CPU Performance, Short Form

For our motherboard reviews, we use our short form testing method. These tests usually focus on if a motherboard is using MultiCore Turbo (the feature used to have maximum turbo on at all times, giving a frequency advantage), or if there are slight gains to be had from tweaking the firmware. We put the memory settings at the CPU manufacturers suggested frequency, making it very easy to see which motherboards have MCT enabled by default.

Video Conversion – Handbrake v1.0.2: link

Handbrake is a media conversion tool that was initially designed to help DVD ISOs and Video CDs into more common video formats. For HandBrake, we take two videos and convert them to x264 format in an MP4 container: a 2h20 640x266 DVD rip and a 10min double UHD 3840x4320 animation short. We also take the third video and transcode it to HEVC. Results are given in terms of the frames per second processed, and HandBrake uses as many threads as possible. 

Handbrake v0.9.9 H.264: LQHandbrake v0.9.9 H.264: HQHandbrake v0.9.9 H.264: 4K60

Compression – WinRAR 5.4: link

Our WinRAR test from 2013 is updated to the latest version of WinRAR at the start of 2017. We compress a set of 2867 files across 320 folders totaling 1.52 GB in size – 95% of these files are small typical website files, and the rest (90% of the size) are small 30 second 720p videos.

WinRAR 5.4 Compression Test

Point Calculations – 3D Movement Algorithm Test v2.1: link

3DPM is a self-penned benchmark, taking basic 3D movement algorithms used in Brownian Motion simulations and testing them for speed. High floating point performance, MHz and IPC wins in the single thread version, whereas the multithread version has to handle the threads and loves more cores. For a brief explanation of the platform agnostic coding behind this benchmark, see my forum post here. We are using the latest version of 3DPM, which has a significant number of tweaks over the original version to avoid issues with cache management and speeding up some of the algorithms.

3DPM: Movement Algorithm Tester (Multi-threaded)

Rendering – POV-Ray 3.7.1b4: link

The Persistence of Vision Ray Tracer, or POV-Ray, is a freeware package for as the name suggests, ray tracing. It is a pure renderer, rather than modeling software, but the latest beta version contains a handy benchmark for stressing all processing threads on a platform. We have been using this test in motherboard reviews to test memory stability at various CPU speeds to good effect – if it passes the test, the IMC in the CPU is stable for a given CPU speed. As a CPU test, it runs for approximately 2-3 minutes on high end platforms.

POV-Ray 3.7 Render Benchmark (Multi-Threaded)

Synthetic – 7-Zip 9.2: link

As an open source compression tool, 7-Zip is a popular tool for making sets of files easier to handle and transfer. The software offers up its own benchmark, to which we report the result.

7-Zip 9.2 Compress/Decompress Benchmark

System Performance Gaming Performance
Comments Locked

18 Comments

View All Comments

  • Marlin1975 - Wednesday, October 24, 2018 - link

    Thanks for the Power Delivery Comparison. That's something that is hard to check for if not in a good review. Keep up the good work.
  • gavbon - Wednesday, October 24, 2018 - link

    Thanks Marlin, appreciated! We're looking to further add to this going forward
  • MrCommunistGen - Wednesday, October 24, 2018 - link

    I actually really appreciate this as well because this kind of information has been hard to come by in the past. I had some questions regarding the included table.

    The ASRock B450 Gaming ITX/ac is listed in the table as using the same controller (ISL95712) in "6+2" mode. The spec sheet for ISL95712 states that it is up to a 4+3 controller. The table also doesn't list any doublers. Doesn't this make the board a "big 3+2" rather than a 6+2?

    I didn't look at any of the other boards in the table closely, but I was just doing a bunch of reading on the ASRock ITX boards recently and so I was looking very closely at that line in the table.

    Thanks for the review!
  • TeutonJon78 - Monday, October 29, 2018 - link

    It would help more if it was accurate.

    The ISL95712 maxes out at 4+3 phases, so it can't run in 6+2 as listed in the chart. On the ASRock boards, then run a 3+2 setup but with two sets of MOSFETs and chokes in parallel, but not doubled. There is only one capacitor per phase. It helps keep the temps down by splitting the current, but it's still only a 3 phase design.

    Plus, it would be more meaningful if the chart was for similar style boards rather than a smattering of different form factors.In this case, the other m-ITX boards.
  • DanNeely - Wednesday, October 24, 2018 - link

    Just thinking about POST times, when UEFI was the next big thing we'd been repeatedly promised that it'd allow 1-2s POST times because unlike BIOS it could start components up in parallel instead of 1 at a time.

    While current boards do post faster than their predecessors the speedup never came close to meeting the hype. Does anyone know why reality fell so short of the promise?
  • Dragonstongue - Wednesday, October 24, 2018 - link

    likely because there is so much crud to siphon through before it completes the process
    the same from going to post to OS log screen to fully booted OS

    am sure there are ways to "speed it up" but often those methods come with chance of something terrible happening and not knowing what took place.

    be happy the new stuff is WAY quicker and more energy efficient then the old 9/10 so even if it takes a wee bit longer to boot up, patience come to those whom wait.

    SSD helps that much I know, but as far as only 1 to 2s to finish post..umm I personally never heard of such promises, the board I am using is AM3/3+ M5A99X EVO (v1) which uses a UEFI based bios design and just going from HDD to SSD made post much quicker and a few changes I made as well to speed it up also helped.

    I personally have more issue with how long it takes to shut down then be fully booted up ^.^
  • Vatharian - Wednesday, October 24, 2018 - link

    As fun as it sounds there are 10 year old systems that run on UEFI. I must say, that boot times are much quicker on them, while supporting much wider range of hardware. Feature-wise it's hit or miss, I have 9 year old board that's capable of booting from NVMe, thanks to drivers being easy to punch in, but it won't allow user mode to write to boot manager area, thus linux bootloader installation will fail (I just have to point the path manually from the UEFI setup).
  • JoeyJoJo123 - Wednesday, October 24, 2018 - link

    Ahmdal's Law, in a way. There is a relative speedup to be had, but desktops don't need to be shutdown/restarted extremely often. Often just letting it go into sleep mode is adequate for me, and leaves my workspace in exactly the same place it was before.

    For laptops or other mobile devices, shutting down is more necessary when moving between work environments, (given that sleep mode will drain battery life over time anyways, hibernate isn't ideal and can clog the main storage disk with a file equal in size to your capacity in memory, etc.). For mobile PC devices, I have personally noted that boot times are appreciably fast.

    Also, personal anecdote, but I've had boards that when set to options for "fast booting", not only would it refuse to take in DEL or F11 prompts to get into BIOS screen when booting up, but it would go from a stable (normal) bootup, to a blazingly fast crash/reboot cycle for 5 loops before landing me back on the BIOS page with stock settings. This is before and after BIOS updates.

    Personal opinion, but I just think it's intentional that for the enthusiast PC market that vendors don't _want_ to speed up boot times because the same users buying these motherboards are enthusiasts likely willing or needing to occasionally go into BIOS and change boot order of disks, or do overclocking features, etc. and making sure users can actually get into BIOS easily and reliably (hence a slow boot with ample time to opt-in to boot to BIOS) may have been one of their intended design goals.
  • Vatharian - Wednesday, October 24, 2018 - link

    There is reverse effect in action: amount of stuff that is set to initialize during boot is staggering. I'd asku you to find 1MB legacy BIOS, while 16MB UEFI is not uncommon, and I am in a possession of motherboard that has 512 MB SPI flash for it (not 512 megabits - half a gigabyte). Then, amount of stuff you can boot from went trough the roof. On top of drivers for everything that need to start up and register, there are at least three different frameworks that monitor hardware and need to set up.
    Then, every damn power controller, voltage regulator, thermal zones, OC chips are smart need to boot up too.

    Then, amount of memory skyrocketed - if you remember old BIOS happily running memcheck kilobyte after kilobyte, until it reaches that 64MB, and compare to you 64GB system, that still gets a cursory check, despite memory NOT being 1024x faster.

    You can cut your boot time down, if you properly disable CSM, disable external controllers (or at least disable their boot rom), properly set up boot sequence to UEFI bootloader, and use GPU with a BIOS that supports GOP, and finally enable fast boot, your PC WILL boot in ~2-3 seconds plus OS.

    On the other hand, while handling servers I am accustomed to boot times on the orders of 20 minutes at extreme case.
  • Death666Angel - Wednesday, October 24, 2018 - link

    My system (H87 with NVME boot mod BIOS) gets to windows faster than my 4k DP monitor wakes up from standby. The DVI one is a bit faster. I guess around 3 seconds maybe?

Log in

Don't have an account? Sign up now