Source report:
“DeepSeek Debates: Chinese Leadership On Cost, True Training Cost, Closed Model Margin Impacts” by SemiAnalysis
DeepSeek took the world by storm. For the last week, DeepSeek has been the only topic that anyone in the world wants to talk about. As it currently stands, DeepSeek daily traffic is now much higher than Claude, Perplexity, and even Gemini.
But to close watchers of the space, this is not exactly “new” news. We have been talking about DeepSeek for months (each link is an example). The company is not new, but the obsessive hype is. SemiAnalysis has long maintained that DeepSeek is extremely talented and the broader public in the United States has not cared. When the world finally paid attention, it did so in an obsessive hype that doesn’t reflect reality.
We want to highlight that the narrative has flipped from last month, when scaling laws were broken, we dispelled this myth, now algorithmic improvement is too fast and this too is somehow bad for Nvidia and GPUs.
Here’s another clue they’re talking out of their arse. They throw FUD on the Huawei partnership with DeepSeek. Turns out, they were full of shit. A real analytics company, not a grifter who professionalised bribing the supply chain for tidbits, would have seen this coming months in advance due to the foundry traces all the way down the supply chain. They’re just aping Demerjian without any of the contacts or industry knowledge but enough VC money to buy puff pieces on them in business insider.