{"data":{"id":2225,"lang":"en","title":"Cerebras Unveils CS-4 Rack System, Claims 30x Faster AI Inference Than GPUs","summary":"Cerebras launched the CS-4 rack system and WSE-3 Turbo chip, claiming up to 30x faster AI inference than GPU-based servers.","category":"ai","category_label":"AI","published_at":"2026-08-21T02:02:25Z","updated_at":"2026-08-21T02:02:25Z","url":"https://techtools.cz/tools/news/article/2225-cerebras-unveils-cs-4-rack-system-claims-30x-faster-ai-inference-than-gpus","api_url":"https://techtools.cz/news-api/v1/articles/2225?lang=en","markdown_url":"https://techtools.cz/news-api/v1/articles/2225.md?lang=en","image_url":"https://www.servethehome.com/wp-content/uploads/2026/08/Cerebras-CS-4-Rack-Hero-Image-scaled.jpg","word_count":182,"reading_time_minutes":1,"sources":[{"url":"https://www.servethehome.com/cerebras-intros-faster-wse-3-turbo-processor-and-first-rack-scale-cs-4-system/","domain":"servethehome.com"},{"url":"https://www.nextplatform.com/compute/2026/08/19/cerebras-overclocks-wse-3-waferscale-engine-to-boost-inference-oomph-in-nexus-cs-4/5289400","domain":"nextplatform.com"}],"translation":{"lang":"cs","id":2298,"title":"Cerebras představil systém CS-4, tvrdí, že je 30× rychlejší v AI inferenci než GPU","url":"https://techtools.cz/tools/news/clanek/2298-cerebras-predstavil-system-cs-4-tvrdi-ze-je-30x-rychlejsi-v-ai-inferenci-nez-gpu","api_url":"https://techtools.cz/news-api/v1/articles/2298?lang=cs"},"content":"Cerebras Systems launched its CS-4 rack-scale server alongside a new wafer-scale chip, the WSE-3 Turbo, claiming up to 30 times more tokens per second per user than GPU-based systems on large language model inference. First shipments are expected this quarter.\n\nThe WSE-3 Turbo packs 4 trillion transistors and 900,000 AI-optimized cores onto a single wafer built on TSMC's 5nm process, doubling its predecessor's compute to 250 PFLOPS and its memory bandwidth to 43.2 petabytes per second, with 44GB of SRAM integrated directly on-chip. The CS-4 rack houses three of these wafers, delivering 750 PFLOPS of total AI compute and 129.6 petabytes per second of memory bandwidth — six times the throughput of a single CS-3 system and ten times higher throughput per watt.\n\nThe CS-4 is the first system built on Cerebras' new \"Nexus\" rack architecture, designed to let the company swap in future wafer generations, including planned CS-5 and CS-6 chips, without redesigning the rack. The launch intensifies competition in AI inference hardware, where Cerebras has positioned wafer-scale chips as a low-latency alternative to Nvidia GPU clusters for serving large models."}}