Date: November 25, 2019
Author(s): Rob Williams
The day many have been waiting for has arrived: AMD and Intel have officially launched their respective next-gen enthusiast processors. In this article, we’re going to follow-up on our 3950X Linux article from last week to introduce both Intel’s Core i9-10980XE and AMD’s Ryzen Threadripper 3960X and 3970X into the test results.
As we did with the launch of AMD’s Ryzen 9 3950X, we’re going to kick off our performance coverage of the company’s newest Threadripper chips, along with Intel’s just-released Core i9-10980XE, under Linux.
We’re going to be following-up on this article with a full Windows-focused view of AMD’s new Threadrippers soon (we already tackled the 10980XE), but since our Linux testing can complete much faster, we’ve gone this route to get performance up in time for embargo.
On the Intel side, getting the i9-10980XE to work with Linux wasn’t even a concern, because the platform itself has remained unchanged. In fact, we’ve been using the same Intel X299 motherboard since the original Core X-series launch (7000-series). This platform is well taken care of support-wise, but the new TRX40 platform from AMD might throw a wrench into that whole process of “booting”.
You might recall that at the Zen 2 launch, Linux was simply unbootable until some fixes were released. This was due to an issue that impacted some Windows software, as well, but acted as a true roadblock for many Linux users, since it was the systemd manager that had to be patched. Today, the situation on Zen 2 is a lot better, but Threadripper has introduced a brand-new caveat.
To get Linux to boot on third-gen Threadripper right now, a boot flag of mce=off (machine check error) needs to be added. The fix for this was provided to me by Michael from Phoronix, who went through the painstaking effort of figuring out what the temporary fix even was. AMD itself doesn’t seem to know of Linux issues when the CPUs are handed out, because reviewers are left to find unavoidable roadblocks on their own.
We haven’t talked to AMD about this issue yet, but as with the Zen 2 launch, we’d expect the company to help fixes get to market sooner than later. We miss the days of simply building a new PC and booting up, but it’s hard to fault AMD too much with the limited number of engineers it has – something that has got to improve with the ongoing success of its Zen products.
Aside from all of that, the topic of conversation for Intel here is the new i9-10980XE, to see how it compares not just against its competition, but also its predecessor, 9980XE. On the AMD side, the Zen 2-based 3960X and 3970X Threadrippers have the potential to be absolute beasts. The Zen 2 architecture itself has proven to be a genuinely impressive jump over the original, and it’s hard to believe that Zen 3 is apparently already on track, and is expected to bring even more notable gains.
Here are both AMD’s and Intel’s lineups for those who want to see where these new chips are placed:
|AMD’s Ryzen & Ryzen Threadripper Lineup|
|3970X||32 (64T)||3.7 GHz (4.5)||144MB||Quad||280W||$1999|
|3960X||24 (48T)||3.8 GHz (4.5)||140MB||Quad||280W||$1399|
|2990WX||32 (64T)||3.0 GHz (4.2)||16+64MB||Quad||250W||$1799|
|2970WX||24 (48T)||3.0 GHz (4.2)||12+64MB||Quad||250W||$1299|
|2950X||16 (32T)||3.5 GHz (4.4)||8+32MB||Quad||180W||$899|
|2920X||12 (24T)||3.5 GHz (4.3)||6+32MB||Quad||180W||$649|
|R9 3950X||16 (32T)||3.5 GHz (4.7)||72MB||Dual||105W||$749|
|R9 3900X||12 (24T)||3.8 GHz (4.6)||70MB||Dual||105W||$499|
|R7 3800X||8 (16T)||3.9 GHz (4.5)||36MB||Dual||95W||$399|
|R7 3700X||8 (16T)||3.6 GHz (4.4)||36MB||Dual||65W||$329|
|R5 3600X||6 (12T)||3.8 GHz (4.4)||35MB||Dual||95W||$249|
|R5 3600||6 (12T)||3.6 GHz (4.2)||35MB||Dual||65W||$199|
|Ryzen w/ Radeon Vega Graphics|
|R5 3400G||4 (8T)||3.7 GHz (4.2)||0.5+4MB||Dual||65W||$149|
|R3 3200G||4 (4T)||3.6 GHz (4.0)||0.5+4MB||Dual||65W||$99|
|Intel Processor Lineup|
|Cores||Clock (Turbo 3.0)||L3||Memory||IGP||TDP||Price|
|i9-10980XE||18 (36T)||3.0 GHz (4.8)||24.75MB||Quad||No||165W||$979|
|i9-10940X||14 (28T)||3.3 GHz (4.8)||19.25MB||Quad||No||165W||$784|
|i9-10920X||12 (24T)||3.5 GHz (4.8)||19.25MB||Quad||No||165W||$689|
|i9-10900X||10 (20T)||3.7 GHz (4.7)||19.25MB||Quad||No||165W||$590|
|i9-9980XE||18 (36T)||3.1 GHz (4.5)||24.75MB||Quad||No||165W||$1,979|
|i9-9960X||16 (32T)||3.5 GHz (4.5)||22MB||Quad||No||165W||$1,684|
|i9-9940X||14 (28T)||3.8 GHz (4.5)||19.25MB||Quad||No||165W||$1,387|
|i9-9920X||12 (24T)||3.4 GHz (4.5)||19.25MB||Quad||No||165W||$1,189|
|i9-9900X||10 (20T)||3.5 GHz (4.5)||19.25MB||Quad||No||165W||$989|
|i9-9820X||10 (20T)||3.8 GHz (4.5)||16.5MB||Quad||No||165W||$898|
|i9-9800X||8 (16T)||3.8 GHz (4.5)||16.5MB||Quad||No||165W||$589|
|i9-9900KS||8 (16T)||4.0 GHz (5.0)||16MB||Dual||Yes||127W||$513|
|i9-9900K||8 (16T)||3.6 GHz (5.0)||16MB||Dual||Yes||95W||$480|
|i9-9900||8 (16T)||3.1 GHz (5.0)||16MB||Dual||Yes||65W||$439|
|i7-9700K||8 (8T)||3.6 GHz (4.9)||12MB||Dual||Yes||95W||$374|
|i7-9700||8 (8T)||3.0 GHz (4.7)||12MB||Dual||Yes||65W||$323|
|i5-9600K||6 (6T)||3.7 GHz (4.6)||9MB||Dual||Yes||65W||$262|
|i5-9600||6 (6T)||3.1 GHz (4.6)||9MB||Dual||Yes||65W||$213|
|i5-9500||6 (6T)||3.0 GHz (4.4)||9MB||Dual||Yes||65W||$192|
|i5-9400||6 (6T)||2.9 GHz (4.1)||9MB||Dual||Yes||65W||$182|
|i3-9350K||4 (4T)||4.0 GHz (4.6)||8MB||Dual||Yes||91W||$173|
|i3-9320||4 (4T)||3.7 GHz (4.4)||8MB||Dual||Yes||62W||$154|
|i3-9300||4 (4T)||3.7 GHz (4.3)||8MB||Dual||Yes||62W||$143|
|i3-9100||4 (4T)||3.6 GHz (4.2)||6MB||Dual||Yes||65W||$122|
There is not much else that can be said here; the only thing left lingering is the actual performance results. As of the time of this article’s publishing, we’re foregoing our usual commentary on each set of results due to time, although they may be slipped in later. We’re trying to get more performance content done ahead of travel this weekend; this really has proven to be an incredible month for launches and benchmarking.
|Techgage Workstation Test System(s)|
|Processors||AMD Ryzen Threadripper 3970X (32C/64T; 3.7GHz)|
AMD Ryzen Threadripper 3960X (24C/48T; 3.8GHz)
AMD Ryzen Threadripper 2990WX (32C/64T; 3.0 GHz)
AMD Ryzen Threadripper 2950X (16C/32T; 3.5 GHz)
AMD Ryzen 9 3950X (12C/24T; 3.8GHz)
AMD Ryzen 9 3900X (12C/24T; 3.8GHz)
AMD Ryzen 7 3700X (8C/16C; 3.6GHz)
AMD Ryzen 5 3600X (6C/12C; 3.8 GHz)
AMD Ryzen 5 3400G (4C/8T; 3.7 GHz)
Intel Core i9-10980XE (18C/36T; 3.0GHz)
Intel Core i9-9980XE (18C/36T; 3.0GHz)
Intel Core i9-9900KS (8C/16T; 4.0 GHz)
Intel Core i9-9900K (8C/16T; 3.6 GHz)
Intel Core i9-8700K (8C/16T; 3.7 GHz)
|Motherboards||AMD X399 (TR3gen): ASUS ROG Zenith II Extreme|
AMD X399 (TR2gen): MSI MEG CREATION
AMD X570: Aorus X570 MASTER
AMD B450: Aorus B450 Pro WiFi
Intel Z390: ASUS ROG STRIX Z390-E GAMING
Intel X299: ASUS ROG STRIX X299-E GAMING
|Cooling||AMD X399 (TR3gen): NZXT Kraken X62|
AMD X399 (TR2gen): Enermax LIQTECH TR4 240mm
AMD X570: Corsair Hydro H115i PRO RGB
AMD B450: AMD Wraith Spire
Intel Z390: Corsair Hydro H100i V2
Intel X299: NZXT Kraken X62
|Chassis||AMD X399: Cooler Master MasterCase H500P Mesh|
AMD X570: Fractal Design Define C
AMD B450: Enermax Equilence
Intel Z390: NZXT S340 Elite
Intel X299: Corsair Carbide 600C
|Graphics||NVIDIA TITAN RTX|
|Memory||G.SKILL Flare X (F4-3200C14-8GFX)|
4x8GB; DDR4-3200 14-14-14
|Et cetera||Ubuntu 19.10 (5.3.0-23 kernel)|
Our Linux configuration is simple overall. We’re using the latest version of Ubuntu (19.10) for our testing here, updated as much as possible. It includes the 5.3 kernel and GCC 9.2.1. A lot of our tests are run with the help of Phoronix Test Suite, with tests locked to a specific version to generate reliable data. We don’t do any real OS configuration outside of disabling sleep, and enforcing the performance power profile.
As for platform configurations, all of them have been tested with the same 32GB kit of DDR4-3200 memory, and outside of enabling the memory spec, the motherboard UEFIs are left stock. An exception to that rule is with ASUS motherboards on Intel platforms. On those, Multi-Core Enhancement is disabled to prevent a vendor-specific automatic overclock. MCE is not the same as regular Intel Turbo, and it needs to be off for true apples-to-apples results. The tested AMD boards do not have the same vendor-specific automatic overclocking feature.
Beyond all of that, all of the platforms have been updated to their latest UEFI, which for AMD means the latest AGESA release is reflected. No special patches are installed for the sake of security mitigation; we simply take the updates that Ubuntu provides us by default.
We’re going to be tackling both video encoding and scientific performance on this page, and while that sounds like only two things, the reality is that there are a countless number of workloads available in each one, and loads of opportunity to see interesting scaling.
We unfortunately forgot about Blackmagic’s RAW Speed Test having a Linux build, so we’ll make sure that’s included next time. What’s new here video-wise is Intel’s SVT encoder libraries, all of which are open-source. For the latter half of the page, we’ll look at some scientific scenarios.
The previous pages have been focused on specific scenarios, like science, rendering, and compiling, so to wrap things up, this page is going to revolve around some more general scenarios, like encryption, compression, and memory bandwidth. For good measure, a chess engine even makes an appearance.
We’ve been lacking more single-threaded tests in the past, but have introduced some SciMark-sourced ones here. We’re still not sure if they are ideal, but they serve the purpose for now. As always, if you have a suggestion of another test we should implement, please leave a comment.
There’s a lot to look over here, considering we’re taking a look at not just one CPU, but three of them from two different vendors. Let’s start with the easiest of the two to talk about: Intel’s Core i9-10980XE.
This Cascade Lake-X chip is an interesting one, largely because it doesn’t exactly pull that far ahead of the outgoing 9980XE. In some cases, performance was actually worse on the newer chip, something that highlights just how close they are to performance. Both chips flip-flopped a lot throughout the review, leading us to feel that they are “the same”. At least roughly.
We’d be remiss to ignore the fact that the 10980XE does support up to 256GB of memory, double that of the 9980XE. It also bumps the officially rated memory speed to 2933, from 2666, although that’s not exactly easy to get excited over that when the new Threadrippers are spec’d for 3200. AMD included a DDR4-3600 kit with the review samples (not used for our apples-to-apples tests), highlighting how confident the company is in Zen 2 being able to handle higher-than-the-usual clocks.
The biggest feature of Intel’s new chip is simply the fact that it costs just under $1,000. The last-gen 9980XE retailed for $2,000, so the AMD effect has definitely been felt, and it’s consumers who benefit. We’d wager Intel’s chip is still priced a bit high considering how close the 3950X gets to it (if not beating it in some cases), for $250 less. Market mindset is going to keep Intel safe for a while, but AMD seems to be rapidly changing things, quicker than normal.
And speaking of Threadripper, as we saw throughout most of the results here, the third-gen parts are a force to be reckoned with. We already have some Windows testing completed, and we can give a quick example of how the new 32-core scores over 16,000 in Cinebench’s multi-thread score, while the old one hit about 11,500. Plus, the new generation has the ability to use even faster memory than the old one, which in our tests does in fact improve things further in some cases.
As always, what processor is right for you depends entirely on whichever workloads are important to you. If we don’t cover the one that you care about, please let us know. We do have a couple of suggestions already to add, and will do so the next time we update our suite (this current update is not even a month old!).
Simply put, the new 32-core Threadripper 3970X impresses the hell out of us. When we ran the first test and saw its result in comparison to the last-gen 32-core, our mouths literally dropped. We didn’t expect as many gains as we saw, especially on the rendering front. We definitely expected improvements with encoding, due to an improved memory design, but what we’re actually seeing is improvements all over the place. The 24-core likewise shows great performance, and while we didn’t have the last-gen 24-core in our charts here for comparison, we can tell from the 32 vs. 32 core comparison that big improvements have been made here. NUMA mode handling has also been improved greatly, although there are still some software tweaks that can be made to fully utilize Threadripper.
The new 32-core costs $200 more ($2,000) than the last-gen part, and that had some scratching their heads. How could AMD actually raise prices? Well, considering how that last-gen part has an internal design that detrimentally impacted certain workloads, and the new one doesn’t have those issues, and it also greatly improves the architecture – the newest Threadripper processors are absolute powerhouses.
Copyright © 2005-2019 Techgage Networks Inc. - All Rights Reserved.