Coffeehouse Thread

3 posts

Forum Read Only

This forum has been made read only by the site admins. No new threads or comments can be added.

Cheap laughs for those who haven't moved to SSD's

Back to Forum: Coffeehouse
  • User profile image

    Crucial quickly released a firmware update for the 5200 hour bug recently found on the popular m4 SSD. This bug would cause systems that reached a total of 5184 hours on their m4 SSD to become unstable. Once restarted systems would be stable until another hour had passed.


    Actually, I did get SSD recently but it's too early to say much about it besides that just to get it working stably I had to figure out to lower the CPU overclocking. I'll let you chuckle at the thought of how much time I wasted figuring that one out (was almost ready to return it to the shop). Atleast I didn't have to lower the clocks by much - if I had to, I'd rather keep overclock than SSD because stuff can be cached to RAM which speed up my real world apps 2-3x or more compared to loading them from SSD.

  • User profile image
    Sven Groot

    I've got an Intel SSD as my OS drive in this machine, and my laptop has RAID0 dual-SSDs. The speed-up is amazing (on my desktop I installed that SSD without upgrading any of the other components, so I know exactly how much speed-up was gained from that).

    Note that I don't overclock. It does seem strange that a disk drive would be affected by the CPU speed, unless the motherboard is doing something wonky.

  • User profile image

    @Sven Groot: I found few similar reports after finding the problem. I would venture a guess that it depends on what's at either end of the cable. The HDD's I have and have had never pushed the bus to the limit even when they're talking just to the HDD's cache (they were using legacy IDE driver).

    The problems prior to lowering the clock appeared as near-complete freezing (audio appeared to go on very very slowly) with no mouse cursor movement. Since I hadn't seen that kind of problem on this computer prior to installing the SSD, my first guess was it had something to do with the controllers. The crashes only appeared to occur then AHCI drivers were used. In benchmarks, the AHCI drivers get sequential speeds that saturate the bus bandwidth. The legacy IDE driver maxes out at under 150 MB/s for reads from the SSD. Eventually I just figured the overclocking might be doing something that only becomes an issue when the bus is used at full speed. This computer still uses FSB overclocking, can have more effect on other components,  unlike the recent Intels where you can get quite far just by increasing the CPU's multiplier if it's unlocked processor.

    I currently have the SSD as secondary drive and redirected certain OS & program folders there. The result is, that VS2010 cold starts now into the project selection in 2-5 seconds depending on if SuperFetch has loaded things into RAM yet. The warm start speed is about 2 sec and cold from SSD 5 sec (just after cold boot, while OS may still be bit busy with checking Windows Updates).

    While the system is now stable in a clean install with the MSAHCI and RST drivers, there's still some tests needed before I could consider enabling them in the registry for my main install.

    One particular issue I found in tests so far is that the warmed up system isn't as snappy when the OS is on the SSD. eg. Consent dialog could take almost 1 second to pop up on the clean install, where I guesstimate it at 0.3-0.5 seconds on my primary HDD Windows install. I'll try a complete clean install again on the SSD once more to verify if this persists or if it's result of trying out various versions of the AHCI drivers or some other tests.

Conversation locked

This conversation has been locked by the site admins. No new comments can be made.