Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

Archer42

Member
  • Content Count

    363
  • Joined

  • Last visited

Awards

This user doesn't have any awards

About Archer42

  • Title
    Member

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. I am going to throw in typical "raid is not backup" warning. If you are planning to use raid at home - stop and think about what you are trying to accomplish and what raid actually does. In short raid is needed to increase availability, trying to eliminate potential downtime due to disk failures. What it does not do - it does not guarantee data safety. Arrays still can fail for various reasons and data still can be lost in multiple ways. In a sense it can be even more risky than single drive because of increased complexity. So in most cases raid
  2. They probably did not want it there for some reason... Also samsung is one of manufacturers which publishes unusually detailed specs for retail ssd-s. Including stuff like speeds after slc cache and slc cache sizes. And they generally do not change hardware without changing model apart from may be newer flash of the same type and other things which do not affect performance. If you look at specs of someone who does change hardware constantly, like kingston or adata, their specs are quite a bit less detailed. They do not. They have "up to" everywhere and do not mention
  3. It should. If you copied all small partitions with boot files and stuff and system partition it should work.
  4. It will probably work. It is just a good habit to always start with imaging because this is much safer way. If you are willing to take risks you can skip the step. Also when reading files from damaged hdd once you bump into bad sectors you'll have to manually skip those files, otherwise it will take unrealistically long time. ddrescue does it for you by skipping bad sectors relatively fast. I understand that you probably have no space to store the image, but may be it is a good moment to think about buying external hdd? Can use it for recovery now and for backups later. Because if ss
  5. Imaging the disk is the first thing that should always be done in any "data recovery" situation. Because such image will provide exact copy of hdd regardless of filesystem state and such. You can write it to another hdd and work with it to recover data, if you mess it up you can always retry once more. Basically once the image is done you are safe in terms of not making things worse and can play around with stuff safely as much as you want. It skips unreadable sectors. Might retry again later if you want to. But if they are completely unreadable nothing can be done about it anyway.
  6. Yes, nvme only and tooltip shows that it is literally presence of dram cache. HMB should be completely separate. May be incorrect though, that's true. In specific case dram is indeed present because it is pretty easy to see physically. There is also this set of utilities which allows viewing a bunch of information, which is not displayed from some database but is read from ssd itself.
  7. I personally tend to take certain precautions, like with any electric equipment left unattended. Use separate low current fuse in outlet it is connected to, make sure nothing easily flammable (paper, cloth) left close to it, use full metal case for stuff that runs overnight etc. Someone might call it stupid, but IMO totally worth tiny bit of trouble. House fires are no fun and they happen when people get careless with stuff. That said i never had an incident at home when something really caught fire. Burned components - sure, but shorted/burned capacitor or melted connect
  8. Are they? In my experience high-end cards rarely last long. 4-5 years is good already. Probably has something to do with power density and physical damage which rapid uneven heating/cooling causes to GPU. Not the reason in this case though... Problem with fixing damage like this is it cannot really be fixed. There are probably a bunch of shorts in there, will probably have to drill a hole there removing burned parts and simply remove one "phase" which is destroyed. Card can work like this, but it will increase load on other components making second failure more probable. That is
  9. It creates full image of hdd skipping bad blocks. It is relatively safe, does not write anything to disk at all, but working with damaged hdd can always damage it further, no way around it. As for how to use - boot in some linux live cd, this should work. Have enough space to store the image somewhere. The tool is commandline, you can read the manual here.
  10. With bad sectors... well you have 2 options IMO. 1. Just boot into OS and copy files you need first. Reinstalling OS and stuff might be a pain, but actual unrecoverable info is always more valuable. 2. Use ddrescue. It is one of the best if not the best tool for the job. Do not run checkdisk before the data is safe on other device. You are right to be scared, it might fix stuff or it might destroy it.
  11. Honestly with cooler like this my opinion would be such temperatures at such load are not normal. Do all fans work? If they do probably yes, needs a disassembly and paste replacement. If you never done it find a friend who did and do it together. Nothing really hard, just need to be careful. Thermal pads are most likely fine, just need to be careful not to damage them. Also one fun way to deal with old noisy fans on card like this is removing decorative cover along with fans from the cooler (leaving heatsink only) and attaching 2-3 120mm fans with zip ties. Does not look prett
  12. Many different BSOD-s with pretty high probability mean memory issues. Also since they are rare issues are, most likely, intermittent. Disable XMP if it is on. Try running memtest86 overnight.
  13. Basically manufacturers do not want to commit to anything. They show useless peak speeds and nothing else, so that they may change hardware and performance as they wish in future without changing model. Look for reviews... Also crystal disk info does show it: https://imgur.com/fCeQ6gK
  14. The only thing I'd be very careful about with such damage is that you've broken off part of the key which prevents you from inserting connector wrong way. Nothing too bad, just need to make sure you are inserting it the right way instead of simply relying on keys.
  15. To a degree may be. But they tend to also use the cheapest controllers without dram, and when everything is combined together they can be bad enough for it to become noticeable. Especially after you fill them with data and use for some time with all the writes OS does. One more reason for it would be price. People expect 128GB to be 2x cheaper than 256GB, they will not buy 128GB if it will be only 10% cheaper. And that does not work. Manufacturing costs, components cost, controller/dram cost, etc all are the same. And the difference between the price of 128GB and 256GB of flash
×