Unverified50% confidenceFactExact time
NVIDIA Pascal架构（2016年）引入了线程级抢占（Thread-level Preemption）
1
Sources
50%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
UnverifiedPascal架构（2016年）引入了线程级抢占（Thread-level Preemption），使GPU上下文切换粒度更细83% similarUnverifiedNVIDIA recently introduced Multi-Device Inference Support for TensorRT73% similarVerifiedNVIDIA introduced Tensor Cores starting with the Volta architecture, which are hardware units specifically designed for matrix multiply-accumulate operations.71% similarUnverifiedNVIDIA recently announced end-to-end support for Descriptor Heaps in Vulkan, with the entire chain from API calls and driver implementation to hardware execution capable of natively handling descriptor heaps71% similarUnverifiedNVIDIA在Ampere架构（RTX 30系，2020年）引入专用INT8 Tensor Core使INT8推理速度相比FP16提升约2倍，NVFP4是为Blackwell（RTX 50系，2025年）定制的4位浮点格式70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/530919API
curl https://kongchang.com/api/v1/knowledge/claims/530919MCP
get_claim(id=530919)