Jump to content
Corsair Community

Corsair Kevin

Corsair Employees
  • Posts

  • Joined

Everything posted by Corsair Kevin

  1. LeDoyen is correct. Each rows' four pins are the same. One row is +12V, the other is ground. The "connectors" are the same, if you pay attention to the keying on the plastic overmold.
  2. Are you still on BIOS v1001? Jordan mentioned he was getting proper speeds with BIOS v1105. Can you try to flash to that version?
  3. Update: In another test one of my own test samples, I saw slightly lower than expected sequential write (not the 1GB/s everyone else has seen), and I used the Windows 10 optimization in disk management. Optimizing the drive in Windows 10 seemed to bring the speeds back to normal, so yes, I agree that this could potentially help. For those who want to try, follow these steps: - Press Win+R - Type "dfrgui.exe" then press enter - Highlight the drive in question - Click the "Optimize" button If you want to make Windows automatically optimize, you can change the scheduling settings.
  4. Was the PSU the only thing you changed?
  5. Are you staying on the rolled back BIOS, Jordan? Also, if you do, can you please keep me updated on the performance? I am more and more inclined to believe it has everything to do with software interaction between firmware, drivers, etc. I really hate seeing that there have been many with the slowdown issue and wish I could do more to help beyond just suggesting RMA'ing and such. :/
  6. Is it a single drive? If you installed NVMe RAID drivers, you can try opening Device Manager, then expanding "Storage Controllers" and uninstalling the "RAID" driver if, and only if, you're certain it's the driver for that drive
  7. It is recommended that you connect however many EPS (CPU power) connectors the motherboard requires. From the motherboard's product page, it looks like you need an 8-pin and a 4-pin. The 8-pin EPS connector splits into two 4-pin connectors on the motherboard side. Ensuring the CPU has proper power delivery will help the CPU reach higher boost clocks with the likes of AMD's Precision Boost.
  8. Thank you! There are so many variables, be it in software and hardware, that can make it somewhat difficult to pinpoint the cause of these errors. That difficulty is only compounded by the fact that in all the units I've received from users on here, the performance is as advertised, whether empty, filled half-way, or even at 90%. I really wish I had more insight into the goings-on of all the cases presented here. :/ Still, I'll keep checking up the thread. I will add your thoughts/notes into the spreadsheet I have with all the issues. Thank you again!
  9. Glad to hear the replacement is working better. Please keep an eye on it and test it as you fill it up. One of the other members had good results from their replacement initially, but saw similar results after a while. Please keep us posted.
  10. I'm back in the office now, sorry for the delay. The results on ATTO are all over the place, which is concerning. Originally, the SSD dev team were thinking that it was just CDM not playing nicely with the new hardware and giving inaccurate readings, though seeing these ATTO results is alarming. Additionally, now that I've been testing more SSDs that were sent to me, I haven't yet been able to replicate the issues described (though I did see some odd sequential read behavior). I'm kind of at a loss of what it can be, since the units I've received aren't exhibiting the same slowdowns, even after multiple, back to back runs. Have any of you tried turning write-back caching on and off? https://support.microsoft.com/en-us/help/324805/how-to-manually-turn-disk-write-caching-on-or-off Running CDM or ATTO as administrator?
  11. I apologize for the inconvenience, and I completely agree with you. It's completely optional and at your discretion, but can you share some comprehensive details on the configurations you've tried? I just want to get a better idea of what's going on. At any point in time, did you test with having only the MP600 connected? You said "after taking the MP600 drive out" when you reinstalled Windows on the 960. Did you leave the MP600 out, or did you also have it occupying the other M.2 slot? Was the SATA SSD still connected as well? I want to rule out any possibility of drive priority/recognition on the motherboard. Any kind of details you can provide would be of great help. Thank you!
  12. If you haven't already, try removing the 960 and having only the MP600 connected, then see if you can go a lengthened period of time without the issue.
  13. If you have submitted an FA RMA on an MP600, can you DM me your name, so I can discuss my findings in more detail with you? Thanks!
  14. Hi everyone, I'm still here. I've just been trying to balance everything out. I have received two more FA RMA units and will test accordingly once some other tests are completed. Should have some comments tomorrow. As of right now, I'm not really sure what's going on. On my testbench (ASRock X570 Taichi), as I've stated, the one RMA'd unit I've personally tested achieved speeds that were on par with the advertised speeds. Again, I'll test the two other units, hopefully tomorrow, and I'll share any findings (i.e. if the results align with what's shown here or if they're proper speeds).
  15. What slots are the actual memory modules seated in?
  16. So, I recently received one unit for the FA RMA and have been running tests. I don't know who the user is on here though. On my testbench, I'm getting performance that are consistent with what we advertise and the same numbers as the units I already had on hand. This particular unit was accompanied by a screenshot showing the 1000MB/s seq write issue, but on my machine it's fine, which honestly kind of points to it being a combination of variables. I'm still running more tests with different fill % and such, and will share any quirks I've found with system configs. My testbench: ASRock Taichi X570 Ryzen 7 3800X 32GB Dominator Platinum RGB 3200MHz
  17. https://www.corsair.com/us/en/psu-cable-compatibility
  18. Vegan is suggesting to pull the metal socket from inside the plastic connector housing, then solder the wire to it. This is the proper way to permanently repair it, as you don't want loose electrical connections. However, I would suggest RMA'ing the PSU at corsair.com/support, especially if you're not familiar with, or comfortable using, a soldering iron.
  19. When these shutdowns occur, is the system idling for a while beforehand? It sounds like it's going to sleep.
  20. While it may not support TCG Opal, there is apparently encryption between the controller and NAND. I personally don't know the full details, but I'm still trying to get more information on it as well.
  21. Directly from our DRAM/SSD team (paraphrased a bit): The MP510 doesn't support hardware encryption. AES 256-bit Encryption refers to software encryption, specifically Windows Bitlocker.
  22. It's the same product. They're the same package with the exception of one having imagery printed directly on it. Using these early packaging blanks, I believe, is a way for us to reduce waste involved with manufacturing and such. Aside from the printing, they're identical.
  23. Hi Rim, did you specifically mention doing an advanced RMA with support? If not, can you PM me your ticket number and I'll get one of the support staff to look it up and I'll see what we can get underway. As for the packaging differences, when we launch completely new products, there's a lot of logistical stuff that goes on in the background, particularly with packaging design. The white boxes were probably the initial packaging units and rather than being wasteful, they were used for retail units with stickers applied. Regarding the IOPS, the numbers could have changed as motherboard manufacturers, AMD, and we release software updates that improve baseline performance.
  • Create New...