@Johnero0@zaa_warudo_ Sorry facts hurt your fragile white ego. Asian dude aren't small like your insecure ๐ค๐ค would lead you to believe, and we steal your jobs because you're incompetent, including making your women cum ๐
@risa666327@vkeiweebgf Japanese women are so pathetic for how much you hate your own men. You think foreigners are much better and larger? Quit your delusion and white worshipping, it's embarrassing
@Oriental_Race@Alt_Logos_35 If your first response to "why Asian men don't date XF" is that they'd reject you, then you're a coward who lost the battle before it's fought. It's never been easier for Asian men to date in the West, and you're staying home with your dick in your hand. Maybe go meet some women
@Oriental_Race@Alt_Logos_35 These are the Asian male cucks that choose inaction and stay single while Asian women worship the most bottom tier white men. More WMAF -> fewer Asian women dating Asian men -> more Asian men stay single if they only choose to date Asian women.
@Fresc0D1gital@zaa_warudo_ I give the best sex my white wife has ever had. Y'all white dicklets can BS about Asians being bad at sex, but you can't even make a woman cum ๐
@MikeLongTerm Fundamental thesis on AMD hardware advantage and rapidly catching up software hasn't changed at all. These movements have all been money going from stocks to bonds
๐ @AMD MI355X just showed competitive performance to NVIDIA B300 on @Kimi_Moonshot K3 serving!
8ร MI355X delivers 24.4% more throughput than HGX B300 (7,567 vs 6,083 tok/s/GPU) at 50 tok/s/user on @SemiAnalysis_ AgentX agentic coding benchmark, and 87.4% more tokens per dollar.
We cover the vLLM optimizations, AMD kernels, and hardware details that closed the gap, plus patterns for spotting bottlenecks so you can get more out of vLLM on your own GPUs.
Thanks @AMD@embeddedLLM@SemiAnalysis_
Full writeup: ๐ https://t.co/jP1Dme7dPn
@Nonbruh@jennifer81xu It's for loser white guys who can't get white girls and Asian girls with low self esteem and self hatred. Both bottom of barrel of their respective groups
@PatrickMoorhead@Arm@AMD So you 1.7x the ARM core count against a last generation AMD EPYC, restrict the power draw to 25kW per rack (nonsensical in data centers), and you're flaunting a case-specific 1.99x and average 1.2x performance advantage? LMAO
@pequityresearch CoWoS is supply forecast. This is a change in DEMAND forecast. Unless deals are being cancelled and CAPEX is being reduced, I don't see how demand is dropping