Chinese semiconductor thread II

Hyper

Junior Member
Registered Member
I've been to niche tech/business conferences like this. "Look at our EUVL" is just a reminder to Chinese customers that "we're still the best so even if you can't buy EUVL from us, buy our DUVL instead of upcoming ones made by Chinese companies". Don't underestimate the cunning of sales and marketing teams. They may lack the hard skills of the RD&E teams but they know how to influence and manipulate human minds for maximum sales revenue.
ASML doesn't lack hard skills of RD&E though.
 

MortyandRick

Senior Member
Registered Member
Same story as Pirelli, they saw that there has been no consequences with robbing the chinese and replayed the same tactic.

I hope Midea have secured Kuka's assets and IPs, I think it's already on the list of the companies to be robbed.
This is a direct result of the BIS 50% rule. This is also why China hit back so hard.

If the US relents, the Dutch will also relent.
 

tphuang

General
Staff member
Super Moderator
VIP Professional
Registered Member
This is not the right thread to be talking about Kuka

Please, Log in or Register to view URLs content!

30亿,工商银行海光芯片服务器大单敲定

10月1日消息,中国工商银行公示2025海光C86架构服务器集采的评审结果,浪潮电子信息产业股份有限公司、中兴通讯股份有限公司、联想(北京)信息技术有限公司位列中标候选人。该项目标的金额高达30亿元,除浪潮夺得标的入场券之外,海光成为最大赢家。
ICBC spent 3B RMB buying Hygon C86 servers, also use Inspur server and ZTE communication

35 亿,科大讯飞投资算力平台、大模型

9月30日,科大讯飞发布公告,计划通过向特定对象发行A股股票募集不超过40亿元资金。此次募集的资金将主要投向两大核心项目:一是“星火教育大模型及典型产品项目”,投资约11亿元,拟投入8亿元募集资金,旨在升级其星火教育大模型并将其深度应用于AI智慧课堂和AI学习机等产品;二是“算力平台项目”,拟投资24亿元,旨在进一步加强其基于国产算力芯片打造的全栈自主可控的大模型训练与推理能力。此次融资将重点支持公司在教育大模型领域的深化应用和国产化算力平台的战略布局。
iFlyTek is spending 2.4B RMB on domestic AI chip for training and inference (they are a long time Ascend customer)

4亿,河南中南信息科技有限公司采购400台算力服务器


9月24日,河南中南信息科技有限公司发布《中南昇腾智算集群租赁项目》竞争性磋商公告,最高限价5000万元/年。该项目将租赁400台算力服务器及配套网络安全设备,以及机房物理环境服务。
Henan DC here also spending 400m RMB on Ascend chips. 400 compute servers.

9月26日,东江湖万卡国产化超智融合算力集群项目签约。该项目位于湖南省资兴市东江湖大数据产业园,由湖南省资兴市与上海武桐树高新技术有限公司签约,总投资15亿元,计划分四期建设并于2026年6月底前完成,内容包括依托国产高端GPU板卡部署700个10KW机柜及配套设备,提供不低于2440PFlops@FP16的总体算力项目建成后,可服务于AI大模型训练和科学计算等多种场景,旨在成为国产化算力中心的示范项目,推动相关产业的智能化转型。
Dongjianghu here also investing 1.5B RMB on domestic high end GPU for 2.44 EFLOPS computation
 
Top