SA260324R-0059
① SA Source
- Source: 開啟完整 SA 文章
- Section:
SRAM and Memory Hierarchy - Line hint:
69
Context Before
We have written about the role of SRAM in the memory hierarchy, but the quick recap is that SRAM is very fast (low latency and high bandwidth) but this comes at the expense of density and therefore cost.
Evidence
SRAM machines such as Groq’s LPU therefore enable very fast time to first token and tokens per second per user
Context After
but at the expense of total throughput, as their limited SRAM capacity quickly gets saturated by weights, with little left over for KVcache that grows as more users are batched. GPUs win for throughput and cost as we have shown.
② Atomic Claim
因此，像 Groq LPU 這類以 SRAM 為核心的機器，可以提供很快的 time to first token，以及很高的每位使用者 tokens per second。
- Epistemic Mode:
INFERRED - Mapping Status:
COMPLETE
③ Semantic Frame
{
"attribute": "USER_LATENCY_AND_TOKEN_RATE",
"context_nodes": [
{
"id": "04_knowledge_base/SRAM",
"label": "SRAM"
},
{
"id": "04_knowledge_base/Time to First Token",
"label": "time to first token"
},
{
"id": "02_companies/Groq",
"label": "Groq"
}
],
"entity": {
"id": "04_knowledge_base/Groq LPU",
"label": "Groq LPU"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": "SRAM machine",
"numeric_mentions": [],
"temporal_mentions": []
},
"value": {
"numeric_mentions": [],
"value_text": "提供很快的 time to first token 與很高的 tokens per second per user"
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | Groq LPU | 04_knowledge_base/Groq LPU |
| context_node | SRAM | SRAM |
| context_node | time to first token | 04_knowledge_base/Time to First Token |
| context_node | Groq | Groq |
⑤ Human Review
請在 Properties 逐項確認：
- 原文 → Atomic Claim 是否忠實
- Atomic Claim → Semantic Frame 是否忠實
- Canonical Entity mapping 是否正確
- Epistemic mode 是否保留原文語氣
- 最後選擇
review_action
Review state
Markdown 內文不是正式 approval。只有 Apply bridge 寫入的 Decision Ledger event 才是正式決策。