00:00 Speaker a
If I were an NVIDIA investor, I was listening to this Patrick. I know, there are some bulls on the street and they are actually more excited about the second half. I’m curious about your opinion on this, and in the reasons that cause excitement, they’re talking about that, as you suggest, there might be this kind of bargaining chip that fits China. They talk about, hey, listen, cloud giant, we hear, they will continue to spend, spend, spend and business with Saudi Arabia and the UAE. Patrick, did I miss something?
01:03 Patrick
No, I think, well, Nvidia has been loaded by the backend for a while, when Blackwell hits, um, um, um, um, um, every dollar per second, and a lot of the super standard is actually waiting, actually waiting for Blackwell. I do think they didn’t buy that much H series because economics is so good for them. I also like NVIDIA’s move to lower Blackwell slightly in Blackwell, uh, by pulling back on an aggressive chassis design, you can actually use many of the same components as the Blackwell 300 in the Blackwell 200. Well, the capital expenditure forecast shows that NVIDIA seems to have done everything. Well, it looks like there is a little less risk than in the first half.
02:59 Speaker a
I want to replace the gears, Patrick, your thoughts on another tech story, which is Salesforce’s acquisition of Informatica for $8 billion. This is the skeptical I saw from rishi jaluria at Royal Bank of Canada (RBC). He said Rishi told his clients that on paper the deal might make sense, bringing Mulesoft, Tableau and Data Cloud together, and we questioned that vision given the lack of integration with Salesforce Assets. Furthermore, even if this logic makes sense, we want to know if there are major advantages to buying information with partners. How did you react to this, Patrick?
04:03 Patrick
So I like the papers here. Well, along with all the CIOs I encountered, one of the top five hurdles for launching and scaling AI generation is the lack of a coherent data management strategy and tools that can extract all of your data. In the AI generation world, you’re heading from the east to the west, not the ML world of the north and south. This means, well, uh, not to say, uh, activate CRM data, but connect CRM with ERP to SCM and connect it to all other datasets. That’s why this makes a lot of sense to me. Well, that’s also parallel to the huge growth in the data cloud we’ve seen with Salesforce. We’ve seen even companies like SAP, and we’ve seen huge growth, right? The predecessor of Big Gen AI action is to make your data line up online. I do think this comment is about the lack of integration, but I hope this integration will happen and needs to be done very quickly. Complexity is another major issue. And if Salesforce can reduce some of the complexity, reduce some of these steps, reduce some of these operational challenges, I think there is something here.
