The time has come to check this place out.

What is interesting about this, if true, is its a recognition that a new class of architecture is needed for AI enhanced/enabled devices.

Assuming the same arch YoY this is a RISC-V based accelerator.

'designed for Meta's AI workloads'

We CT scanned an Apple Vision Pro! We also scanned two Meta headsets. Here’s what we found inside, and what it says about the two companies’ approach to AR/VR and to hardware development in general. 🧵

.Pat Gelsinger shared exciting news in today's #IntelVision keynote – our next processor family, Lunar Lake, will have 3x AI performance on both the GPU and NPU. AND, it will have more than 100 platform TOPS and more than 45 NPU TOPS. Absolutely incredible.

It makes sense Apple is driving the narrative for Apple Vision Pro in the enterprise. This is the market where adoption will take place first before going mainstream to consumers. There are numerous business/productivity/creativity applications for work.…

We all knew it was coming, because what else was Jon Masters 🏴‍☠️ working on :), Google now officially with an @arm CPU.

All three hyperscalers now with @arm cpus for their software stack. This gets super interesting as the ratio of CPUs to GPUs gets closer in future datacenter.…

In our latest episode, Carolina Milanesi (She/Her) and I hit on some important angles of the DOJ v. Apple case. Proving consumer harm will be extremely difficult and many don't seem to see the game the DOJ is playing that would set a dangerous precedent.

Found the FCC report for Humane AI Pin including teardown photos!

It runs the Snapdragon 720G

More great news for the effort of getting the US back to ~20% of the world's leading edge chips by 2030.

Indie developers are the best.

A Holistic Analysis of Nvidia.

How deep is the moat? Are they more susceptible to competition than many realize?

@benbajarin and Jay Goldberg 顾忠南 dive into a full bull and bear case for @Nvidia.

Chapters 👇

(00:30) Competition
(06:32) From Intel and AMD
(12:51) From Startups…

We haven’t had over a 5 in the Bay Area for at least a few years.

Hope all my east coast friends are too shaken up :).

$TSM's customers want whole new fabs, CoreWeave's customers want whole new data centers.

bloomberg.com/news/articles/…

The game for AI inferencing will change drastically when it's common to have enough compute to run 100b+ models locally.

Computing is the new currency.

This is something I notice many people don't understand, including the DOJ lol. Consumers have a choice, and many keep choosing Apple (with 99% satisfaction !) But they don't keep choosing Apple because of some unwilling lock-in or Apple's 'leverage.'

This is an excellent read. Doug hits on something I have been thinking about for a while that we need to think of the entire system (sum of the parts) as the architecture not just the design of the silicon.

Found this quote from Sam Altman at a insvestor symposium. Lots of nuggets, but he makes the case many of us have been making about why bringing a range of Gen AI on device is necessary. Cloud is expensive and the power grid is a limiter.

'Look, I think compute is going to be the…