Operating a GPU at maximum performance using a 16-slot setup with 4x bandwidth and Gen 2 speeds
Operating a GPU at maximum performance using a 16-slot setup with 4x bandwidth and Gen 2 speeds
I've installed my GTX 660 OEM 1.5GB into the PCI-e 3.0 x16 slot of a Dell Vostro 270 motherboard to maximize bandwidth. I recently acquired an Intel DQ77MK board with a PCI-e x16 slot, but it operates at x4 speed and Gen 2.0 performance. Placing my GPU there would significantly limit its speed due to bandwidth constraints. The original board supports on-board discrete graphics running at x16 Gen 3 speeds, though it needs an integrated GPU. My current setup uses a Xeon E3-1230 V2 CPU, which lacks an integrated graphics solution. Would it be feasible to route the GPU through the dedicated PCI lanes for discrete graphics instead of using the slower x4 Gen 2 configuration? My current specs include 2x8GB RAM at 1600MHz.
This data is sourced from GPU-Z. A screenshot is available if needed. The manual indicates receiving 16 PCI-E lanes from the CPU in the x16 slot, excluding onboard graphics.
Ensure all additional parts are connected; this can limit the available data rate per lane. Refer to the user guide for the schematic and confirm slot availability.
I haven't received the new motherboard yet, but I've included a screenshot of the GTX 660 on the Dell Vostro 270 board. Once I get the new one, I'll try it out and share the bandwidth results. I just noticed an error from when I bought it—it claimed a x4 Gen 2 speed 16x slot.
The motherboard features a mini-PCI/MSATA port that I plan to utilize for Wi-Fi. I’ll keep you updated once it’s delivered.
I just got the motherboard and it features two X16 slots, though one is actually an X4 slot with a cutout on the PCIe side to accommodate the X16 connector. All the data transfer is functioning properly now.