[Engine] apply_nv_defaults labels sm_89 (Ada) as "sm_86 (Ampere)"

Open Beginner friendly
#6,229 1 comment 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 5 days

Nobody has claimed this yet.

Assessment

Difficulty
2/5
Estimated time
1-3 hours
Newbie friendliness
73/100
Issue type
Bug
Clarity
Clearly specified
Activity status
Active
Tech stack
cpp
Domain
backend

Research direction

Read src/core/nv_defaults.cpp, especially the startup label function and the (8,9) branch in nv_default_env. Check how the architecture label and performance environment variable are selected for compute capability 8.9. Build and run strata --serve on an Ada card; done when any applicable startup label identifies sm_89 (Ada) and uses the Ada-specific escape hatch.

Written by the indexing model from the issue text.

Description

needs-format

"Area":

[Engine]

"Desktop app, Terminal or source?":

Source (START-HERE.bat, setup.sh or a build)

"What happened":

In src/core/nv_defaults.cpp, the startup label function checks:

const bool sm86 = mj == 8;

This is true for both sm_86 (Ampere: RTX 3060, A4000) and sm_89 (Ada: RTX 4070,
RTX 6000 Ada). If any switch is in the table for an Ada card, the startup line
prints "sm_86 (Ampere)" and names STRATA_SM86_PERF as the off switch, which is
wrong for Ada.

The table for (8,9) is empty today (no switches win on Ada, measured in my
benchmark issue), so the wrong label never prints at runtime. But the logic is
incorrect for the moment someone adds a switch, and it means there is no
STRATA_SM89_PERF escape hatch documented for Ada owners who want to verify
the empty table is correct on their card.

"Smallest steps or command":

  1. Add any switch to the (8,9) branch in nv_default_env
  2. Build and run strata --serve on a compute 8.9 card
  3. Observe the startup line says "sm_86 (Ampere)" instead of "sm_89 (Ada)"

"Reproduction status":

Unconfirmed static analysis (I read code; I did not see it fail)

"Strata version":

Strata 0.2.15 · engine 0.1.60 (also checked at HEAD 0.1.61)

"Graphics card, memory and system":

RTX 6000 Ada Generation 48 GB · 192 GB RAM · Windows 11

"Model, size and settings":

Qwen3.8-Flash-Next Unsloth UD-Q5_K_XL · 524288 context · vision on0

Dominant language
C++
Stars
11.6k
Forks
1k
Avg merge
18d 13h
Merged PRs (30d)
4

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from Niko1221/Strata

All issues in Niko1221/Strata

Similar issues

More C++ issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.