F\u002Feliovp.com\u002Fintroducing-paitons-free-evaluation-models\u002F",[314,400,351],{"path":637,"title":638,"description":639,"date":640,"slug":641,"image":642,"originalUrl":643,"categories":644},"\u002Fblog\u002Fpaiton-dramatically-faster-startup-and-performance-for-llama-3-1-405b","Llama 3.1 405B: sneller starten met Paiton op MI300X","Bekijk Paiton-benchmarks voor Llama 3.1 405B op acht AMD MI300X-GPU's, met opstarttijd, tensorparallelisme, throughput en latency.","2025-06-12T20:15:23","paiton-dramatically-faster-startup-and-performance-for-llama-3-1-405b","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fservingscreenshot.webp","https:\u002F\u002Feliovp.com\u002Fpaiton-dramatically-faster-startup-and-performance-for-llama-3-1-405b\u002F",[314,400,351,316,354,556,645,558,646,647,648,351,649,650],"Koude start","Grafiekcompilatie","Llama 3.1 405B","LLM-optimalisatie","Opstartlatentie","Tensor-parallellisme",{"path":652,"title":653,"description":654,"date":655,"slug":656,"image":657,"originalUrl":658,"categories":659},"\u002Fblog\u002Fpaiton-fp8-beats-nvidias-h200-on-amds-mi300x","Paiton FP8 verslaat NVIDIA's H200 op AMD's MI300X","Vergelijk Paiton op AMD MI300X met NVIDIA H200 voor Llama 3.1 70B FP8: throughput, wachttijd tot het eerste token en latency bij diverse batchgroottes.","2025-06-08T19:12:40","paiton-fp8-beats-nvidias-h200-on-amds-mi300x","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fblognewfp8.webp","https:\u002F\u002Feliovp.com\u002Fpaiton-fp8-beats-nvidias-h200-on-amds-mi300x\u002F",[314,400,351,316,556,660,521,442,401,402,356,647,661,662],"Koude startoptimalisatie","Model serving","vLLM-optimalisatie",{"path":664,"title":665,"description":666,"date":667,"slug":668,"image":669,"originalUrl":670,"categories":671},"\u002Fblog\u002Fmi300x-vs-h200-vs-rx-7900-xtx-vs-tenstorrent-n300s-with-vllm","MI300X, H200, RX 7900 XTX en n300s: vLLM-benchmarks","Vergelijk MI300X, H200, RX 7900 XTX en Tenstorrent n300s met vLLM: throughput, gemodelleerde tokenkosten en hardwarebeperkingen voor Llama 3 8B.","2025-05-09T14:03:58","mi300x-vs-h200-vs-rx-7900-xtx-vs-tenstorrent-n300s-with-vllm","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fcomparisontenstor.webp","https:\u002F\u002Feliovp.com\u002Fmi300x-vs-h200-vs-rx-7900-xtx-vs-tenstorrent-n300s-with-vllm\u002F",[314,400,351,516,316,435,618,448,672,673],"RX7900XTX","tenstorrent",{"path":675,"title":676,"description":677,"date":678,"slug":679,"image":680,"originalUrl":681,"categories":682},"\u002Fblog\u002Fclusterpl-empowering-gpu-cluster-investors-with-real-world-financial-insights","ClusterP&L: financiële modellen voor GPU-clusters","Ontdek hoe ClusterP&L kosten, rendement en investeringsscenario's voor GPU-clusters modelleert, met risicoanalyses en exporteerbare rapporten.","2025-05-03T10:52:22","clusterpl-empowering-gpu-cluster-investors-with-real-world-financial-insights","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fcomparisonscenarios.webp","https:\u002F\u002Feliovp.com\u002Fclusterpl-empowering-gpu-cluster-investors-with-real-world-financial-insights\u002F",[314,400,315,516,436,617,683,448,684],"MI325X","P&L-calculator",{"path":686,"title":687,"description":688,"date":689,"slug":690,"image":691,"originalUrl":692,"categories":693},"\u002Fblog\u002Fcranking-out-faster-tokens-for-fewer-dollars-amd-mi300x-vs-nvidia-h200","AMD MI300X versus NVIDIA H200: Qwen3-32B met Paiton","Vergelijk Qwen3-32B-benchmarks op AMD MI300X met Paiton en NVIDIA H200, met resultaten voor throughput, latency en hardwarekosten.","2025-05-02T21:10:30","cranking-out-faster-tokens-for-fewer-dollars-amd-mi300x-vs-nvidia-h200","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002F3ac59a73-2466-4422-b7e5-ef2e4a8ca58e.webp","https:\u002F\u002Feliovp.com\u002Fcranking-out-faster-tokens-for-fewer-dollars-amd-mi300x-vs-nvidia-h200\u002F",[314,400,351,616,435,617,694,448,351,357],"MI300",{"path":696,"title":697,"description":698,"date":699,"slug":700,"image":701,"originalUrl":702,"categories":703},"\u002Fblog\u002Fpower-meets-precision-high-density-modular-data-center-for-nvidia-nvl-deployments-1-2-mw","Modulaire datacenters voor NVIDIA NVL: 1 tot 2 MW","Ontdek modulaire datacenterontwerpen voor NVIDIA NVL-systemen, met aandacht voor vermogen, vloeistofkoeling, redundantie en uitrolplanning.","2025-05-02T14:09:59","power-meets-precision-high-density-modular-data-center-for-nvidia-nvl-deployments-1-2-mw","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Feliovp_critical-1mw-pod_rev-2_transparent.webp","https:\u002F\u002Feliovp.com\u002Fpower-meets-precision-high-density-modular-data-center-for-nvidia-nvl-deployments-1-2-mw\u002F",[314,315,704,535,320,423,321,322,705,706,325,707],"1-2MW datacenter","NVIDIA Blackwell","NVIDIA NVL","Precisiekoeling",{"path":709,"title":710,"description":711,"date":712,"slug":713,"image":714,"originalUrl":715,"categories":716},"\u002Fblog\u002Fexamining-ai-agents-in-the-medical-field-ai-that-speaks-dicom","AI-agents in de medische wereld: AI die DICOM spreekt","Ontdek een lokale AI-agent die DICOM-gegevens opzoekt en bekijk tests van beeldmodellen met geanonimiseerde medische beelden.","2025-04-11T14:45:48","examining-ai-agents-in-the-medical-field-ai-that-speaks-dicom","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fhealthcareblog-1.webp","https:\u002F\u002Feliovp.com\u002Fexamining-ai-agents-in-the-medical-field-ai-that-speaks-dicom\u002F",[314,400,516,316,616,435,717],"Zorg",{"path":719,"title":720,"description":721,"date":722,"slug":723,"image":724,"originalUrl":725,"categories":726},"\u002Fblog\u002Feliovp-bv-your-trusted-partner-for-supply-chain-resilience-amidst-new-u-s-tariffs","Amerikaanse invoerheffingen en AI-leveringszekerheid in 2025","Lees ElioVP's visie uit april 2025 op Amerikaanse invoerheffingen en leveringszekerheid voor AI-servers, HPC-systemen en modulaire datacenters.","2025-04-04T10:01:27","eliovp-bv-your-trusted-partner-for-supply-chain-resilience-amidst-new-u-s-tariffs","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Ftariffsshipping.webp","https:\u002F\u002Feliovp.com\u002Feliovp-bv-your-trusted-partner-for-supply-chain-resilience-amidst-new-u-s-tariffs\u002F",[314,476,616,435,727,728,729,730],"Invoer","Taiwan","Invoerheffingen","Trump",{"path":732,"title":733,"description":734,"date":735,"slug":736,"image":737,"originalUrl":738,"categories":739},"\u002Fblog\u002Fwhy-ai-agents-are-the-future","Waarom AI-agenten de toekomst zijn","Ontdek AI-agents voor ERP, CRM, financiën en klantondersteuning, met praktijkvoorbeelden en een traject van procesanalyse tot pilot en uitrol.","2025-03-23T22:06:59","why-ai-agents-are-the-future","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Ferp2.jpeg","https:\u002F\u002Feliovp.com\u002Fwhy-ai-agents-are-the-future\u002F",[314,400,516,616,740,741],"AI-agenten","ERP",{"path":743,"title":744,"description":745,"date":746,"slug":747,"image":748,"originalUrl":749,"categories":750},"\u002Fblog\u002Fthe-rise-of-open-source-ai-model-optimization","De opkomst van open-source AI-modeloptimalisatie","Verken trends in opensource-AI-optimalisatie: kwantisatie, mixture-of-experts-modellen, hardwaregerichte afstemming, RAG en edge-AI.","2025-03-22T20:59:23","the-rise-of-open-source-ai-model-optimization","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Friseofopensource.jpeg","https:\u002F\u002Feliovp.com\u002Fthe-rise-of-open-source-ai-model-optimization\u002F",[314,400,476,751,435,443,448],"AI-nieuws",{"path":753,"title":754,"description":755,"date":756,"slug":757,"image":758,"originalUrl":759,"categories":760},"\u002Fblog\u002Fintroducing-our-benchmarking-tool-powered-by-dstack","Maak kennis met onze benchmarktool, gebouwd op dstack","Ontdek onze benchmarktool met dstack: herhaalbare vLLM-tests, automatische parameterreeksen en prestatierapporten voor lokale GPU's en de cloud.","2025-03-20T14:21:59","introducing-our-benchmarking-tool-powered-by-dstack","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fbenchmarktool.jpeg","https:\u002F\u002Feliovp.com\u002Fintroducing-our-benchmarking-tool-powered-by-dstack\u002F",[314,400,351,616,435,761,762,618,351],"benchmark","LLM",{"path":764,"title":765,"description":766,"date":767,"slug":768,"image":769,"originalUrl":770,"categories":771},"\u002Fblog\u002Foptimizing-qwq-32b-by-qwen-amd-mi300x-vs-nvidia-h200","QwQ-32B optimaliseren (door Qwen): AMD MI300X versus NVIDIA H200","Vergelijk throughput en latency van QwQ-32B op AMD MI300X met Paiton en NVIDIA H200, bij kleine batches en meer gelijktijdige aanvragen.","2025-03-19T21:41:44","optimizing-qwq-32b-by-qwen-amd-mi300x-vs-nvidia-h200","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fpaiton4.jpeg","https:\u002F\u002Feliovp.com\u002Foptimizing-qwq-32b-by-qwen-amd-mi300x-vs-nvidia-h200\u002F",[314,400,351],{"path":773,"title":774,"description":775,"date":776,"slug":777,"image":778,"originalUrl":779,"categories":780},"\u002Fblog\u002Feliovp-featured-on-amd-tech-talk-podcast","Eliovp te gast in de AMD Tech Talk-podcast","Beluister Elio Van Puyvelde en Jim Greene in de AMD Tech Talk-podcast over het ontstaan van ElioVP en de hardware- en softwarediensten voor AI.","2025-03-19T07:53:39","eliovp-featured-on-amd-tech-talk-podcast","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Ftechtalkjimgreene.jpeg","https:\u002F\u002Feliovp.com\u002Feliovp-featured-on-amd-tech-talk-podcast\u002F",[314,435,781,782,783],"Jim Greene","Podcast","Tech Talk",{"path":785,"title":786,"description":787,"date":788,"slug":789,"image":790,"originalUrl":791,"categories":792},"\u002Fblog\u002Ffurther-optimizing-amd-powered-inference-with-paiton","AMD-inferentie verder optimaliseren met Paiton","Bekijk Paiton-benchmarks voor DeepSeek R1 Distill Llama 8B op AMD MI300X, gericht op throughput en latency bij kleinere batchgroottes.","2025-03-13T06:18:30","further-optimizing-amd-powered-inference-with-paiton","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fpaitonpost3.webp","https:\u002F\u002Feliovp.com\u002Ffurther-optimizing-amd-powered-inference-with-paiton\u002F",[314,400,351,435,793,794,617,618,683,351,357],"DeepSeek","H100",{"path":796,"title":797,"description":798,"date":799,"slug":800,"image":801,"originalUrl":802,"categories":803},"\u002Fblog\u002Fa-first-look-at-paiton-in-action-deepseek-r1-distill-llama-3-1-8b","Paiton-benchmarks: DeepSeek R1 Distill Llama 3.1 8B","Vergelijk standaard- en Paiton-versies van DeepSeek R1 Distill Llama 3.1 8B op AMD MI300X, met benchmarks voor throughput en latency.","2025-01-31T09:11:02","a-first-look-at-paiton-in-action-deepseek-r1-distill-llama-3-1-8b","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fpaitonpost2.webp","https:\u002F\u002Feliovp.com\u002Fa-first-look-at-paiton-in-action-deepseek-r1-distill-llama-3-1-8b\u002F",[314,400,351,435,793,794,617,618,683,351,357],{"path":805,"title":806,"description":807,"date":808,"slug":809,"image":810,"originalUrl":811,"categories":812},"\u002Fblog\u002Fai-model-optimization-with-paiton","AI-modeloptimalisatie met Paiton","Lees hoe Paiton modelcompilatie, aangepaste kernels en kernelfusie inzet om AI-inferentie op AMD GPU's te optimaliseren.","2025-01-30T19:53:25","ai-model-optimization-with-paiton","\u002Fasset\u002Fimages\u002Fblog\u002Fimported\u002Fpaitonpost1.webp","https:\u002F\u002Feliovp.com\u002Fai-model-optimization-with-paiton\u002F",[314,400,351,435,794,617,618,683,351,357],1789853167081]