Golden data on March 3rd, Lenovo Group recently announced that based on the Lenovo Wentian WA7780 G3 server, it has achieved the industry's first single-machine deployment of the DeepSeek-R1/V3 671B large model at a lower than the industry-recognized 1TGB memory (actually 768GB) to support a smooth experience for 100 concurrent users. According to Lenovo's test data, in a 512 TOKEN standard test environment, the system can support 100 concurrent users to continuously obtain a stable output of 10 TOKENs per second, with the initial TOKEN response time compressed to within 30 seconds.
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Lenovo AI sunucusu ilk kez yerel dağıtım DeepSeek tam kanlı büyük modeli 1TB'den az destekliyor, 100 eşzamanlı
Golden data on March 3rd, Lenovo Group recently announced that based on the Lenovo Wentian WA7780 G3 server, it has achieved the industry's first single-machine deployment of the DeepSeek-R1/V3 671B large model at a lower than the industry-recognized 1TGB memory (actually 768GB) to support a smooth experience for 100 concurrent users. According to Lenovo's test data, in a 512 TOKEN standard test environment, the system can support 100 concurrent users to continuously obtain a stable output of 10 TOKENs per second, with the initial TOKEN response time compressed to within 30 seconds.