News for hackers. Stuff that matters.
Re:Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) by timschmidt on Saturday May 31, 2025 @03:29AM (#44142607)
> I have trouble getting excited about Cerebras. SRAM scaling is dead, so short of figuring out how to 3D stack their wafer scale chips

AMD and TSMC are stacking SRAM on the chip scale. I imagine they could accomplish it at the wafer scale. It'll be neat if we can get hundreds of layers in time, like flash.

Your analysis seems spot on to me.

Re:Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) by nsteel on Saturday May 31, 2025 @05:27AM (#44143079)
Assume you meant Intel, rather than AMD?
Re:Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) by timschmidt on Saturday May 31, 2025 @05:49AM (#44143149)
Re:Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) by nsteel on Saturday May 31, 2025 @05:26PM (#44147015)
Yes, and it's TSMC enabling this. Lots of TSMC's customers going this route, not just AMD. Seemed odd to call out AMD as if they've got any special sauce here.
Re:Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) by timschmidt on Saturday May 31, 2025 @05:40PM (#44147100)
My choices can seem odd to you, that's fine. Have a nice day!
Re:Cerebras achieves 2,500T/s on Llama 4 Maverick (400B) by latchkey on Saturday May 31, 2025 @06:18PM (#44147324)
More on the CPU side than the GPU side. GPU is still dominated by HBM.