Liam Coffey

[Liam Coffey @_Tiraas]

You’re looking at the wrong resumes. Search for Lego and unconventional things in experience and interests. You’ll hit gold. Trust me. The most passionate and capable can be the ones that don’t grade well and have the worst presentation. I’ve had a career in teaching and guiding people as they’ve taken their first steps in industry on the job from all levels of education. The ones that use university to formalise their experience. This generation in the last 8 years has been stellar. Maybe, it’s that’s just in the UK and Canada. Maybe.

[Liam Coffey @_Tiraas]

@HotAisle Ditto

[Liam Coffey @_Tiraas]

@KyleBennett @HotAisle Especially after this winter up north.

[Liam Coffey @_Tiraas]

@KyleBennett @HotAisle On a beach in San Diego. yes.

[Liam Coffey @_Tiraas]

@HotAisle @Vpoile1 Elio, you legend!

[Liam Coffey @_Tiraas]

@contextconor That's cute, our founding agents are all collectively called Brian.

[Liam Coffey @_Tiraas]

We are in conversations with a select group of strategic and financial partners. If you are an accredited investor with deep experience in AI infrastructure or semiconductors, we welcome private conversations.

[Liam Coffey @_Tiraas]

We have been quietly building something that attacks one of AI’s biggest unsolved problems: the memory wall.

Constantly moving massive model weights between memory and compute is the real driver of high inference cost and power consumption. We think there is a better way.

🧵 1/8

[Liam Coffey @_Tiraas]

If you would like to learn more about the Archality Inference Array™ and our direction, reply or DM us. #AIInference #MemoryWall #BMAC #WeightResident

[Hot Aisle @HotAisle]

I've stayed quiet on this for about a year now.

So good to see TAC coming to the light.

You want fast inference, this will be it.

[The Archality Corporation @thearchality]

We have been quietly building something that attacks one of AI’s biggest unsolved problems: the memory wall.

Constantly moving massive model weights between memory and compute is the real driver of high inference cost and power consumption.

We think there is a better way.

🧵 1/8

[Liam Coffey @_Tiraas]

@HotAisle yes we do.

[Liam Coffey @_Tiraas]

@hthieblot It’s not you, it’s them.
Keep going.

[Liam Coffey @_Tiraas]

@hthieblot @rfvivas We’re building performant silicon with 1/300th of the staff usually required. Ideation to tapeout and through silicon-back. We’re the people nobody has heard of but everyone at AMD has depended to turn profit on chips through the last 8 years. Now we’re doing it without corp.

[Liam Coffey @_Tiraas]

@hthieblot Read up on the legalities/get advice on this to Hubert. You do not want to be out of compliance on any fund raising if you’re serious. Doing this publicly is silly; in both meanings.