feat(nccl): add nccl-tests all_reduce_perf for GPU bandwidth testing
- Dockerfile: install cuda-nvcc-13-0 from NVIDIA repo for compilation - build-nccl-tests.sh: downloads libnccl-dev for nccl.h, builds all_reduce_perf - build.sh: runs nccl-tests build, injects binary into /usr/local/bin/ - platform: RunNCCLTests() auto-detects GPU count, runs all_reduce_perf - TUI: NCCL bandwidth test entry in Burn-in Tests screen [N] hotkey Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -8,12 +8,13 @@ import (
|
||||
)
|
||||
|
||||
const (
|
||||
burnCurGPUStress = 0
|
||||
burnCurModeQuick = 1
|
||||
burnCurModeStd = 2
|
||||
burnCurModeExpr = 3
|
||||
burnCurRun = 4
|
||||
burnCurTotal = 5
|
||||
burnCurGPUStress = 0
|
||||
burnCurModeQuick = 1
|
||||
burnCurModeStd = 2
|
||||
burnCurModeExpr = 3
|
||||
burnCurRun = 4
|
||||
burnCurNCCLTests = 5
|
||||
burnCurTotal = 6
|
||||
)
|
||||
|
||||
func (m model) enterBurnInTests() (tea.Model, tea.Cmd) {
|
||||
@@ -48,9 +49,13 @@ func (m model) updateBurnInTests(msg tea.KeyMsg) (tea.Model, tea.Cmd) {
|
||||
return m.burnRunSelected()
|
||||
case burnCurModeQuick, burnCurModeStd, burnCurModeExpr:
|
||||
m.burnMode = m.burnCursor - burnCurModeQuick
|
||||
case burnCurNCCLTests:
|
||||
return m.burnRunNCCL()
|
||||
}
|
||||
case "f", "F", "r", "R":
|
||||
return m.burnRunSelected()
|
||||
case "n", "N":
|
||||
return m.burnRunNCCL()
|
||||
case "1":
|
||||
m.burnMode = 0
|
||||
case "2":
|
||||
@@ -70,6 +75,13 @@ func (m model) burnRunSelected() (tea.Model, tea.Cmd) {
|
||||
return m.hcRunFanStress()
|
||||
}
|
||||
|
||||
func (m model) burnRunNCCL() (tea.Model, tea.Cmd) {
|
||||
m.pendingAction = actionRunNCCLTests
|
||||
m.screen = screenConfirm
|
||||
m.cursor = 0
|
||||
return m, nil
|
||||
}
|
||||
|
||||
func renderBurnInTests(m model) string {
|
||||
var b strings.Builder
|
||||
|
||||
@@ -110,8 +122,15 @@ func renderBurnInTests(m model) string {
|
||||
}
|
||||
fmt.Fprintf(&b, "%s[ RUN SELECTED [R] ]\n", pfx)
|
||||
|
||||
fmt.Fprintln(&b)
|
||||
pfx = " "
|
||||
if m.burnCursor == burnCurNCCLTests {
|
||||
pfx = "> "
|
||||
}
|
||||
fmt.Fprintf(&b, "%s[ NCCL BANDWIDTH TEST [N] ] (all_reduce_perf, NVLink/PCIe bandwidth)\n", pfx)
|
||||
|
||||
fmt.Fprintln(&b)
|
||||
fmt.Fprintln(&b, "─────────────────────────────────────────────────────────────────")
|
||||
fmt.Fprint(&b, "[↑↓] move [space/enter] select [1/2/3] mode [R/F] run [Esc] back")
|
||||
fmt.Fprint(&b, "[↑↓] move [space/enter] select [1/2/3] mode [R/F] run [N] nccl [Esc] back")
|
||||
return b.String()
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user