June 3, 2011
|
Duration |
Confidence |
Remarks |
| Less than 1 pass |
0% |
Totally pointless. Status still 100% unknown. |
| 1 complete pass |
50% |
Able to declare "not known to be failed" |
| 1 hr (at least 2 passes) |
60% |
Slightly stronger versions of above declaration |
| 2 hrs |
70% |
| 4 hrs |
80% |
| 8 hrs |
90% |
Minimum "meaningful" mem diag test |
| 24 hrs |
95% |
Basic ISS* mem diag test |
| 48 hrs |
98% |
Standard ISS mem diag test |
| 72 hrs |
99% |
Comprehensive ISS mem diag test |
| 1 week |
99.5% |
Long term tests |
| 2 weeks |
99.9% |
| 3 weeks |
99.99% |
| Longer than 3 weeks |
99.99% |
Practically pointless. No further gain. |
Non-Contiguous Cumulative Equivalency
Bravo Policy & Formula
- any sub-hour session is excluded from consideration
- all sessions older than 1 week are ineligible
- equivalency = current session +
all eligible previous sessions x 0.5
- confidence declaration is "as of conclusion of current test
session"
Examples
- 30 min, 1.5 hrs, 2.5 hrs today: 0 + 1.5 + 2.5 = 4 hrs
equivalence
- 4 hrs yesterday, 4 hrs today: 4 /2 + 4 = 6 hrs equivalence
- 1 hr per day for the past month: 6 / 2 + 1 = 4 hrs equivalence
- 100 hrs last yr, 8 hrs yesterday: 0 + 8 = 8 hrs equivalence
Tutorial
How to run: Simply boot with special diagnostics boot
disc (CD or USB) on Bravo Sonata workstations, or press F2, F8. On
notebooks, and other non-Bravo workstations, you might need to
deal with
boot
device sequence.
(could be tricky)
Program Screen Layout
Top Left: System specs, including memory size detected
Top Right: Current Pass/Test/Pattern/Percentage Status
Bottom Section: Status of overall session statistics, followed by a
running log of all errors encountered thus far (if any).
What to look for & report: 1. WallTime
2. Pass 3. Errors
NOTE: Any error would be displayed in red.
Generally, once an error is encountered, you should immediately report
it. Most likely, you'd then be advised to terminate the test, after
obtaining whatever details that might pertain your situation.
*ISS - In-Shop Service see
glossary |