Hello, this is my first post and I'm glad to meet you.
Well, it's 6 years since I built my actual dp55kg based system and time has come to jump into SSD world. And... why not do it at beast mode?
http://www.kingston.com/es/ssd/consumer/shpm2280p2 SSD PCIe HyperX Predator - 240 GB, 480 GB | Kingston
I've bought this. And it's working. And it boots Windows 10 without problems. And it rocks. And it makes no noise. And performance is getting me crazy because it cannot run properly.
Here, at spanish Noticias3D fórum, I've posted a Little review and the problem. PCIe x4 slot, announced in the manual as 2.0, is not working at 2.0. In fact I don't know which mode it's working with.
http://foro.noticias3d.com/vbulletin/showthread.php?t=445376&p=5557265&viewfull=1# post5557265 Comunidad N3D
It's an old thread intented to know if the drive could boot or not, because time ago there was a doubt about if the Predator could boot or not. In fact, I wrote to Kingston and they said to me that they did not compromise to guarantee that the drive could work the way I wanted to in my "old" motherboard. Then, i converted that thread into a review and discussion about this Kingston SSD drive.
Help me, please.
Sorry Francisco; there are two overlapping discussions going on and I am mixing them up...
Are you saying that you have a current generation - GTX1060 - graphics card that is working reasonably well both at the BIOS and O/S level? This is good to know. I have not heard of many folks who have been successful - especially at the BIOS level - using current generation cards with boards this old. What specific brand and model number of card is it and what firmware version is it running?
Seeing some issues with PCIe performance is not something that I include in this (working reasonably well) category. I am not sure that the the data being displayed by the BIOS is correct. There was a case that I remember (unfortunately, only vaguely) a year or two back where it was shown to not be correct. Have you displayed this information from within the Windows environment for verification?
But, Esteban, If I change the motherboard, it's obvious, there's a high probability that the problem would dissapear. This thread is intended to find a solution within my DP55KG.
Recently I've built a new pc for my father, with Core i5 and an Asus ITX iB150 based chipset motherboard with M.2 slot. Although I've not tested, I know my Predator SSD will work at full speed.
The graphics assembler is KFA, known as GALAX around the world. He he, let me tell, that the way it's working, is the way I expected to, and is the way expected to because of that logical things due to hardware description of the motherboard, as well as predator ssd should work . Pity I didn't this with my old ATI 5850, but the BIOS always figured out 2,5GT/s.
https://www.techpowerup.com/gpudb/b3730/galax-gtx-1060-ex-oc-black GALAX GTX 1060 EX OC Black | techPowerUp GPU Database
Once I install the Predator SSD into the x8 slot, it shows at BIOS 5GT/s and the speed test shows correct numbers:
NOTICE: Old fotograph, here I have the ATI 5850 installed into x16 slot.
Correct transfer rates.
I am not sure how to answer this. When the board was designed, the concept of PCIe NVMe disk support was not even on the drawing board. There could be all sorts of issues as a result. This board is 6 chipset generations and 7 years out of date. It was end-of-life'd four years ago and has been out of interactive support for three. In the meantime, Intel exited the Desktop Boards business. The developers are gone, dispersed to other organizations and, in some case, other companies (or, like me, retired). There is no one left to investigate this issue. It is what it is; you are not going to solve it, plain and simple. If the board is not performing well enough for you, replace it.
If you want to understand why there is a difference between what happens with the x16/x8 sockets versus the x4 socket, understand that the x16 and x8 sockets expose PCIe lanes output directly by the processor while the x4 connector exposes PCIe lanes output indirectly by the (P55) PCH. The DMI 2.0 bus connecting the processor to the PCH is equivalent in throughput to 4 PCIe lanes. The DMI bus' bandwidth is shared across the x4 connector, the two x1 connectors, the PCI connectors, a 1394a controller, a GBe controller, the internal SATA controllers, a secondary SATA controller, multiple USB 2.0 controllers, the SMBus controller, the LPC Bus controller, the SPI Bus controller and, finally, the HD Audio Codec interface. That's enough interfaces that it completely tired out my arms typing it all in. If you ask me (though I am just making a silly guess), the x4 connector may not have been programmed for PCIe 2.0 because there wasn't deemed to be enough throughput (left) to warrant the need for it.
Thank you for the throughout explanation, Scott.
In case you require any inquiries answered in the future, feel free to reach us here or any other support platform, Fran_Spain