{
 "id": "hy-memory-30b-a3b",
 "name": "Hy-Memory 30B-A3B",
 "repo": "tencent/Hy-Memory-30B-A3B",
 "lab": {
  "name": "Tencent",
  "country": "China"
 },
 "released": "2026-10-15",
 "parameters": 30.06,
 "activeParameters": 3,
 "maxContext": 262144,
 "architecture": "MoE 128 experts + 1 shared · 48 layers, full attention (GQA, 4 KV heads × 128)",
 "page": "https://studiotvai.com/vram-requirements/hy-memory-30b-a3b",
 "calculator": "https://studiotvai.com/llm-vram-calculator?model=hy-memory-30b-a3b",
 "weightsGiB": {
  "bf16": 55.99,
  "fp8": 28.46,
  "q4_k_m": 16.95
 },
 "kvCacheGiBPerRequest": [
  {
   "context": 8192,
   "bf16": 0.75,
   "fp8": 0.375
  },
  {
   "context": 32768,
   "bf16": 3,
   "fp8": 1.5
  },
  {
   "context": 131072,
   "bf16": 12,
   "fp8": 6
  },
  {
   "context": 262144,
   "bf16": 24,
   "fp8": 12
  }
 ],
 "recurrentStateGiBPerRequest": 0,
 "gpusNeeded": {
  "scenario": "one request of 8192 tokens, BF16 KV cache; vLLM on data-center GPUs, llama.cpp elsewhere",
  "gpus": [
   {
    "gpu": "rtx-4090",
    "name": "GeForce RTX 4090",
    "memoryGB": 24,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 3,
     "fp8": 2,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-5090",
    "name": "GeForce RTX 5090",
    "memoryGB": 32,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 2,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "rtx-pro-6000",
    "name": "RTX PRO 6000 Blackwell",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "mac-m5-max-128",
    "name": "Mac M5 Max 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "strix-halo",
    "name": "Ryzen AI Max+ 395 128 GB",
    "memoryGB": 96,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "dgx-spark",
    "name": "DGX Spark (GB10)",
    "memoryGB": 128,
    "engine": "llama.cpp",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h100",
    "name": "NVIDIA H100",
    "memoryGB": 80,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "h200",
    "name": "NVIDIA H200",
    "memoryGB": 141,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   },
   {
    "gpu": "b200",
    "name": "NVIDIA B200 (HGX)",
    "memoryGB": 180,
    "engine": "vLLM",
    "byFormat": {
     "bf16": 1,
     "fp8": 1,
     "q4_k_m": 1
    }
   }
  ]
 },
 "ggufFiles": null,
 "runsWith": [],
 "verified": null,
 "updated": "2026-10-16T18:03:26.209Z"
}