maodouchong 发表于 2025-01-27 12:12 BREAKING: DeepSeek has restricted registration to its services, ONLY allowing users who have a mainland China mobile phone number to register. 我就说嘛
maodouchong 发表于 2025-01-27 12:12 BREAKING: DeepSeek has restricted registration to its services, ONLY allowing users who have a mainland China mobile phone number to register. 我就说嘛
Gavin Baker 1) DeepSeek r1 is real with important nuances. Most important is the fact that r1 is so much cheaper and more efficient to inference than o1, not from the $6m training figure. r1 costs 93% less to *use* than o1 per each API, can be run locally on a high end work station and does not seem to have hit any rate limits which is wild. Simple math is that every 1b active parameters requires 1 gb of RAM in FP8, so r1 requires 37 gb of RAM. Batching massively lowers costs and more compute increases tokens/second so still advantages to inference in the cloud. Would also note that there are true geopolitical dynamics at play here and I don’t think it is a coincidence that this came out right after “Stargate.” RIP, $500 billion - we hardly even knew you. Real: 1) It is/was the #1 download in the relevant App Store category. Obviously ahead of ChatGPT; something neither Gemini nor Claude was able to accomplish. 2) It is comparable to o1 from a quality perspective although lags o3. 3) There were real algorithmic breakthroughs that led to it being dramatically more efficient both to train and inference. Training in FP8, MLA and multi-token prediction are significant. 4) It is easy to verify that the r1 training run only cost $6m. While this is literally true, it is also *deeply* misleading. 5) Even their hardware architecture is novel and I will note that they use PCI-Express for scale up. Nuance: 1) The $6m does not include “costs associated with prior research and ablation experiments on architectures, algorithms and data” per the technical paper. “Other than that Mrs. Lincoln, how was the play?” This means that it is possible to train an r1 quality model with a $6m run *if* a lab has already spent hundreds of millions of dollars on prior research and has access to much larger clusters. Deepseek obviously has way more than 2048 H800s; one of their earlier papers referenced a cluster of 10k A100s. An equivalently smart team can’t just spin up a 2000 GPU cluster and train r1 from scratch with $6m. Roughly 20% of Nvidia’s revenue goes through Singapore. 20% of Nvidia’s GPUs are probably not in Singapore despite their best efforts. 2) There was a lot of distillation - i.e. it is unlikely they could have trained this without unhindered access to GPT-4o and o1. As @altcap pointed out to me yesterday, kinda funny to restrict access to leading edge GPUs and not do anything about China’s ability to distill leading edge American models - obviously defeats the purpose of the export restrictions. Why buy the cow when you can get the milk for free?
不抛点亏很大啊 BREAKING: This is not a memecoin. This is Nvidia, $NVDA, the most valuable company in the world before today. It is down 17%. It lost $560 billion in market cap today so far, the largest in market history.
回复 1楼 千渔千寻 的帖子 ADAM @AdameMedia BREAKING: Nvidia is down 15% representing a wipeout of $450 billion! The NASDAQ is on track to lose $2 TRILLION All thanks for DeepSeek AI from China. DeepSeek is the biggest paradigm shift and single most disruptive product in a long time. It also exposes how much money is being funneled into what increasingly feels like a massive tech scam. They showed you can do AI MUCH cheaper. A few million compared to so many billions. And it’s all open source (which is a massive win for humanity!)
Ai相关都要爆了，估值都得重新定价，这可不是短期波动倒车接人
科技股得废
不得不说，你太牛了！！
你太牛了👍
即使没nvda也没关系
我是去年年底还没到140就卖了
去年9月跌到90几元的时候，大家都在说经济危机来了，我有上来说只是一个 dip, 后来涨到140的时候，那个波段的最高点，我也有上来说我要全部清仓了，我一直在做 NVDA 的波段，低吸高抛。 后来经常在这里被骂，就不太说了。
对啊， deepseek再牛也是囤了老黄的H100
让子弹飞几天吧，我等着加仓呢，哈哈
所以QQQ其实不好做
IWM好做多了，基本上每次大跌很快都可以回到ATH，QQQ一旦下去了有可能回不来，而且深不见底
是的， 直接把川普刚宣布雄心勃勃的5000亿的AI“星际之门”计划干成废柴了。
那天孙正义，Sam 几个站台，动辄投资5000亿，一个几个Trillion的大盘子开始吹的时候，我脑海里第一时间就想到了文贵。
很多人心态崩盘了。
还真是，stargate刚宣布就完蛋了，简直超级打脸
抛开和sam，孙正义的个人恩怨不说，马斯克脑子至少是清醒的，对AI也是有想法的，他一开始就知道这个星际之门计划是个大忽悠，根本不靠谱。 可惜那是川普的面子工程之一。他只能悻悻在一边挖苦几句，不敢站出来炮轰。
是想拉拢中东向东大靠拢？ 投资中国AI?
可能就是刚好做出来了，正常发布
这周真的太开心了。这才周一。
回转了一些
nvda股东笑出不出来😭
就deepseek那点投资量，需要中东大佬们投资吗？国内投资方现在排队都可能排到西湖边了
躲被窝自己继续去骗自己吧，不是真的，都是假的，连说100遍就行了。
被cyber attack 了
hahaha，你们之前吹Temu和前一阵吹A股的时候也是这样的
川普也被他们给忽悠了。也许他心里清楚，但在资本利益面前也没辙，他也不想让彼此尴尬。
这里很多人生活在公元前
今天终于轮到我笑一笑了。。 不仅在150清空了nvda，还在533清空了qqq。。。
蘑菇头高兴死了，名声赚了有不用花钱
难道对方不是用的女大的gpu吗？只不过不是最顶尖的而已。我记得那里说用了1000块Gpu, 小公司一听高兴了，赶紧的上，只要5百万，不要太便宜啊。
谁干的？ nsa? Open ai, 还是自导？
我多谢你哦， LOL
对的
我刚刚注册了，没有中国电话，用谷歌邮箱注册就行。
不要试图叫醒假装睡觉的人。
重要的是，只有7B！普通电脑都能运行试试了！
1) DeepSeek r1 is real with important nuances. Most important is the fact that r1 is so much cheaper and more efficient to inference than o1, not from the $6m training figure. r1 costs 93% less to *use* than o1 per each API, can be run locally on a high end work station and does not seem to have hit any rate limits which is wild. Simple math is that every 1b active parameters requires 1 gb of RAM in FP8, so r1 requires 37 gb of RAM. Batching massively lowers costs and more compute increases tokens/second so still advantages to inference in the cloud. Would also note that there are true geopolitical dynamics at play here and I don’t think it is a coincidence that this came out right after “Stargate.” RIP, $500 billion - we hardly even knew you.
Real: 1) It is/was the #1 download in the relevant App Store category. Obviously ahead of ChatGPT; something neither Gemini nor Claude was able to accomplish. 2) It is comparable to o1 from a quality perspective although lags o3. 3) There were real algorithmic breakthroughs that led to it being dramatically more efficient both to train and inference. Training in FP8, MLA and multi-token prediction are significant. 4) It is easy to verify that the r1 training run only cost $6m. While this is literally true, it is also *deeply* misleading. 5) Even their hardware architecture is novel and I will note that they use PCI-Express for scale up.
Nuance: 1) The $6m does not include “costs associated with prior research and ablation experiments on architectures, algorithms and data” per the technical paper. “Other than that Mrs. Lincoln, how was the play?” This means that it is possible to train an r1 quality model with a $6m run *if* a lab has already spent hundreds of millions of dollars on prior research and has access to much larger clusters. Deepseek obviously has way more than 2048 H800s; one of their earlier papers referenced a cluster of 10k A100s. An equivalently smart team can’t just spin up a 2000 GPU cluster and train r1 from scratch with $6m. Roughly 20% of Nvidia’s revenue goes through Singapore. 20% of Nvidia’s GPUs are probably not in Singapore despite their best efforts. 2) There was a lot of distillation - i.e. it is unlikely they could have trained this without unhindered access to GPT-4o and o1. As @altcap pointed out to me yesterday, kinda funny to restrict access to leading edge GPUs and not do anything about China’s ability to distill leading edge American models - obviously defeats the purpose of the export restrictions. Why buy the cow when you can get the milk for free?
求多发帖多回帖
你怎么知道老黄没有抛。。。
东大扇起川大来简直不留情面
不抛点亏很大啊
BREAKING: This is not a memecoin.
This is Nvidia, $NVDA, the most valuable company in the world before today.
It is down 17%.
It lost $560 billion in market cap today so far, the largest in market history.
我觉得，AI的核心从业人员，肯定早就明白这是个大bubble，很多事情被吹得，夸张的太厉害，芯片显卡算力什么的，但是呢，外行反正也不懂，也没有别的国家/企业，能搞出第二个，那么大家就都秘而不宣，揣着明白装糊涂，让别人接着吹，吹的神乎其神才好呢
但是，这个东西，就怕有有人能造出第二个，而且这个人不乐意，和你入伙，一起接着神秘。而是，马上公之于众，魔术的秘密，普通群众惊呼，“原来就这！？”，从此跌下神坛，再无任何神秘色彩
只不过，这次，来的也太快了！连几年都没有，就几个月吧，就被demysify了
还好吧，很多人成本低着呢
因为被大规模cyber attack 了
懂王为啥叫懂王？没有谁比我更懂llm
AGI还没出来，说什么泡泡。多了解一下这个行业吧。SAM和梁都说有希望，就是十年的事，老菜帮们没一点眼界的都可以退休了。
re
这里充满了股坛神棍
最终消费者受益
前两天有个A股十多年十几%效益的神人，去请教下。
早就出来了
ADAM @AdameMedia
BREAKING:
Nvidia is down 15% representing a wipeout of $450 billion!
The NASDAQ is on track to lose $2 TRILLION
All thanks for DeepSeek AI from China.
DeepSeek is the biggest paradigm shift and single most disruptive product in a long time.
It also exposes how much money is being funneled into what increasingly feels like a massive tech scam.
They showed you can do AI MUCH cheaper. A few million compared to so many billions.
And it’s all open source (which is a massive win for humanity!)
这么大的新兴公司怎么都不会死，只是过去两年把后面30年的财都发了
死猫还得跳三下了，，，不着急抄底
你买A股，比如四大银行股，存个20年，年回报率都在10%以上。我自己一个A股，近20年没动，年回报率接近10%，如果是茅台的话，更加厉害。问题是，普通人没有这种定力。
我也等着呢，看能进场吗，不着急
问题是普通人不是神仙，要不都发了
不着急，让事情发酵一下
没等我加仓今天就开始涨了，打脸来的太快了
终于有大投行开始质疑美国IT科技业的高估值了。
DeepSeek's emergence could call U.S. tech's 'stratospheric valuations' into question, says Deutsche Bank
恭喜恭喜，刚看了一下涨了7%了