AITOPAI热报

Latest

Every published item, in order · 117 items
BlogsNewsPostsPapersPapers →
Oct 22
Oct 21
ModelsUS vs ChinaMulti-sourceSelected77°09:43
DeepSeek Open-Sources DeepInfer Inference Framework for V4.1 Flash

Open-sourcing the framework that runs DeepSeek's own production fleet gives operators a rare look at inference internals optimised for both NVIDIA and Huawei Ascend hardware, a dual-platform angle few comparable stacks like vLLM or SGLang emphasise. Teams running DeepSeek models on Ascend 950 in Southeast Asia gain a first-party option instead of community ports. Deterministic sampling is currently NVIDIA-only, so check hardware fit before adopting.

#DeepSeek#DeepInfer#open-source models#inference
Event
SuperTechFans11 sourcesSource
ProductsBuildersMulti-sourceSelected08:37
vLLM Matches DeepInfer Decode Performance on DeepSeek-V4.1-Flash

Open-source inference stacks closing the gap with a lab's proprietary serving setup changes the economics of self-hosting large MoE models. Teams weighing DeepSeek's API against running their own GPUs can use these Wide-EP and EPLB results to estimate revenue per GPU at off-peak prices. The GB200 and GB300 benchmarks also give infrastructure planners concrete throughput targets for the latest NVIDIA hardware.

#vLLM#DeepSeek#DeepInfer#inference
Event
官方账号vLLM@vllm_project7 sourcesSource
Oct 20
PapersBuildersSelected11:47
Addressed-Memory Writes Fail at Reasoning in Qwen3 and DeepSeek Beds

Memory write-injection looks like a cheap way to add knowledge to a deployed model, but these results show the added facts cannot be chained into multi-hop reasoning by the model itself — orchestration must live in the application layer. Builders wiring RAG-style memory into agents should chain written atoms explicitly in prompts or system logic, and benchmark vendors should adopt coverage, synergy and distortion reporting. The tested beds include Chinese frontier models Qwen3 and DeepSeek, making the findings broadly relevant across model families.

#memory#DeepSeek#Qwen#RAG
arXiv: DeepSeek@Wei Gao et al. (4)Source
ProductsBuildersMulti-sourceSelected08:42
SGLang Low-Latency Mode Launches with NVIDIA and AMD Support

The same serving stack now optimises both NVIDIA and AMD paths, giving operators a concrete reason to benchmark MI355X deployments against GB300 rather than defaulting to CUDA. Teams running agentic inference traces can enable the mode through SGLang and expect P90 latency gains without changing models. The plugin system also means third-party kernels — starting with GLM-5.3, K3, and MiniMax M3 — will arrive without waiting for a core release.

#SGLang#NVIDIA#AMD#DeepSeek-V4.1-Flash
Event
官方账号LMSYS Org (SGLang)@lmsysorg11 sourcesSource
Oct 19
ModelsBuildersSelected18:37
DeepSeek-V4.1-Flash-1017 builds full projects from scratch

Self-feedback reinforcement learning that targets whole-project output, not single-file answers, signals a shift in coding benchmarks toward multi-file, runnable deliverables. Developers evaluating open-weight models for end-to-end scaffolding can compare V4.1-Flash-1017 against Claude and GPT coding models on game and 3D tasks before committing to a stack.

#DeepSeek#V4.1-Flash-1017#coding assistants#reinforcement learning
AI Will@FinanceYF5Source
Oct 18
ModelsUS vs ChinaSelected06:10
DeepSeek-V4.1-Flash-1017 lands on Ollama with US hosting

Zero-retention US hosting lowers the compliance barrier for teams that previously avoided Chinese models on data-residency grounds, letting Ollama Cloud serve as an evaluation path before any API commitment. Practitioners can test it immediately via 'ollama run deepseek-v4.1-flash:1017-cloud' or plug it into Claude Code and OpenCode as a cheaper per-task agent model. Against Claude models, the pitch is lower cost per completed task rather than raw capability.

#DeepSeek#Ollama#coding assistants#agents
ollama@ollamaSource
Oct 17
ModelsUS vs ChinaMulti-sourceSelected19:04
DeepSeek-V4.1-Flash-1017 replaces Gemini 4.1 Flash on RSS site

A real-world swap test between DeepSeek's latest flash model and Google's 4.1 Flash on a production translation pipeline, with thinking intensity set to High as the variable. Developers running cheap bulk-translation or summarisation workloads can copy this setup to benchmark latency, output length and quality on their own content. Those already using DeepSeek Harness should update to v0.2.1-rc.1 before adopting the new model.

#DeepSeek#DeepSeek-V4.1-Flash#translation#reasoning models
Event
向阳乔木@vista89 sourcesSource
Sep 17
AITOP16:09
Anthropic拟10月在新加坡设亚太第五个办事处 开...

Anthropic拟10月在新加坡设亚太第五个办事处 开放本地招聘 数据显示，按人口比例计算，新加坡的Claude.ai使用水平在121个国家中排名第二。 我们观察到，人工智能公司Anthropic计划于10月在新加坡设立其亚太地区的第五…

AITOPRead
Aug 13
Aug 7
AITOP17:04
出口结构剧变：中国卖给世界的，不再是产品而是解法

中国出口"新新三样"：从打火机到爬墙机器人，世界不再只是买便宜货 我们观察到一组清晰的数据信号：中国出口的"新新三样"已经从过去的服装、家具、家电，换成机器人、AI模型和创新药物。这不是产业升级的修辞游戏，而是发生在澳大利亚和巴西的实证样…

AITOPRead
Aug 2