The AMD Ryzen Threadripper 1950X and 1920X Review: CPUs on Steroids
by Ian Cutress on August 10, 2017 9:00 AM ESTCPU Office Tests
The office programs we use for benchmarking aren't specific programs per-se, but industry standard tests that hold weight with professionals. The goal of these tests is to use an array of software and techniques that a typical office user might encounter, such as video conferencing, document editing, architectural modeling, and so on and so forth.
All of our benchmark results can also be found in our benchmark engine, Bench.
Chromium Compile (v56)
Our new compilation test uses Windows 10 Pro, VS Community 2015.3 with the Win10 SDK to compile a nightly build of Chromium. We've fixed the test for a build in late March 2017, and we run a fresh full compile in our test. Compilation is the typical example given of a variable threaded workload - some of the compile and linking is linear, whereas other parts are multithreaded.
One of the interesting data points in our test is the Compile, and it is surprising to see the 1920X only just beat the Ryzen 7 chips. Because this test requires a lot of cross-core communication, the fewer cores per CCX there are, the worse the result. This is why the 1950X in SMT-off mode beats the 3 cores-per-CCX 1920X, along with lower latency memory support. We know that this test is not too keen on victim caches either, but it does seem that the 2MB per core ratio does well for the 1950X, and could explain the performance difference moving from 8 to 12 to 16 cores under the Zen microarchitecture.
PCMark8: link
Despite originally coming out in 2008/2009, Futuremark has maintained PCMark8 to remain relevant in 2017. On the scale of complicated tasks, PCMark focuses more on the low-to-mid range of professional workloads, making it a good indicator for what people consider 'office' work. We run the benchmark from the commandline in 'conventional' mode, meaning C++ over OpenCL, to remove the graphics card from the equation and focus purely on the CPU. PCMark8 offers Home, Work and Creative workloads, with some software tests shared and others unique to each benchmark set.
Strangely, PCMark 8's Creative test seems to be failing across the board. We're trying to narrow down the issue.
SYSmark 2014 SE: link
SYSmark is developed by Bapco, a consortium of industry CPU companies. The goal of SYSmark is to take stripped down versions of popular software, such as Photoshop and Onenote, and measure how long it takes to process certain tasks within that software. The end result is a score for each of the three segments (Office, Media, Data) as well as an overall score. Here a reference system (Core i3-6100, 4GB DDR3, 256GB SSD, Integrated HD 530 graphics) is used to provide a baseline score of 1000 in each test.
A note on context for these numbers. AMD left Bapco in the last two years, due to differences of opinion on how the benchmarking suites were chosen and AMD believed the tests are angled towards Intel processors and had optimizations to show bigger differences than what AMD felt was present. The following benchmarks are provided as data, but the conflict of opinion between the two companies on the validity of the benchmark is provided as context for the following numbers.
Facebook
Twitter
RSS
347 Comments
View All Comments
Lolimaster - Friday, August 11, 2017 - link
Adding useless singe thread benchs for people that will pay $600+ for many cores is plain stupidity like that useless open a pdf test.casperes1996 - Saturday, August 12, 2017 - link
Why do you bother replying to these, Ian? I love your enthusiasm about what you do, and am happy that you reply to comments, but as you state yourself, no matter what you say, you'll be called a shill on more than a weekly basis by either side no matter what you do. Intel shill, AMD shill, Apple shill, Nvidia shill and so on. There's no stopping it, because you just can't please the people who go into something wanting a specific result. Well, you can if you give them that result, but sometimes, facts aren't what you want them to be, and some people don't accept that.Cheers, mate
Diji1 - Thursday, August 10, 2017 - link
You sound like a crazy person.Notmyusualid - Saturday, August 12, 2017 - link
@ Diji1You are correct.
He is implying just because we want HEDT platforms / chips, that we care nothing for single-threaded performance.
He is a true AMD fan-boi, as you will see over time.
lordken - Thursday, August 10, 2017 - link
@Johan Steyn: while I agree with you that the Intel piece with PR slide at the top was a little bit lame, I even lolled at "most scalable" part (isn't something like "glued" zen the most scalable design?) I think this review is good and goes also around architecture etc., there were few instances during reading when it seemed odd wording or being unnecessarily polite toward intel's shortcoming/deficit but I cant even remember them now.Though I was surprised about power numbers, as Toms measured much higher W for 7900X , 160-200 and with TTF even up to 250-331 , but here 7800/7900x had only ~150W.
Also this sentence is odd
"All the Threadripper CPUs hit around 177W, just under the 180W TDP, while the Skylake-X CPUs move to their 140W TDP." move to their? They are above the TDP...why not state it clearly?
tamalero - Thursday, August 10, 2017 - link
Im sratching my head on power consumption as well. Almost all reviewers shows that the i9's consume more than threadripper.Could be the motherboard used?
Some used the Zenith and other reviewers used the ASROCK retail one.
smilingcrow - Thursday, August 10, 2017 - link
Power consumption can vary a lot depending on the type of task and the exact nature of that task.So you should expect a lot of variation across reviews.
Johan Steyn - Thursday, August 10, 2017 - link
The amount of tests on games in this review is unbalanced. Also read my reply to Ian.Extremely well stated: " unnecessarily polite toward Intel's shortcoming" Sometimes I think these guys think all are complete mindless drones.
carewolf - Thursday, August 10, 2017 - link
It has always been like this here. This was pretty neutral by Anandtech standards, they even admitted it when it was faster.Johan Steyn - Thursday, August 10, 2017 - link
Please read my response to Ian, I think you are not looking close enough to what is happening here.