2017 Hardware Thread

Cliff S said:

The newer CPU's are not necessarily faster or better.

• They have better power management like digital thermometer for Skylake & Kaby Lake(they save more power or throttle when the CPU gets too warm, not good to desktop users that OC, but keep the CPU cooler for most other users),
• more instruction sets(i.e. codecs>hello 4K Netflix (only with the 7th gen)),
• tighter/better security features(hello SGX, MPEX, DEP, and SME),
• better virtualization features(SVE & SVM),
• and better general CPUID features.

Here is a video from Linus on New Years Eve 2016 ya'll might find interesting:

Good info there.

Anyway this is why I feel people should be looking more towards the tech rather than pure performance numbers. When I look at my still running Haswell system, gaming or real world performance usage, I can't say it's noticeably slower than my Kaby lake system. Benches may tell a "slightly" different story, but real word "feel", not so much. Would you notice a few frames here or there when gaming? Would you notice a few Nano seconds here or there performance wise? I seriously doubt it.

Tech wise though, especially motherboard tech (Z87 vs. Z270) is where Haswell falls short - native USB 3.1, Thunderbolt 2, native NVMe support, more PCI lanes to name a few.

Gaming wise the either system games as well as the other. Benchmark wise, it's mostly about the GPU rather than CPU. So yeah, in my opinion it's now about the motherboard tech and CPU instructions, rather than flat out performance.

All that said, I'd rather have a 2017 system with a new CPU, than a 2013 one :)

My two cents.

OldMike65 said:

Well Drew you did a beautiful job building your new rig, Very good cable management also. That is my sore spot for me, as I really need to get better at that part. Nice Job!!!!

Thanks Mike,

Didn't have to do any re-cabling with this this since all I had to do was replace the CPU and MB. The cabling was actually done when I did the Skylake build last year. Check this post - Thoughts on Phanteks’ Enthoo Luxe PC case. Features two systems.

Thanks.

sygnus21 said:

Good info there.

Anyway this is why I feel people should be looking more towards the tech rather than pure performance numbers. When I look at my still running Haswell system, gaming or real world performance usage, I can't say it's noticeably slower than my Kaby lake system. Benches may tell a "slightly" different story, but real word "feel", not so much. Would you notice a few frames here or there when gaming? Would you notice a few Nano seconds here or there performance wise? I seriously doubt it.

Tech wise though, especially motherboard tech (Z87 vs. Z270) is where Haswell falls short - native USB 3.1, Thunderbolt 2, native NVMe support, more PCI lanes to name a few.

Gaming wise the either system games as well as the other. Benchmark wise, it's mostly about the GPU rather than CPU. So yeah, in my opinion it's now about the motherboard tech and CPU instructions, rather than flat out performance.

All that said, I'd rather have a 2017 system with a new CPU, than a 2013 one :)

My two cents.

I always thought, with gaming, it wasn't maximum FPS, but a nice steady FPS(so no micro stuttering occurs), with the minimum FPS optimally being the monitor's refresh rate.
Anything above the refresh rate is just, literally, hot air.
At least the was mas my understanding as a non-gamer:)

This one is for Andrew, for selecting a GPU:

listen closely starting at 1:44

Cliff S said:

I always thought, with gaming, it wasn't maximum FPS, but a nice steady FPS(so no micro stuttering occurs), with the minimum FPS optimally being the monitor's refresh rate.

Yeah, that's true, but using my comparison between my Haswell and Kaby system - if using the same GPU, there's no performance difference - micro stuttering, etc... This is why I say games are more GPU than CPU dependent.

In short, given the same GPU, there's really no difference between the last iteration of CPU's where gaming is concerned - notably Haswell, Skylake, Kaby.

Cliff S said:

I always thought, with gaming, it wasn't maximum FPS, but a nice steady FPS(so no micro stuttering occurs), with the minimum FPS optimally being the monitor's refresh rate.
Anything above the refresh rate is just, literally, hot air.
At least the was mas my understanding as a non-gamer:)

Yes, overall ease of use and higher graphics in more powerful systems make it easier and smoother. When I played original MOHA series on some dual core and 510MB GPU it was much more difficult. Now hen I tried it on this system, not only that I can play it it at top graphics and details but seams much easier to play without getting my a** being blown up 3 times in same mission.

I'm going to try Der8auer's way of finding a good OC for my card. On my 1080 Aero 8GB OC it is the power limit that gets me most, sometimes temp, and rarely, the voltage limit(I leave them showing in sys tray from Afterburner's monitoring).

Cliff S said:

I'm going to try Der8auer's way of finding a good OC for my card. On my 1080 Aero 8GB OC it is the power limit that gets me most, sometimes temp, and rarely, the voltage limit(I leave them showing in sys tray from Afterburner's monitoring).

1st thing I noticed in your screen shot are your temps.....47° for Avg .... Right now mine are at around 31° Just seems kinda warm. I do have 5 fans plus Water Cooled thou. :) and I'm at 4.9 also.

OldMike65 said:

1st thing I noticed in your screen shot are your temps.....47° for Avg .... Right now mine are at around 31° Just seems kinda warm. I do have 5 fans plus Water Cooled thou. :) and I'm at 4.9 also.

There are way too many variables to really compare those numbers in a given moment. Was your system idle while Cliff's was running. What's you room temps vs. Cliff's? What did you have running at time reading was taken vs. Cliff's. System configuration, etc... You get the point :)