{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreiavkhfaap73k2pxph73dczhq3ikkp6iyi24my36vhrn35ulgrevmy",
"uri": "at://did:plc:ctcqaqa3tlbhfwmfefti346h/app.bsky.feed.post/3mk4l7x5zo5z2"
},
"coverImage": {
"$type": "blob",
"ref": {
"$link": "bafkreibgcidchjiebkxw6qm5fe22trlwolnicw6s2gitdpo5eeqsh5oldm"
},
"mimeType": "image/png",
"size": 1024567
},
"path": "/pro/inference-requires-infrastructure-that-is-closer-to-users-and-more-energy-efficient-antimatter-debuts-global-ai-network-built-to-bypass-grid-bottleneck-with-a-400-000-gpu-roll-out-by-2030-across-1000-data-centers",
"publishedAt": "2026-04-22T21:15:00.000Z",
"site": "https://www.techradar.com",
"tags": [
"Pro"
],
"textContent": "Antimatter targets growing inference demand with global rollout of modular data centers designed to operate where electricity supply is already available.",
"title": "\"Inference requires infrastructure that is closer to users and more energy-efficient\": Antimatter debuts global AI network built to bypass grid bottleneck with a 400,000 GPU roll out by 2030 across 1000 data centers"
}