I am not an AI algorithm engineer, but the recent developments in China's AI models—from last month's rising star, Deepseek, to this month's newcomer, Manus—have had a significant impact on the demand for AI chips (at least within China).
Love the post - hadnt seen a good market sizing. This feels bearish for NVDA and all AI stock. H800 cost how much $40k? If 1MM of them are needed that’s only $40bn in sales. Doesn’t seem like enough to justify a $3Tn NVDA valuation unless it’s at least 100x that. That’s before we even consider that chips get better and inference gets more efficient at a rapid clip
Love the post - hadnt seen a good market sizing. This feels bearish for NVDA and all AI stock. H800 cost how much $40k? If 1MM of them are needed that’s only $40bn in sales. Doesn’t seem like enough to justify a $3Tn NVDA valuation unless it’s at least 100x that. That’s before we even consider that chips get better and inference gets more efficient at a rapid clip
Hi Chris, please note this is only China portion. I am not sure if other nations will follow Deepseek's optimization route.
If 1.1m h800 is needed...then it's like pretty much all h800 has ever been made right?
Follow the same logic, how far china can go with inferencing even with h20 or huawei chip?