The AMD 3rd Gen Ryzen Deep Dive Review: 3700X and 3900X Raising The Bar
by Andrei Frumusanu & Gavin Bonshor on July 7, 2019 9:00 AM EST** = Old results marked were performed with the original BIOS & boost behaviour as published on 7/7.
Benchmarking Performance: CPU Legacy Tests
3DPM v1: Naïve Code Variant of 3DPM v2.1
The first legacy test in the suite is the first version of our 3DPM benchmark. This is the ultimate naïve version of the code, as if it was written by scientist with no knowledge of how computer hardware, compilers, or optimization works (which in fact, it was at the start). This represents a large body of scientific simulation out in the wild, where getting the answer is more important than it being fast (getting a result in 4 days is acceptable if it’s correct, rather than sending someone away for a year to learn to code and getting the result in 5 minutes).
In this version, the only real optimization was in the compiler flags (-O2, -fp:fast), compiling it in release mode, and enabling OpenMP in the main compute loops. The loops were not configured for function size, and one of the key slowdowns is false sharing in the cache. It also has long dependency chains based on the random number generation, which leads to relatively poor performance on specific compute microarchitectures.
3DPM v1 can be downloaded with our 3DPM v2 code here: 3DPMv2.1.rar (13.0 MB)
x264 HD 3.0: Older Transcode Test
This transcoding test is super old, and was used by Anand back in the day of Pentium 4 and Athlon II processors. Here a standardized 720p video is transcoded with a two-pass conversion, with the benchmark showing the frames-per-second of each pass. This benchmark is single-threaded, and between some micro-architectures we seem to actually hit an instructions-per-clock wall.
CineBench 11.5 and 10
Cinebench is a widely known benchmarking tool for measuring performance relative to MAXON's animation software Cinema 4D. Cinebench has been optimized over a decade and focuses on purely CPU horsepower, meaning if there is a discrepancy in pure throughput characteristics, Cinebench is likely to show that discrepancy. Arguably other software doesn't make use of all the tools available, so the real world relevance might purely be academic, but given our large database of data for Cinebench it seems difficult to ignore a small five minute test. We run the modern version 15 in this test, as well as the older 11.5 due to our back data.
Facebook
Twitter
RSS
447 Comments
View All Comments
Tkan215 - Monday, July 8, 2019 - link
true the future is more cores. People and customers should feel awake that single core aint the future its just a stopping rock. more cores !Tkan215 - Monday, July 8, 2019 - link
yes i called it a tie because of the margin of error and patches were not taken into account. also, Intel get enormouse game support so really many factors as they are not equal playing groundwatzupken - Sunday, July 7, 2019 - link
Intel's bad moment just started. Clearly while there are some areas where Intel chips are still doing well, however the victories are significantly lesser now. Looking at the power metrics, they lost the fab advantage, so they are now in the disadvantage. To top it off, Intel is still charging monopolistic prices on their existing chips. Have not really seen the rumored price cuts, which may be too little and too late.StrangerGuy - Sunday, July 7, 2019 - link
IMO the $200 CPU landscape is now buy 3600 non-X, or get ripped off by Intel anything even if the latter for cheaper by $50.mikato - Thursday, July 11, 2019 - link
Yeah I really wish a 3600 was tested.Maxiking - Sunday, July 7, 2019 - link
Intel is waiting for 10nm, considering the fact AMD didn't even match Skylake prepatches performance... IF Intel fixes the 10nm, AMD will be be smashed to the ground. If it is a big if, but it is a fact.Mahigan - Sunday, July 7, 2019 - link
AMD actually beat Intel on a clock for clock basis now. What you're seeing is Intel's higher boost clocks saving the day (somewhat).If Intel can't go past 5GHz with their 10nm, due to the new core design, and only are able to get say 10-15% more performance per clock then Gen3 Ryzen will most likely end up, with its 7nm+ and improvements AMD aren't done making, in tough competition.
just4U - Sunday, July 7, 2019 - link
Intel won't be doing any smashing anytime soon there Max.. I was damn pleased with the overall value/performance of my 2700x in comparison to my highly overclocked 8700K (4.9Ghz) and basically shrugged of the 9 series intel. The addition of a 12core.. with great performance levels really changes the game.Even if Intel brings something out it's not going to destroy anything. All we've seen over the past 5 years is small bumps upwards in performance.
Korguz - Sunday, July 7, 2019 - link
Maxiking intel has been waiting for 10nm for 204 years now.. and they are still kind of waiting for it. skylake prepatch ? as in specture and meltdown ? um.. kind of need those fixes/patches in place, even if it means a performance hit.. but by all means.. get skylake, dont fix/patch it, and worry about that.. and spend more.. its up to you... either way.. zen2.. looks very good....Targon - Monday, July 8, 2019 - link
What RAM was used in the Intel system? The Ryzen system used DDR4-3200, but it's CL16, not CL14 RAM. That CAS latency difference would be enough for Ryzen to at least tie the 9900k if not beat it in the gaming tests.