【AMD Zen 处理器测试曝光，IPC接近Haswell】AMD Zen Engineering Sample Benchmarks Leak Out – Summit Ridge CPU Faster Than The Intel Core i5 4670K In AotS Benchmark
4790四核Boost 3.8GHz，4670K四核Boost 3.6GHz，换算过来多的线程在同频下带来了4670K 118.4%的性能，也就是大概4.72个核心
那么按照之前帖子说过的 Zen ES 8核满载3.05GHz的话，换算过来IPC会是Haswell的86%。
This is the day we have been waiting for. The first verified benchmarks of the upcoming AMD Zen processor have just leaked out via the Ashes of the Singularity benchmark database. This is the first indication of real world performance which does not originate from AMD’s marketing material. That said however, the benchmarks given in this article are that of the Engineering Sample of AMD Zen and NOT entirely representative of the final product. As such, the benchmarks only serve to give us a preliminary understanding of what we can expect from the final product.
【等了这么多天终于等到了 - AMD Zen 处理器的首个奇点灰烬性能测试。这是Zen架构的首个实际性能指向，而不是ppt。然而，这测试的Zen属于工程样品，不能完全代表最终产品的性能。所以这个测试只代表产品的初步性能。】
Preliminary Zen processor benchmarks are out – show impressive IPC gains over AMD’s FX-8350
【早期Zen处理器性能测试 - IPC大幅提升】
A few days ago, something remarkable happened. The OPN (Order Part Number) of AMD’s Zen Engineering Sample leaked out. This was 1D2801A2M88E4_32/28_N. The first part of the number indicates the unique identifier of the Zen ES CPU while the second part (32/28) reveals the base clock as well as the turbo clock. This means we are looking at a base clock of 2.8 Ghz and a turbo of 3.2 Ghz.It goes without saying that the frequencies of the engineering sample are probably not final and could get an uplift by the time the product actually hits the shelf.
【几天前Zen处理器的订单编号曝光 - 1D2801A2M88E4_32/28_N，(32/28) 代表基础频率2.8，Boost频率3.2。可以肯定的是，最终产品频率会更高】
With this context, we continue on to the main spectacle of today: Zen ES benchmarks. The exact flavor is the Ashes of the Singularity benchmark. We can quickly confirm that the benchmarks are indeed of a Zen processor because:
【然后是今天的主菜：Zen ES版 奇点灰烬测试。从上面的信息很容易判断这是Zen处理器，因为：】
It lists the OPN Zen has previously been identified with (1D2801A2M88E4_32/28_N),【订单编号是之前曝光过的】
It lists another OPN that we have yet to identify Zen with but appears to be a different revision of the same (2D2801A2M88E4_32/28_N)【还有另一个处理器是不同版本的订单编号】
So in short, we have at least two variants of Zen being tested here. I’ll refer to them as 1D and 2D.【有两种版本，我称之为1D和2D】
We can see that the profile has previously tested Carrizo engineering samples of AMD – so this isn’t a newly minted spoofer profile.【测试者之前测过CarrizoES】
So without any further ado, here are the benchmarks:
Right of the bat, we can see that the Zen 1D ES sample is performing rather well. Across all 4 settings, the CPU is able to earn a consistent CPU frame-rate of 58 frames per second. The benchmark in question was conducted with the AMD Radeon RX 480, so it goes without saying that we cannot make direct comparisons from just this benchmark alone. We can however, compare this setup to similar setups where only the CPU has been replaced (which we have done so down below).
【立马可以发现的是，Zen 1D ES性能很不错。四种画质下都能保持58FPS。单独一个测试不能直接比较性能，但我们可以用不同CPU的配置来比较。】
There doesn’t seem to be any difference at first sight between the 1D and 2D variants. They are both purportedly clocked at the same rates and the only difference that we can see is the fact that the 2D variant fares much worse in terms of performance. There could be multiple reasons for why this is the case. It is possible that this particular variant is actually a cut down core and the benchmark is reading it incorrectly or even that its simply another revision being tested that did not like the AotS benchmark (or was simply tested before being optimized). In any case, because of these reasons, I decided to focus on the 1D variant for our comparison.
Zen ES Performance Comparison (AotS Benchmark)【Zen ES 性能比较】
Since the Zen benchmark was conducted using an RX 480 graphics card, I was able to find comparable configurations (RX 480 GPU + Standard 1080p Preset) and see how the CPU compares to other processors of this tier. The first thing you will notice is that the Zen ES variant is clocked pretty low. This means that the benchmark is somewhat skewed right from the start. I did not bother to adjust the numbers for clock rates because AotS database does not record the exact frequency of the CPU being tested.
Even with a clear clock speed handicap the Zen ES shows impressive performance.
We should however be able to get a rough idea of how Zen fares among its competitors. The Zen ES sample is immediately faster than most of the competition. Even with a base clock of 2.8 Ghz, we can see that it manages to beat out the Intel Core i5 4670k which has a base clock of 3.4 Ghz. It is a full 38% faster than the AMD FX-8350 and around 10% faster than the i5 4670k. However, in its current state, and in this particular benchmark, the Zen ES 1D variant is around 11% slower than the Core i7 4790 (which has a base clock of 3.6 Ghz).
If by the time Zen ships, its clock speeds are similar to that of the FX 8350, it should easily be able to go toe to toe with the Intel’s i7 offerings (if this benchmark is any indication of how we can expect performance to scale across other applications). If it is able to achieve a clock speed of 4.0 Ghz, which is a 25% increase over the current clocks, it should be able to easily out perform Intel’s i7 CPUs (the i7 4790 in this case). According to recent updates from AMD’s Lisa Su – Zen will be shipping in limited quantities by year end, with full market availability in 2017. With that, this particular write up comes to a close. I have included screenshots of all the benchmarks used in this article, including the 2D variant and the processors used for comparison.
【如果Zen正式出货时，频率能接近FX 8350的话，就能赶上Haswell 四核i7（如果这次的性能提升属实）。如果能达到4.0Ghz，就超过了i7。这篇文章说明 Zen在今年底的少量出货离我们又近了一步。】
[Update: the tester has pulled all the Zen ES benchmarks form the online database so please refer to the screenshots posted in the gallery above] All the benchmarks used in this article can be freely accessed by the links given below:
Zen ES (Variant 1D):
Zen ES (Variant 2D):