Unverified50% confidenceFactExact time
Hy-Memory-30B-A3B's 262K max context is a configured value; the GGUF uploader measured only up to 32K in practice.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/17/2026
First Seen
Valid until: 1/15/2027
Sources
Related Entities
Related Claims
UnverifiedFor actually deploying Hy-Memory-30B-A3B, the realistic bar is a 24 GB GPU or 32 GB Mac running Q4_K_M, while going to Q8_0 with long context requires 64 GB.72% similarVerifiedA single sequence of 32K context length for LLaMA-2 70B can require approximately 80 GB of KV cache in FP16.71% similarVerified以LLaMA-3 70B（80层、64头、每头128维）为例，单个token的KV Cache约占2.6MB（FP16精度），32K上下文的KV Cache总量约达83GB71% similarVerifiedThe NVIDIA H20's peak BF16 compute is approximately 296 TFLOPS, significantly reduced compared to the H100's 989 TFLOPS, but it retains 141 GB HBM3 memory and 4.0 TB/s memory bandwidth.70% similarUnverifiedThe current maximum capacity of a single HBM3e stack is approximately 36 GB, the DDR5 bus width is only 64 bits, whereas HBM can reach 4096 bits.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1054163API
curl https://kongchang.com/api/v1/knowledge/claims/1054163MCP
get_claim(id=1054163)