Ryan Gould
speaker
56 appearances
3 recordings
1 series
first heard May 2026
last heard 13 Aug
Ryan Gould’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 3 in all, peaking in Aug 2026 with 1.
Appearances
And when you look at across the street and think about some of these other sort of chip inference type plays, efficiency plays that are coming out on the market, like you know, we've been covering Ten's Torrent Ed, you know it well.
These are all types of deals where I think companies like Anthropic are realizing that there is actually
massive value here and there are still quite few of them at scale.
So it's kind of a scarcity play I think at this point.
Yeah, so they're in advance talks to acquire a modular for about four billion dollars.
It looks like that deal could be announced as soon as the coming weeks.
I think Ed, this is part very much of this wider AI race for infrar uh race in AI for inference technology.
Basically that's the process where you go from training models to actually putting them into production.
And we've seen, I think, probably bookended by both the Grok uh NVIDIA licensing deal that announced just before
Christmas, uh we all remember then on Christmas Eve, uh and maybe also the San Bernova deal with Intel and uh Vista Equity Partners, the private equity firm.
Um this is just really an arms race right now to see how quickly we can start to lower some of these inference costs for enterprise.
Yeah, I mean there's a quite a quite a few uh famous VC firms in there, really.
I mean, Google Ventures, General Catalyst.
General Catalyst is in there for quite a significant chunk of the money they were valued at uh just over a billion dollars uh last September, I believe.
Um so this would kind of be a fairly quick turnaround for those guys involved there.
And I think this kind of also gets at something else, which is we've been hearing that Qualcomm has been looking at companies like TenceTorrent, again, another player.
That sort of sits in and around this this inference AI sphere.
Um modulo itself though, I mean, really is that software layer.
It's it's putting those hardware chips and making them accessible to cross all different models, so that's ARM, um, you know, Intel, Nvidia, uh, and in this case soon to be Qualcomm, uh, should this all come to plan.
It does, Ed.
Showing 21–40 of 56 · page 2 of 3
← Previous
Next →