The AMD Zen and Ryzen 7 Review: A Deep Dive on 1800X, 1700X and 1700
by Ian Cutress on March 2, 2017 9:00 AM ESTThe Ryzen Die
Throughout the time leading up to the launch of Ryzen, AMD reaffirmed its commitment to at least +40% IPC improvement over Excavator. This was specifically listed as a goal relating to performance, at an equivalent energy per cycle, resulting in a 40% increase in efficiency. At the Tech Day, AMD listed an overall 2.7x (or 270%) performance per watt improvement, split into the following:
Obviously a number of benefits come from moving the 28nm TSMC process to GloFo’s 14nm FinFET process which is used via a Samsung license. Both the smaller node and FinFET improvements have been well documented so we won’t go over them here, but AMD is stating that Zen is much more than this as a direct improvement to immediate performance, not just efficiency. While Zen is initially a high-performance x86 core at heart, it is designed to scale all the way from notebooks to supercomputers, or from where the Cat cores (such as Jaguar and Puma) were all the way up to the old Opterons and beyond, all with at least +40% IPC.
The first immediate image out of the presentation is the CPU Complex (a CCX), which shows the Zen core design as a four-CPU cluster with caches. This shows the L2/L3 cache breakdown, and also confirms 2MB of L3 per core with 8 MB of L3 per CCX. It also states that the L3 is mostly inclusive of the L2 cache, which stems from the L3 cache as a victim cache for L2 data. AMD is stating that the protocols involved in the L3 cache design allow each core to access the L3 of each other core with an average (but range) of latencies.
Over the next few pages, we’ll go through the slides. They detail more information about the application of Simultaneous Multithreading (SMT), New Instructions, the size of various queues and buffers, the back-end of the design, the front-end of the design, fetch, decode, execute, load/store and retire segments.
Facebook
Twitter
RSS
574 Comments
View All Comments
EchoWars - Thursday, March 2, 2017 - link
No, apparently the failure was in your education, since it's obvious you did not read the article.Notmyusualid - Friday, March 3, 2017 - link
Ha...sharath.naik - Thursday, March 2, 2017 - link
I think you missed the biggest news in this information dump. The TDP is the biggest advantage amd has. Which means that for 150watt server cpu. they should be able to cram a lot more cores than intel will be able to.Meteor2 - Friday, March 3, 2017 - link
^^^This. I think AMD's strength with Zen is going to be in servers.Sttm - Friday, March 3, 2017 - link
Yeah I can see that.UpSpin - Thursday, March 2, 2017 - link
According to a german site, in games, Ryzen is equal (sometimes higher, sometimes lower) to the Intel i7-6900K in high resolution games (WQHD). Once the resolution is set very low (720p) the Ryzen gets beaten by the Intel processor, but honestly, who cares about low resolution? For games, the probably best bet would be the i7-7700K, mainly because of the higher clock rate, for now. Once the games get better optimized for 8 cores, the 4-core i7-7700K will be beaten for sure, because in multi-threaded applications Ryzen is on par with the twice expensive Intel processor.I doubt it makes sense to buy the Core i7-6850K, it has the same low turbo boost frequency the 6900K has, thus low single threaded performance, but at only 6 cores. So I expect that it's the worst from both worlds. Poor multi-threaded performance compared to Ryzen, poor single threaded performance compared to i7-7700K.
We also have to see how well Ryzen can get overclocked, thus improving single core performance.
fanofanand - Thursday, March 2, 2017 - link
That is a well reasoned comment. Kudos!ShieTar - Thursday, March 2, 2017 - link
Well, the point of low-resolution testing is, that at normal resolutions you will always be GPU-restricted. So not only Ryzen and the i7-6900K are equal in this test, but so are all other modern and half-modern CPUs including any old FX-8...The most interesting question will be how Ryzen performs on those few modern games which manage to be CPU-restricted even in relevant resolutions, e.g. Battlefield 1 Multiplayer. But I think it will be a few more days, if not weeks, until we get that kind of in-depth review.
FriendlyUser - Thursday, March 2, 2017 - link
This is true, but at the same time this artificially magnifies the differences one is going to notice in a real-world scenario. I saw reviews with a Titan X at 1080p, while many will be playing 1440p with a 1060 or RX480.The test case must also approximate real life.
khanikun - Friday, March 3, 2017 - link
They aren't testing to show what it's like in real life though. The point of testing is to show the difference between the CPUs. Hence why they are gearing their benchmarking to stress the CPU, not other portions of the system.