The ASUS X399 ROG Zenith Extreme Motherboard Review: Top Tier Threadripper
by E. Fylladitakis on July 17, 2018 10:30 AM EST- Posted in
- Motherboards
- Gaming
- Asus
- ROG
- Overclocking
- ThreadRipper
- X399
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 for these tests, making it very easy to see which motherboards have CPU core enhancements enabled by default.
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.
In this test we can see that ASUS is pushing the Ryzen Threadripper 1950X the most by default, with the X399 ROG Zenith Extreme landing at the top of our chart. The performance difference in this test is relatively small, yet clearly measurable. Note that the three faster motherboards in this chart are all temporarily overclocking the processor above 4 GHz by default, with the difference being that ASUS is just pushing it significantly harder.
Rendering - LuxMark v3.1: link
As a synthetic, LuxMark might come across as somewhat arbitrary as a renderer, given that it's mainly used to test GPUs, but it does offer both an OpenCL and a standard C++ mode. In this instance, aside from seeing the comparison in each coding mode for cores and IPC, we also get to see the difference in performance moving from a C++ based code-stack to an OpenCL one with a CPU as the main host.
In this tests we are also getting about the same results as before, with the ASUS X399 ROG Zenith Extreme landing at the top of our performance charts, outperforming the competition by a small, yet measurable margin.
Rendering - 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 – 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.
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.
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).
Facebook
Twitter
RSS
25 Comments
View All Comments
jordanclock - Tuesday, July 17, 2018 - link
It's nice to have drivers available in situations where you either don't have immediate internet access or where you cannot access the internet because your OS install doesn't include necessary drivers, like for graphics or network adapters.twtech - Tuesday, July 17, 2018 - link
This is far from the worst example, but not having built a system in a few years - it seems like I've come back to find every motherboard now looks like some blinged-out kids toy from the dollar store. Lights, weirdly-shaped plastic shrouds with big gamer logos that have no functional purpose, etc.Yuriman - Wednesday, July 18, 2018 - link
The retail PC segment is in decline. Manufacturers are doing everything they can to attract a new generation of buyers.Awful - Tuesday, July 17, 2018 - link
I have no use case that requires 32 cores, 10Gb networking, or even wifi in a desktop so I wouldn't buy one of these. But I still waaannnt one!Alien959 - Wednesday, July 18, 2018 - link
I want to comment about m.2 cooling. According to numerous respected internet outfits and even the JEDEC specification and testing, nand flash lasts longer at higher temperatures and is recommended to be above 40c. The only part that needs cooling is the controller, but almost all cooling plates cool the nand.kazoOC - Thursday, July 19, 2018 - link
Notice one thing: most boards share the m2 heatsink with the PCH, which is responsible for pci and sata. It will always get relatively warm while still keeping the m2 drives away from throttling.Now, aftermarket m2 heatsink are another story but still a valid choice in poorly ventilated cases. Just take care to avoid contact with nand chips by peeling off portions of the thermal pad.
virpuain@gmail.com - Wednesday, July 18, 2018 - link
A quick glare suggests this VRM will be dissipating 30W at 1.4V@176A. That being said the 32 cores TR better come with a maximum TDP of 250W.For the pricetag this board is lacking alot in the VRM side of things, especially if you consider this a very premium board ( with AAA+++ premium pricetag ) with a VRM that is actually worse than what you have on a few AM4 boards like the X370 Taichi and C7H.
This VRM is pretty much like what you have with the X370 GT7, a $ 120 board for AM4.
Oxford Guy - Thursday, July 19, 2018 - link
ASUS and Gigabyte both sold hybrid air/water VRM cooling, beginning in 2013 with ASUS — for quad core CPUs. But, no — we don't need water cooling for VRMS on a board like this. Instead, we need tiny fans and LEDs.a351must2 - Wednesday, July 18, 2018 - link
I just have to comment on this ... I have this motherboard and would say the heatsink for the included 10GB card is necessary. I've actually been having problems with the 10GB network dropping offline randomly and traced it down to the card overheating. The heatsink is big, but it also needs some airflow near it (my case provides none). I now have an additional slot fan mounted that moves air across the 10GB card and my other addon cards (older intel dual GB nic and a 9211 sas controller).Also, if this review had been done 9 months ago when I built mine, there would've been some mention of the buggy bios and memory support. I made the mistake of buying unsupported DDR3200 ram (Corsair kit for Intel) and after reading a bit was feeling lucky I got it to run at 3000. A bios update in January got it to run at 3200 though and I believe most of the fan speed control issues have been resolved.
As for power, when messing with overclocking mine I've had the 1950X draw over 350 watts by itself using this board. That'll easily cover a 250Watt TDP 32 core processor ... I'm sure we'll see the limits when the new processors become available though.
Timur Born - Friday, July 20, 2018 - link
Thanks for the article.I find the DPC Latency section misleading. There is no mention what power profile and BIOS (C-states) settings were used and if the values reported are maximum or average values. Look at this example:
Highest measured interrupt to DPC latency (µs): 354,514229
Average measured interrupt to DPC latency (µs): 2,480283
Quite the difference. And 354 µs maximum still is not a problem even at lowest audio buffer settings.