Unverified50% confidenceBenchmarkExact time
In the BPE tokenization (cl100k_base, 512/50) scenario, Chunkr achieved 38 MB/s, while Chonkie led at 151 MB/s.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/5/2026
First Seen
Valid until: 1/3/2027
Sources
Related Entities
Related Claims
VerifiedTaking LLaMA-2 70B as an example, when processing a 4096-token context with a batch size of 32, the KV Cache can consume over 80 GB of memory.66% similarUnverified主流BPE词表规模通常为32K-128K，全量成对相似度矩阵的内存占用可超过数十GB64% similarVerifiedLLaMA-2-70B's KV Cache for a single request at a context length of 128K can occupy more than 160 GB of VRAM.63% similarVerifiedLLaMA-2 7B's KV Cache occupies approximately 2 GB with a 4096-token context, and can surge to over 16 GB when extended to a 32K context61% similarUnverifiedA user running a Threadripper platform with 512GB DDR4 memory achieved approximately 8 tokens/s running DeepSeek 671B by offloading expert layers to RAM.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/977587API
curl https://kongchang.com/api/v1/knowledge/claims/977587MCP
get_claim(id=977587)