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 leave the BIOS settings at default and memory at JEDEC (DDR4-2133 C15) for these tests, making it very easy to see which motherboards have MCT enabled by default.

Blender 2.78: link

For a render that has been around for what seems like ages, Blender is still a highly popular tool. We managed to wrap up a standard workload into the February 5 nightly build of Blender and measure the time it takes to render the first frame of the scene. Being one of the bigger open source tools out there, it means both AMD and Intel work actively to help improve the codebase, for better or for worse on their own/each other's microarchitecture.

Rendering: Blender 2.78

The SLI Plus leads all other boards tested so far by a 10 second margin with a time just over 3 minutes and 10 seconds. The MSI boards both have multi-core turbo enable by default and seems to hold longer on the SLI Plus than the Arctic on this set of tests. 

Rendering – POV-Ray 3.7: 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.

Rendering: POV-Ray 3.7

POV Results show the SLI Plus is in the middle of the motherboards tests, a negligble difference behind the Gaming Pro Carbon AC, while the Tomahawk Arctic runs away with this benchmark holding MCT/E longer in this shorter test. 

Compression – WinRAR 5.4: link

Our WinRAR test from 2013 is updated to the latest version of WinRAR at the start of 2014. 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.

Encoding: WinRAR 5.40

The SLI Plus sits in the middle of the pack for WinRAR with the Pro Carbon AC leading the pack and the othersr more tightly grouped together. Another result of MCT/E. 

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.

Encoding: 7-Zip

The SLI Plus in 7-zip scores last in this test, however nearly all results are within a percentage point or so; a margin of error difference in many cases. 

Point Calculations – 3D Movement Algorithm Test: 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.

System: 3D Particle Movement v2.1

The SLI Plus is bringing up the back up the back here, several percent behind the leader and other peers. Clocks appeared to be the same running this test as the others, including mesh at 2.7 GHz. At this time, we are not sure why there is a lower result compared to the Tomahawk or other boards in the round up.

Neuron Simulation - DigiCortex v1.20: link

The newest benchmark in our suite is DigiCortex, a simulation of biologically plausible neural network circuits, and simulates activity of neurons and synapses. DigiCortex relies heavily on a mix of DRAM speed and computational throughput, indicating that systems which apply memory profiles properly should benefit and those that play fast and loose with overclocking settings might get some extra speed up. Results are taken during the steady state period in a 32k neuron simulation and represented as a function of the ability to simulate in real time (1.000x equals real-time).

System: DigiCortex 1.20 (32k Neuron, 1.8B Synapse)

The SLI Plus climbs back up towards the top of this tightly group set of results from DigiCortex 1.20, though all are very close. 

System Performance Gaming Performance
Comments Locked

20 Comments

View All Comments

  • Lolimaster - Wednesday, November 29, 2017 - link

    *buy
  • BillyONeal - Wednesday, November 29, 2017 - link

    Threadripper has all the same problems. Huge socket == small area for VRM, and 180W TDP. At least in my testing TR 1950X consumes well over 250W when running at 3.9 GHz.
  • Lolimaster - Wednesday, November 29, 2017 - link

    Why would you buy a handycapped X299 when AMD's X399 platform is better in every possible way?
  • mkaibear - Wednesday, November 29, 2017 - link

    Ah, still spreading your usual fear, uncertainty and doubt I see.

    I'll reply the same way I did before;

    https://www.anandtech.com/bench/product/1904?vs=19...

    That's why (note not identical parts because it's a 12 core TR vs an 8 core i7 - but they are as close as I can get in terms of costs). If I went the other way and went with a 10 core i9 vs the 16 core TR then we see roughly the same pattern of behaviour.

    Threadripper wins in the multithreaded tests so long as the workload suits it but for the many benchmarks it's per-core speed which is more important than number of cores.

    In essence, if your work requires fast cores and quite a few threads then you're better off with the i7 or i9, if it utilizes loads of threads but speed is less important then you're better off with the TR.

    So; given that there are obvious use cases for both processors I'm afraid I can't agree that "Threadripper X399 is better in every possible way".

    I note you didn't bother to reply before, but it'll be interesting to see if you manage to string a coherent sentence together this time.
  • mkaibear - Friday, December 1, 2017 - link

    ...hey look at that, he doesn't bother to reply. What a shock...
  • BillyONeal - Wednesday, November 29, 2017 - link

    Because you have a workload Threadripper isn't as good at? Building the compiler my 7980XE box is ~40% faster than my 1950X box. TR has much better price/performance but that's not an end-all, be-all metric.
  • Lolimaster - Wednesday, November 29, 2017 - link

    Even better when the 1950X is $800 right now.
  • BillyONeal - Wednesday, November 29, 2017 - link

    Agreed; at that price the 1950X invalidates the 1920X even more than it used to. (Saving $200 on a CPU when the rest of the system costs ~1.5k seems like a waste)
  • BenJeremy - Thursday, November 30, 2017 - link

    More VROC nonsense. Officially, you are not supposed to be able to buy the VROC upgrade key - vendors are supposed to provide you with a means to get one. Until you PAY MORE ($126 for VROCSTANMOD), you will be stuck with a hobbled system that cannot create bootable RAID-0 arrays unless you use Optane SSDs.

    Meanwhile, X399/ThreadRipper offers bootable RAID-0 for your NVMe array for FREE and it's been demonstrated at higher speeds than Intel's promised with VROC.
  • Edward190 - Saturday, December 2, 2017 - link


    https://www.cartoonhdappz.com/

Log in

Don't have an account? Sign up now