跨国深谈← 返回←Walter Goodwin

① 他说过的话

8 条

它相当于芯片领域的前沿模型——如果你能找到一种办法，在结构上切分出三到六个月的优势，你将会赢下所有那些部署。
It’s the equivalent of the frontier model for the chip space is if you can just find a way to structurally carve out a three to six months advantage, you will be winning all of those deployments.
—— Walter Goodwin · [00:30]

指向原始笔记的链接

响应更快的聊天机器人，差不多就是快速推理领域里的「更快的马」。
The snappier chatbot is kind of the faster horses of kind of fast inference.
—— Walter Goodwin · [13:21]

指向原始笔记的链接

经济性实际上坍缩为本质上一个东西：你所使用的内存的每 GB 成本。
the economics actually collapsed down to essentially a kind of cost per gigabyte of the memory that you’re employing
—— Walter Goodwin · [14:42]

指向原始笔记的链接

新的 LLM 迫切希望拥有比现在高出几个数量级的内存带宽，以便运行得更快。
a new LLM desperately wants to have orders of magnitude more bandwidth to memory in order to run faster
—— Walter Goodwin · [16:33]

指向原始笔记的链接

另一个，我认为我们不能失去的，是需要进行非常深思熟虑、在架构上微妙的下注，因为你最终真正造出来的芯片仍然需要有三年以上的使用寿命。
There is also, I think, what we don’t lose is the need to place really thoughtful and subtle bets architecturally because the chip you actually end up making does still need to have a three year plus useful lifespan.
—— Walter Goodwin · [19:42]

指向原始笔记的链接

另一部分，我认为这对 AI 实验的递归自我改进同样成立：当驱动实验的思考的智能水平越来越高、而实验因此成为瓶颈时，我认为我们会做更多的思考。
The other part there, and I think this is true of RSI for AI experiments as well, is as our intelligence levels on the thinking that then drives that experiment gets higher and higher, And as that experiment therefore becomes the bottleneck, I think we do more thinking.
—— Walter Goodwin · [26:57]

指向原始笔记的链接

今天有个有点像笑话的说法：那些第一方自研努力，它们的主要目的是压低人们付给 NVIDIA 的价格。
There’s a bit of a joke today that the sort of first party efforts, their primary purpose is to reduce the price that people pay NVIDIA.
—— Walter Goodwin · [32:12]

指向原始笔记的链接

在我部署足够多的那款芯片、从而也获得那种 5 倍计算效率提升之前的九个月里，我可能会死掉。
I could die in the nine months before I get to deploy enough of that chip that I’ve also now gained that kind of 5X in computational efficiency.
—— Walter Goodwin · [34:43]

指向原始笔记的链接

② 出现在这些集

1 集

• 《内存带宽才是 AI 芯片的下一个战场：Fractile 创始人 Walter 的押注》 — 作为嘉宾

③ 他谈到的

点进去有真内容 —— 本页主要出口

Fractile · NVIDIA · Broadcom · TSMC · Grok · Cerebrus · AMD · 推理 · 内存带宽 · ASIC

④ 也在聊「智能体」的人