Sorry McSteel, for slightly hijacking (unintentionally) your thread. Hopefully it helped provide you with some important info in relation to this build though.
Nothing to be sorry about, it's a sub-topic that is indeed relevant to both this build and my interests. Your contributions have helped, thanks for that!
Anyone who wants VT-d support (1155 socket), regardless if you have a non-K series CPU or not, will require a board with the Q67 or Q77 chipset, not the 'enthusiast' Z77 chipset.
Back to this, as it begs the question: Who would actually
want VT-d, and why?
Namely, VT-d is an addition to already existing VT-x (among other VM-oriented instruction sets), and it's there for four reasons.
1. I/O dev assignment. Basically this provides the administrator(s) with a hardware-assisted ability to distribute I/O device resources amongst running VMs, as per the control/communication needs. It's basically a way to implement a software-emulated and soft-managed KVM switch which works in much the same way a hardware one would work with actual hardware. Neat, but is it really needed?
2. DMA remapping. This means that an admin can isolate a DMA address range and assign it exclusively to a VM. It also means that there will be less overhead when accessing a DMA-capable device (an HDD, for example), but only when the device itself hasn't exhausted the number of parallel pipelines it can service. Again, neat, but is it a must-have?
3. IRQ remapping. This is probably the most useful of the bunch. It isolates everything from everything, interrupt-wise, while also eliminating the need to doubly translate IRQs between host and guest, which again reduces overhead. Would be nice to have, but not essential.
4. Reliability & Reporting support. Namely, this is a hard-soft interface, similar to Intel Management Engine for motherboards/chipset, it should help with usage tracking and error logging/debugging. Useful when things go wrong or when optimizing, but still not something one can't live without.
Add to this that none of these features are of any use in a standard Host OS <-> Guest OS VM configuration and it looks like I actually don't need to take virtualization instructions into considerations. It's only really beneficial to VMs that are run in Hybrid or Hypervisor mode. Since I won't be running Hyper-V (no need to) nor will I run a hybrid kernel, Q77/67 would be wasted on me. Besides, the only MoBos available here that carry those chipsets are Gigabyte GA-Q77M-D2H, ASUS P8Q67-M DO/USB3/TPM/SI (long-winded name) and Intel DQ77MK and DQ77CP. None of these look very appealing to me. And I don't mean the aesthetics.
Perhaps Intel's logic was just that. Only those who intend to run 4-10 VMs with a Hypervisor kernel would fully benefit from VT-d, so Intel simply does away with it where it doesn't really belong - the Gamer, Enthusiast and SOHO environments.
Thanks for all the help so far, guys! I'm happy that I asked for opinions here, and I hope I can be of at least some use to you in the future.
Anyone who wants VT-d support, regardless if you have a non-K series CPU or not, will require a board with the Q67 or Q77 chipset, not the 'enthusiast' Z77 chipset. Sigh. I searched high and low for this information when I researched my build, and yet I find it many months after the fact.... goddammit
I knew with Intel it wouldn't be that simple.
