🚨 BREAKING: @Cerebas just stopped pitching wafer-scale as a monolith.
we got to the supernova cerebras event this week and have some thoughts on their CS-4 launch.
it's a beast.
but the CS-4 story is not just about tokens per second ... (which cloaked 120b above 4.4k tps per
Fireworks and Baseten are the two largest inference clouds and I think of them as *an* index for open-source AI adoption.
Also first time seeing SpaceXAI on the list of “fastest growing” IIRC.
SambaNova! Will wonders never cease.
This is extremely bullish for ASICs (Like SambaNova SN50)
GPU engineers are desperately trying to reproduce the primitives that are built into the ASIC hardware and in effect just validating the market exists
If you believe TileRT is cool, imagine it with 10x throughput, 2-4x
Upper90 has committed up to $400M in debt financing to General Compute, one of the largest facilities ever raised for an ASIC cloud.
It means more capacity for us to build the fastest tokens available anywhere, sooner.
Full story in @TechCrunch today:
