Hacktoberfest 2026:维护者为十月标记出来的 issue,仍然开放、适合新手。 浏览 Hacktoberfest issue

Ball and Stick: 171 ms per frame for a 5184-atom water box (CPK: 20 ms)

未关闭
#224 2 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

还没有人认领这个 Issue。

评估

难度
4/5
预计耗时
3-5 天
新手友好度
48/100
Issue 类型
缺陷
描述清晰度
基本清楚
活跃度
活跃
技术栈
c
领域
performance

调研方向

Start with the tools/viewer-bench benchmark and profile its water-box Ball and Stick case using perf record; the issue notes that Xvfb is sufficient. Compare where CPU time is spent in moiv/viz versus the rasteriser, using the CPK case as a reference. Done when the profile identifies the repeated per-frame cost behind the 171 ms rendering time.

由索引模型根据 Issue 内容生成。

描述

Measured with tools/viewer-bench (branch wip/viewer-bench) on an NVIDIA RTX 3050, 800×800, mean ms per frame:

System Ball and Stick CPK
water, benzene, Cr(CO)6 0.5–0.8 0.4–0.5
benzene + MO isosurface 1.0 –
water box, 5184 atoms 171 20

Render caching AUTO vs OFF makes no difference in any case.

The same atoms draw 8.5× faster in CPK, so the cost is in the bond path of Ball and Stick. A few thousand cylinders in immediate-mode GL take a few ms on this GPU; 170 ms suggests CPU work repeated every frame (bond finding or geometry rebuilt per render) rather than GL submission. Rewriting moiv's drawing to vertex buffers would therefore probably not address it.

Next step: a CPU profile (perf record) of the benchmark's water-box Ball and Stick case; Xvfb is sufficient, as the profile separates moiv/viz time from the rasteriser. Relevant only for large systems (MD boxes, proteins); teaching-sized molecules render in under 1 ms.

主要语言
C
星标
10
派生
8
平均合并
21 分钟
30 天内合并 PR
1

环境准备

这个项目没有提供开发容器、Dockerfile 或贡献指南,环境需要你自己搭建:先看它的 README,通用步骤见我们的新手贡献指南。

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

FriendsofECCE/ECCE 的其他 Issue

查看 FriendsofECCE/ECCE 的全部 Issue

相似的 Issue

更多 C Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。