{"contentHash":"1aab984405197c982b3dc7cb684e6d14d9c82875587b2b31b1ed8a0a84711ce1","protocol":"bitcoin","authorPubkey":"7fba547cafad4e1b88dcf44872562e396a190ac6d50294a040ff2ac5ed4e0405","text":">https://www.theregister.com/ai-and-ml/2026/11/03/british-fintech-revolut-slashes-ai-coding-costs-10x-by-cutting-out-claude-and-going-self-hosted/5307345 >To serve GLM-5.3 at scale, Revolut deployed 56 Nvidia B200 GPUs, equivalent in GPU count to seven eight-GPU DGX B200 systems. But as Abitbol noted, the greatest performance gains didn't come from adding GPUs, but rather tuning the model's hyperparameters along with switching to NVFP4 quantization, implementing disaggregated prefill and decode, and tuning context lengths. Looks like local is getting more popular.","topicHash":"913c62c2ee31793c7e8517600dbc8f0c337afae133700a128a7d8ffb8a21c93e","topicName":"technology","parentHash":"2bf97aff2cbc951346b411c0a5a267b2f13e546a379920f7a2e769c01c9a4022","costSunk":"0","biasSupport":0,"biasOppose":0,"biasVouch":0,"biasNull":0,"blockHeight":null,"confirmations":0,"timestamp":"2026-11-04T01:04:56.304Z","tripFp":"7e7d5fd4abb3","viewCount":0,"powDifficulty":0,"relation":0,"ephemeral":true,"nonce":"9567e27bfc250752","sig":"92a762e6b980be587819f32ab62db552283ed41693047d68813a2d3479c1688c06b98df16a04dcecae6aa57434effeb564bafae06335254cc46b0be16a1134bb","ephStamp":{"blockHashHex":"000000000000000000006cc405be30d8170fe98c6b6108b3e6e0fac75502690b","nonceHex":"244d010000000000"},"replyCount":1,"chainReplyCount":0,"vaporReplyCount":1,"clusterSize":2,"receiptHeight":974464,"expiryHeight":976624,"authorMarks":{"since":"2026-11-03T22:04:37.654Z","via":"sight","inscribed":0,"likes":0,"biases":0},"lang":"en","servedBy":{"host":"6af58b1822e0cabfe57d76540f745d059b438f489a796ec6a705c6540233845e","t":"2026-11-08T20:02:50.199Z","sig":"df6d18606da0c7a9cda737182cd445888a5c05c7303161139b22e6de2de005d748f80277ff7d08f63376922b79ba695115dd93f26e35e96534fd3a706f7daee7","h":"00000000000000000000cd4b0ec60a001c6ea85c4680ce040485c9d4d9866bdc"}}