Running RS twice on same drive amlost 57% faster??

  • Release Candidate 6
    Guest:
    We are at a “proposed final” true release candidate with nothing known remaining to be changed or fixed. For the full story, please see this page in the "Pre-Release Announcements & Feedback" forum.
    /Steve.
  • Be sure to checkout “Tips & Tricks”
    Dear Guest Visitor → Once you register and log-in:

    This forum does not automatically send notices of new content. So if, for example, you would like to be notified by mail when Steve posts an update to his blog (or of any other specific activity anywhere else), you need to tell the system what to “Watch” for you. Please checkout the “Tips & Tricks” page for details about that... and other tips!

    /Steve.

orion2921

New member
Jan 9, 2021
3
0
Not sure whats going on behind the curtains here. Just ran back to back runs on the same drive the speed increase a good deal.

Code:
Driv Size  Drive Identity     Location:    0      25%     50%     75%     100
---- ----- ---------------------------- ------- ------- ------- ------- -------
 81  2.0TB WDC WDS200T2B0B               275.9   525.3   534.3   541.5   536.8

                  Benchmarked: Saturday, 2021-01-09 at 16:14

Code:
  AHCI '.' and IDE '-' drive reset & diagnostic latencies in milliseconds:

  61.607

  Associating BIOS drive [80-81]:

  [ 80 81! ]

  BIOS drive count: 2

  BIOS Inter Bus Location Dev Port Intr
  ---- ----- --- -------- --- ---- ----
   80   USB  PCI  0:29:0  17                                                 

  BIOS Location Type Clas Pg Vend Devi Prt Intr Cmmd Ctrl Mast Driv Link Size
  ---- -------- ---- ---- -- ---- ---- --- ---- ---- ---- ---- ---- ---- -----
   81   0:17:4  AHCI 0106 01 8086 8D62  1   11  < AHCI Cntrl >  6.0  6.0 2.0TB

BIOS Size  Drive Identity     Location:    0      25%     50%     75%     100
---- ----- ---------------------------- ------- ------- ------- ------- -------
 81  2.0TB WDC WDS200T2B0B               475.7   538.1   541.8   541.9   539.3

                  Benchmarked: Saturday, 2021-01-09 at 16:15
 
Not sure whats going on behind the curtains here. Just ran back to back runs on the same drive the speed increase a good deal.
Yes, we've seen this often and your question is a good one. I'm not sure that we even have a theory about it. =I= would be inclined to suggest that if a readback of a very troubled sector "frightened" the SSD controller, it might have recovered the data then swapped it into a new location, thus "healing" the drive for the second and future passes. But it's unknown whether SSD controllers will do that on the fly.
 
  • Like
Reactions: lovecakeboy
Suspected this SSD was having issues for some time now. Crazy a speed test would wake it up, least in that spot. Thanks Steve!
 
What's the size of the area RS reads? I mean how big are chunks at 0, 25, 50% etc.? This particular model appears to have some SLC NAND, 1 GB sized to cache, vs the rest TLC NAND.

=I= would be inclined to suggest that if a readback of a very troubled sector "frightened" the SSD controller, it might have recovered the data then swapped it into a new location

Yeah, rather than 'frightened' it tripped some threshold triggering the data to be stored at a different location, but I assume that's what the "" was there for. If we leave SLC cache out of the picture then I think it's the best working theory for now. Unless there's some quantum effect or whatever that causes NAND cells that are read to be 'refreshed' in some way. It's not far fetched to assume intelligent firmware behaves like that, upon detecting reads dropping below certain specs, reallocate the data, invalidate original pages and queue them for the garbage collector. Conventional hard drives not doing this does not mean SSDs don't, specially since you lose nothing and it helps maintain data integrity and performance and an SDD has this option without sacrificing any spares as you would with conventional drives. You've got everything in place to reallocate data whenever you desire. There's plenty of research suggesting and actual drives that even do this type of refreshing as part of their standard maintenance. It could also explain difference between difference we see level 2 vs level 3. Level 3 'forces' reallocation of all sectors you read/write. Level 2 perhaps only triggers reallocation of slower pages. Level 3 shows more dramatic differences than level 2 if I'm not mistaking.
 
Last edited: