注册并分享邀请链接,可获得视频播放与邀请奖励。

Daebak 앤디 的个人资料封面
Daebak 앤디 的头像

Daebak 앤디 (@MyDaebakCafe)

@MyDaebakCafe
0 正在关注    0 粉丝
I am actually disgusted by the timing of this drop. And this will get a bit more technical. Let’s do a real technical teardown of this "1,385-sigma Blackwell weight fingerprint" report that OpenAI and Anthropic are breathlessly promoting today, because banking on nobody reading past the headline is how DC narratives get manufactured. ​First, look at the actual CUDA maths: when PyTorch initialises model weights (torch.nn.init) with a fixed PRNG seed, curand/Philox parallelises random number generation across the GPU’s Streaming Multiprocessors (SMs). Because Hopper has 132 SMs and Blackwell has 148 or 152 SMs, the exact same seed lays out initial weights in a different spatial pattern depending on the GPU’s SM count. In huge MoE models, certain low-gradient null-spaces or unrouted slices preserve faint directional traces of those step-0 weights even after training. ​So yes, brute-forcing the seed and SM grid can show that a model’s step_0 weight initialisation script ran on a 148/152-SM Blackwell GPU. Now read what the summary explicitly admits in the fine print: 1. Initialisation is not full training: The report literally states that these signatures show how the weights were created, and "support (but do not prove) the inference that Blackwell also participated in model training". Initialising a model checkpoint takes seconds on a single rented dev node before loading the safetensors onto a standard Hopper cluster. The "2,000 to 3,000 Blackwell GPUs" and "$500 million" cluster numbers do not come from the weight fingerprint and Depue admits in his own thread that he simply "estimated" cluster sizes from release schedules and model details. ​2. Zero geolocation and zero proof of smuggling: The final sentence of the executive summary explicitly admits that the findings "do not establish the physical location of the chips, how they were acquired, or whether export control regulations were violated". Depue himself rented a cloud GB200 instance just to run his verification test. Commercial labs that are not on the Entity List renting bare-metal Blackwell nodes in offshore clouds (Singapore, Japan, or the Middle East) is completely legal under BIS rules and leaves the exact same SM fingerprint. Now ask yourself why a former OpenAI researcher dropped this on November 12, the exact day Jensen Huang was pushed out, with OpenAI and Anthropic immediately stepping in to "independently verify" it. US closed-source labs are watching their paid API moats get eaten alive by free Chinese open-weights models. Conflating a step-0 PRNG initialisation seed with a "$500M smuggled Blackwell cluster" hands DC hawks the exact pretext OpenAI and Anthropic want to regulate global cloud compute and kneecap open-source AI.
显示更多
0
18
394
60
转发到社区
Just so everybody gets a clear image about this. ​If the board removes him as CEO, several clear boundaries apply: • Every share Jensen Huang already owns outright is his legal property. He retains full voting rights for those shares and can hold them, borrow against them, or dump them on the open market whenever he chooses (subject only to standard SEC insider trading blackout windows if he remains on the board, or Rule 144 volume limits if he steps down completely). • The only shares a board can claw back or cancel upon termination are unvested performance stock units (PSUs) or restricted stock units (RSUs) that have not yet met their time or performance milestones. Given that Jensen co-founded Nvidia in 1993 and holds roughly 3.5% of the company outright, the vast majority of his multi-billion-dollar stake is long since vested. • Because he is Nvidia's largest individual shareholder, firing him against his will creates an immediate existential headache for the board. As an outside shareholder, he could launch a proxy fight to replace the board directors at the next annual meeting, start a rival silicon venture and poach Nvidia's top CUDA and architecture leads, or threaten to liquidate tens of billions of dollars in stock, cratering the share price for institutional investors. In short, firing a founder who owns over $100 billion in vested stock does not strip him of power; it just turns him into an activist shareholder with infinite capital and zero fiduciary obligations to protect the board.
显示更多
JUST IN: Nvidia $NVDA prepares to fire CEO Jensen Huang, WSJ reports.
0
215
10.5K
627
转发到社区