The AMD Ryzen 7 1800X Performance Review

[SKYMTL]

Could this finally be it? Throughout the course being drip fed information about AMD’s new Ryzen architecture over the last eighteen months, that question has been bouncing through the depths of my mind. Would this be the CPU architecture that finally pushes the missteps of Phenom K10 and Bulldozer out the proverbial window and leverages AMD back into competition with Intel? Could AMD relive its K7 and K8 microarchitecture glory days? Based on the slivers of hope delivered at countless press days and briefing calls the technology which was first explained as “something new”, then as Zen and Summit Ridge and now finally as Ryzen took shape.

A review of Ryzen can’t begin without bringing up the painful past of high expectations and subsequent crushing reality checks of AMD’s last five or so years. The Bulldozer, Steamroller, Piledriver and Excavator architectures effectively failed to live up to the hopes that Intel’s juggernaut would face some competition. Meanwhile, the efforts that were put into the APU Fusion initiative fizzled out in short order when the performance of those aforementioned processor cores acted as a dragging anchor for an entire lineup.

Despite those challenges, AMD persevered and had a hallelujah moment as they slowly but surely engineered a core technology that would leapfrog them back into a position of competitiveness. So here I am today and after taking this new architecture through the wringer and I couldn’t be happier. At the risk of blowing this review wide open at its onset let me say this: the wait for Ryzen and its Zen architecture has been well worth it.

At the onset of designing Zen about four years ago, AMD set out an extremely ambitious goal for themselves: they wanted to increase instructions per clock over their Excavator cores by about 40%. At first you may scoff at a number like that but remember Excavator was based upon the half-decade old Bulldozer architecture. In other words, it isn’t too far of a stretch to expect an intergenerational leap of 40% between a current processor design and one that’s five years old.

Now a 40% boost for Excavator in relation to what Intel was accomplishing with Broadwell, Skylake and Kaby Lake would certainly have put AMD into direct competition with some of the best CPUs on the market. However, as time went on there was a realization that Zen could accomplish quite a bit more. The end result is an incredible 52% IPC increase over Excavator. Not only should this be enough to overcome many of Intel’s current generation processors but Team Blue may be doing some serious soul searching about their upcoming 8th generation Core products as well.

Even before launch, there has been a bit of confusion over the naming scheme of AMD’s newest processors but it’s actually pretty straightforward. The Ryzen part of this equation is simply the model name designation for the enthusiast and prosumer level CPUs based upon the Zen microarchitecture. Meanwhile, the “7” point towards the model segment with –for the time being- Ryzen 7 being the top flight 8-core, 16-thread parts while other series like Ryzen 5 will be released later this year with 6-core (12 thread) and 4-core (8 core) variants.

Moving onto the 1700X in the example above with the 1xxx being a generation indicator and the 7 referring to the performance level of that particular chip. At launch, AMD will have 1800-series and 1700-series CPUs but expect lower tiers to quickly roll out as well. The double zero here is being used as extra space in case there’s additional speed bins of a processors launched in the future.

The last letter (or lack thereof) in this whole equation is perhaps the most important. An “X” means the processor utilizes the maximum implementation of AMD’s Extended Frequency Range (XFR) while the lack of a letter here denotes a so-called standard desktop processor. It doesn’t end there either. “G” means the processor has an integrated GPU while “T” and “S” are reserved for low power desktop processors without and with GPUs. Finally the “H”, “U” and “M” series will be reserved for high performance, standard and low voltage mobile CPUs.

With all of that setup out of the way, let’s take a look at what AMD is initially launching within the Ryzen lineup. Naturally, there are some common threads running throughout. The flagship Ryzen 7 series will consist of eight core, 16 thread (yes, AMD has finally enabled simultaneous multithreading) processors with 4MB of local L2 cache and 16MB of shared L3 cache along with DDR4 2400MHz support. Make sure you read page 5 of this article since that 2400MHz spec is tenuous at best due to official memory frequency support ranging from 1866MHz to 2667MHz depending on the type of DIMMs you use.

The key points of differentiation lie in the clock speeds with the $499 Ryzen 7 1800X featuring a Base clock 3.6GHz which runs up to 4GHz (or 4.1GHz if there’s thermal room for XFR to kick in). AMD bills this as the least expensive 16-thread processor on the market and it also happens to be -on paper at least- the most power efficient with a TDP of just 95W. Compare this to Intel’s $1100, 140W i7-6900K and it isn’t hard to see where this distinction comes from.

The Ryzen 7 1700X is effectively the same processor as the Ryzen 7 1800X but due to its slightly lower operational frequencies it will retail for just $399. Personally I think this will likely be the sweet spot of the Ryzen 7 lineup for the time being.

AMD’s initial launch is rounded out by the Ryzen 7 1700 which has a significantly lower Base clock than its bigger brothers but that also leads to a miniscule TDP of just 65W. With this processor you also get a lower XFR ratio of 50MHz versus 100MHz on the X-series parts but that $329 price point could compare favorably to Intel’s own $350 i7-7700K.

Much of AMD’s focus for Ryzen’s marketing efforts has been directed towards its price / performance ratio against Intel’s similar offerings. It is more than evident that Intel has been taking advantage of a commanding lead in the x86 market to keep processor prices high while rolling out limited innovations from one generation to the next. No better example of this can be seen than the $1100 i7-6900K and $1700 i7-6950X. AMD on the other hand is launching their Ryzen series with some extremely disruptive price positioning.

Disruptive is a great term but while they may seem inexpensive to everyone who has been force-fed Intel’s pricing structure for the last few years, all of these Ryzen 7 processors represent the most expensive AMD desktop offerings in recent memory. Like it or not $500 is still A LOT to pay for a processor. Hence why I think the 1700-series may quickly become the darlings of this particular launch.

For those of you who are used to AMD providing a simple drop-in upgrade solution for your age-old motherboard, start thinking of Ryzen as a top-to-bottom system upgrade. The new 1331-pin AM4 socket isn’t compatible with previous generation CPUs and you’ll need new DDR4 memory but this platform will form the basis of many AMD platforms to come. While Summit Ridge Ryzen CPUs are compatible right now, expect upcoming Raven Ridge APUs (which also use the Zen architecture alongside Vega graphics) to share the same AM4-based foundation. Indeed, many current AM4 motherboards have simple display outputs so they can be used for APUs as well. This is a big step away from the bifurcated and disjointed AM3/FM2 approach of yesteryear.

Before I move on from this already-too-long introduction, I need to talk about availability since that happens to be the biggest question surrounding every launch these days. AMD has already launched pre-orders for all of the CPUs I have mentioned above with an availability date of March 2nd. There’s no question AMD needs a strong launch to backstop what looks like a great architecture. Retailers and my contacts within the distribution channels have indicated there will be sufficient quantities of the Ryzen 7 1800X to fill pre orders while the other processors may initially be in short supply.

Regardless of intangibles like availability, the Ryzen launch is finally here and if you got to this point without jumping ahead to the benchmarks, congrats! There’s much more to this review than a simple introduction!

 

[SKYMTL]

The Zen Architecture; An Overview

The Zen Architecture; An Overview

While AMD has done a great job in detailing their Zen architecture, this review is a perfect place to go over some of its highlight points as they pertain to Ryzen processors. Remember, Zen will be around in some shape or form for at least the next half decade so it was imperative that AMD design it to not only excel in current workloads but also position its feature set to deal with future tasks as well.

The path to Zen has always followed four primary goals: the architecture needed to have significantly improved single threaded performance, simultaneous multithreading was an absolute necessity, it needed to boast great efficiency and module scalability had to be improved.

For the enthusiast market, the Ryzen 7 processors will hold a preeminent flagship place for the forseeable future. Their cores make use of an advanced 14nm FitFET manufacturing process and pack a total of 4.8 billion transistors. Unlike past AMD CPU designs, other than the eight physical cores, a lot of I/O capabilities have been built into Ryzen in an effort to streamline platform design and speed up critical communication pipelines.

Other than 16 native Gen3 PCI-E lanes, there’s also four more lanes dedicated to NVMe or SATA storage solutions and four USB 3.1 Gen1 ports. All of these will play a key role in future motherboard designs.

The scalability aspect of AMD’s goals was achieved by creating a very simple modularized building block called the CPU Complex or CCX. Each of these CCX’s contains four cores which use simultaneous multithreading technology to process up to eight concurrent threads in parallel, 64K of L1 cache, 512KB core-specific L2 cache and 8MB of general L3 cache which can be shared across all four cores.

These CPU Complexes can be used individually as a simple high efficiency four core, eight thread part or combined to make larger, more capable processors for higher end markets. Meanwhile, individual cores within each CCX can be disabled without impacting overall performance metrics. For example, Ryzen 7 has two of these modules while Ryzen 5 makes use two CCX’s but disables a pair of cores to create a 6-core, 12-thread CPU. The possibilities really are endless.

Tying the CCX’s together is AMD’s newfound Infinity Fabric which is essentially a high speed interconnect that’s meant to facilitate on-die communications and aid with the integration of other components like onboard graphics or sound solutions.

In order to address the performance end of the equation, AMD very much focused on the way Zen goes about executing its workloads. Not only has the instruction scheduler been significantly expanded but its resource pool has also been augmented. In plain English this means the scheduler can send information to the execution units at a much quicker pace than in previous designs.

There has been a lot of talk about machine intelligence and deep learning in the last year as scientists from all over the world over attempt to build computer networks that can think for themselves. For their part AMD has taken some of those highbrow concepts and have built an artificial network -albeit a simple one that isn’t Terminator-level smart- inside the Zen microarchitecture.

Called Neural Net Prediction, it builds a model of the decisions driven by software code execution and anticipates future needs, can pre-loads instructions and then choose the best path through the CPU for workloads. As such, a Zen-based processor could get faster over time as it “learns” your usage habits.

Every modern processor has some form of prefetch algorithm built into its design but AMD is hoping to take this to the next level with Smart Prefetch. This is an effort to boost execution stream performance so data can be fed through the core at a faster, more efficient pace. Smart Prefetch is supposed to anticipate the location of future data accesses by applications and then utilizes a high level algorithm to learn application data access patterns and model its responses in parallel. It will then prefetches vital data into local cache so it’s ready for immediate use.

As we already mentioned in the CCX description above, there have been some pretty major revisions to cache hierarchy. When combined with the new prefetcher, these changes allow for a lower level cache nearer to the core netting up to 5X greater cache bandwidth into a core.

 

[SKYMTL]

Going Deep With SenseMI, XFR & More

Other than the antiquated manufacturing process being used for their production, current generation AMD processors haven’t benefitted in any way from the advancements being rolled into successive APU generations. Remember, today’s highest-end CPU the FX-9590 uses a Piledriver microarchitecture that was introduced nearly a half decade ago. Meanwhile both Steamroller and Excavator have been launched without any paralleling updates within the FX-series.

As you can imagine this has all led to Zen incorporating countless improvements in the power efficiency front. Actually I misspoke; AMD isn’t actually trying to make their next generation processors consume less power. Rather it’s all about allowing them to achieve maximum performance while minimizing power.

In order to achieve these goals the aforementioned adaptive power and clock gating technologies have been combined under three primary terms under the SenseMI umbrella: Pure Power, Precision Boost and XFR.

In order to achieve these goals the aforementioned adaptive power and clock gating technologies have been combined under three umbrella terms: Pure Power, Precision Boost and XFR.

Very much like the older AMD initiatives like PowerTune and Enduro, Pure Power monitors core operations in real time with hundreds of sensors which log temperature, speed and voltage. It then optimizes the silicon responsiveness to enhance efficiency across all P-States while also allowing the chips to hit their idle rates much faster.

While the Pure Power algorithm adaptively manages and logs various core functions it works hand in hand with Precision Boost to maximize operational frequencies. The “precision” part of this equation is due to the Boost algorithm’s ability to accomplish extremely fine-grain 25MHz clock speed adjustments in an effort to squeeze every last ounce of attainable performance out of the silicon. In addition, each core can be clocked (or parked) individually depending upon the situation.

Those minute 25MHz increments work in parallel with the aforementioned sensors that pull data from the core every millisecond. Voltage can also be adjusted on the fly at 0.6mV for even more granularity. The end result is a CPU design that can respond extremely quickly to changing workloads and whose frequency over time looks more like that of a modern GPU. Indeed, we can see how AMD is leveraging their education in the GPU world to make more efficient x86 processors.

By this point it should be obvious that every Zen-based processor will feature an effective frequency range denoted by a Base and Precision Boost clock. However, AMD didn’t stop there and have implemented something of an ode to enthusiasts with XFR. This so-called Extended Frequency Range is supposed to scale clock speeds depending upon core temperatures and thus will reward cooler running systems with a single thread clock speed that could extend above the chip’s stated maximum Precision Boost limit. This is a very important distinction to make: XFR is only meant for single thread workloads.

XFR is included on all Ryzen processors, but to different extents. Processors with the "X" designation can extend their Boost frequency by 100MHz while all other non-X parts could see up to a 50MHz increase under the right conditions.

 

[SKYMTL]

Say Hello to the AM4 Platform

Say Hello to the AM4 Platform

AMD’s 990FX flagship platform may have been around for the better part of six years, but much of its epic staying power is an attestation of how forward looking it was back in 2011. It had PCIe lanes aplenty so adding third party controllers was easy, memory support was broad and pricing remained respectable. Naturally, some minor revisions like PCIe 3.0 compatibility were added over time but for the most part the 990FX boards stood the test of time.

With Ryzen comes a new AM4 socket and also an updated chipset lineup with some very interesting offerings. More importantly, these new processors take a page out of the APU playbook and integrate a large amount of I/O functionality directly onto the die, thus acting like a quasi-SoC.

Even if you don’t read any further, make sure you remember this small bit of information: The AM4 platform is meant to serve as a foundation for both Ryzen CPUs and upcoming Raven Ridge APUs. Hence, many supporting motherboards will integrate display outputs into their designs. AMD also expects this particular platform to be around until at least 2020 provided next generation technologies like PCI-E 4.0 and the advent of high speed DDR5 don’t necessitate a pin-out change to the processors themselves.

Each Ryzen 7 processor has 16 native PCI-E 3.0 lanes which are dedicated for graphics use. On a higher end X370 motherboard, these can be configured towards either a single x16 graphics solution or a dual x8 partition. This design is exactly the same as what Intel has been pushing on their more mainstream Z-series platforms for years now. It can also be considered a step backwards from 990FX which had the capability to run dual x16 slots, much like the enthusiast X-series Intel chipsets.

This PCI-E lane layout is actually quite odd considering AMD has routinely highlighted their ongoing support for three or more Radeon graphics cards. NVIDIA meanwhile has stepped away from triple and quad card setups, rather deciding to focus their efforts on delivering strong single / dual GPU configs. Some vendors could launch more capable motherboards with PLX PCIe multipliers but that may be some time off.

There’s plenty of other functionality here too. AMD has packed in four additional on-die Gen3 PCIe lanes that are set aside for high bandwidth NVMe or SATA storage solutions and guarantees they communicate directly with the processor over a dedicated bus. Four USB 3.1 Gen1 (the artist formerly known as USB 3.0) are also included which will come into play with the upcoming X300-based boards, but more on that later.

While the processor itself has more than enough connectivity for basic needs, the X370 chipset is where most of the fun happens on this particular platform. It is connected to the processor or APU via a quartet of PCI-E lanes for quick communications and includes eight general purpose PCI-E 2.0 lanes and two native USB 3.1 Gen2 ports alongside six USB 3.1 Gen1 and six USB 2.0 connections. That native USB 3.1 Gen2 support is a key element here since partitioning chipset-bound PCI-E lanes and adding third party controllers won’t be necessary. This could keep board costs down.

On the mass storage side of the equation, there’s a pair of SATAe ports which can be effectively segmented into four standard SATA Gbps connections if a motherboard vendor doesn’t want to utilize that already-dead standard. Finally, AMD has added four native SATA 6Gbps ports as well.

While much of this page has been dedicated to the enthusiast level X370, there will be plenty of other motherboard options cascading downwards into lower price points. The B350 for example will offer very similar functionality when compared to X370 while still offering overclocking. Granted, there are a few reductions in available USB 3.1 Gen 1, SATA and general purpose PCI-E connections but other than the loss of dual graphics capabilities, this chipset looks pretty capable.

The A320 chipset is the budget-focused member of this particular family and there’s good reason for this. It receives I/O cuts across the lineup and it won’t support overclocking in any way. This is all to be expected from a chipset that will target entry level markets.

In my opinion the most interesting addition to this lineup has been conveniently pushed to the bottom of the chart you see above: the dedicated small form factor “chipsets”. This is one of the first times an entire desktop motherboard range is being conceived at a platform level and the way AMD has gone about this is quite interesting. Since the Ryzen 7, 5 and 3 processors feature basic built-in I/O functions (USB 3.1. Gen1 and four PCI-E lanes dedicated to high bandwidth storage) and mini ITX form factors typically don’t require tons of storage options AMD decided to revise their chipset design.

That revision took on an extreme form since the X300 and A/B300-based motherboards don’t have a secondary chipset at all. Rather, in the place of the traditional chipset is a simply Trusted Platform module that handles security, BIOS management and other baseline tasks.

Since there isn’t a chipset per se, the quartet of Gen3 PCI-E lanes that are normally used for chip to chip communications between the Ryzen CPU and its associated controller have been repurposed. These can now be used by motherboard vendors to feed into third party controllers for key features like USB 3.1 Gen2, SATA, WLAN / Bluetooth and so on. This is also where that ALC1220 codec comes into play since it can be easily enabled to provide a high fidelity sound solution on a compact mini ITX platform.

It will be really interesting to see what motherboard vendors can accomplish with this new type of simplified enthusiast platform.

 

[SKYMTL]

Memory Compatibility - Ryzen's Achilles' Heel?

Memory Compatibility - Ryzen's Achilles' heel?

Back in the bad old days when every memory module had a green PCB and no heatspreader, you may have noticed that the labels on those modules had extra information that looked like 1Rx8, 2Rx8, 1Rx4, or similar. That information largely disappeared from consumer-oriented memory kits due to higher quality memory controllers that supported a greater range of memory configurations. Regrettably, with Ryzen, those who want to use four memory sticks or just want to run at high memory clocks are going to have do their homework.

This fact is revealed by this simple table:

The key focus point here is Rank, which is simply a way to describe how memory chips are grouped together to form a 64-bit interface on a memory module. Achieving a 64-bit data bus width is important because modern mainstream processors have two (or four in the case of Intel's Broadwell-E) 64-bit memory buses, which are otherwise known as dual channels, one channel for each bank of two slots.

Currently, the most common rank configuration is eight 8-bit (x8) memory chips, connected together into a rank, and typically (but not always!) located on one side of the PCB. But what if you need to create a larger capacity module but only have those 8-bit memory chips? You need to separate the memory chips into two groups/ranks, each of which is 64-bit and each of which needs to be accessed asynchronously (i.e., the ranks cannot be accessed simultaneously). In doing so, you have just created a dual rank memory module.

For consumers, rank is a somewhat confusing term, especially when a module is labelled as 1Rx8 or 2Rx8. The number in front of the R refers to the numbers of ranks - single rank, dual rank, or even quad rank - and the number after the X is the data bits width of the memory chips, either x4, x8, or x16. For our purposes, we can exclude both quad rank and x4 since those two elements are only found on registered ECC memory intended for servers or workstations. In fact, we can even forget about x16, since it is borderline impossible to find, and not used by any of the popular memory manufacturers.

So essentially, a memory module with 1Rx8 written on it is single-ranked and uses x8 (8-bit) memory chips. Whereas one with 2Rx8 on the label is dual-ranked and also uses x8 (8-bit) memory chips. The drawback with using higher ranked modules is that modern CPUs have memory controllers that have a maximum number of ranks that they can address.

Increasing the number of ranks that need addressing causes increased load on the memory controller and thus decreases the memory speed that it can handle... and that is what we are seeing in the table above. Clearly, Ryzen has a somewhat weak memory controller when compared to the one found in Intel's Kaby Lake processors. By comparison, those Intel chips natively support four dual-rank DIMMs at DDR4-2400, and since Intel is extremely conservative in the memory department, can actually handle those four DIMMs at overclocked speeds of up to DDR4-3600. Maybe AMD are being equally conservative, but based on early reports, running dual rank memory kits above DDR4-2400 is proving difficult at this time.

How do you avoid dual rank memory kits? Well, first and foremost, most memory manufacturers will be launching AMD-specific models immediately. However, there are also a few other ways of determining whether a memory kit is single-ranked or dual-ranked, and this is worthwhile information especially for those who currently own DDR4.

One thing you can do is open AIDA64, scroll to and expand the 'Motherboard' header and click on 'SPD'. As you can see on the 'Module Size' row, our Corsair DDR4-4000 8GB kit is single-ranked.

The other method is that "ranks" can just be interpreted as "sides". While the two terms are not related from a technical standpoint, in reality they are strongly correlated. If you take a look at this memory support list (.PDF) for a random GIGABYTE Z270 motherboard, you will see that nearly every single rank (1Rx8) memory kit is also single-sided (SS). There are exceptions to this, but they are rare and mostly found in 'OEM' type memory, the green PCB stuff.

From a consumer point-of-view, there is no reason to buy dual rank modules, unless you need 16GB modules, in which case you currently have no other option. Some people will point out that dual rank modules are sometimes faster - they can achieve higher bandwidth numbers due to better interleaving, at the expense of higher latency - but the fact that they are harder to find and can bottleneck memory overclocking makes them unattractive in our opinion.

This leaves Ryzen buyers in the unfortunate situation of trying to navigate something of a minefield when putting together their new systems. To play it safe, we'd recommend you stick with 16GB dual channel kits that don't run higher than 2666MHz or dual channel 32GB kits that don't exceed 2400MHz. Higher speed memory can and will cause boot-up issues.

 

[SKYMTL]

Test Setups & Methodology

Test Setups & Methodology

For this review, we have prepared a number of different test setups, representing many of the popular platforms at the moment. As much as possible, the test setups feature identical components, memory timings, drivers, etc. Aside from manually selecting memory frequencies and timings, every option in the BIOS was at its default setting.

For all of the benchmarks, appropriate lengths are taken to ensure an equal comparison through methodical setup, installation, and testing. The following outlines our testing methodology:

A) Windows is installed using a full format.

B) Chipset drivers and accessory hardware drivers (audio, network, GPU) are installed.

C)To ensure consistent results, a few tweaks are applied to Windows 10 and the NVIDIA control panel:

• UAC – Disabled
• Indexing – Disabled
• Superfetch – Disabled
• System Protection/Restore – Disabled
• Problem & Error Reporting – Disabled
• Remote Desktop/Assistance - Disabled
• Windows Security Center Alerts – Disabled
• Windows Defender – Disabled
• Screensaver – Disabled
• Power Plan – High Performance
• V-Sync – Off

 

[SKYMTL]

System Benchmarks: AIDA64

AIDA64 Extreme Edition

AIDA64 uses a suite of benchmarks to determine general performance and has quickly become one of the de facto standards among end users for component comparisons. While it may include a great many tests, we used it for general CPU testing (CPU ZLib / CPU Hash) and floating point benchmarks (FPU VP8 / FPU SinJulia).


CPU PhotoWorxx Benchmark

This benchmark performs different common tasks used during digital photo processing. It performs a number of modification tasks on a very large RGB image:

This benchmark stresses the SIMD integer arithmetic execution units of the CPU and also the memory subsystem. CPU PhotoWorxx test uses the appropriate x87, MMX, MMX+, 3DNow!, 3DNow!+, SSE, SSE2, SSSE3, SSE4.1, SSE4A, AVX, AVX2, and XOP instruction set extension and it is NUMA, HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.

CPU ZLib Benchmark

This integer benchmark measures combined CPU and memory subsystem performance through the public ZLib compression library. CPU ZLib test uses only the basic x86 instructions but is nonetheless a good indicator of general system performance.

CPU AES Benchmark

This benchmark measures CPU performance using AES (Advanced Encryption Standard) data encryption. In cryptography AES is a symmetric-key encryption standard. AES is used in several compression tools today, like 7z, RAR, WinZip, and also in disk encryption solutions like BitLocker, FileVault (Mac OS X), TrueCrypt. CPU AES test uses the appropriate x86, MMX and SSE4.1 instructions, and it's hardware accelerated on Intel AES-NI instruction set extension capable processors. The test is HyperThreading, multi-processor (SMP) and multi-core (CMP) aware.

CPU Hash Benchmark

This benchmark measures CPU performance using the SHA1 hashing algorithm defined in the Federal Information Processing Standards Publication 180-3. The code behind this benchmark method is written in Assembly. More importantly, it uses MMX, MMX+/SSE, SSE2, SSSE3, AVX instruction sets, allowing for increased performance on supporting processors.

FPU VP8 / SinJulia Benchmarks

AIDA’s FPU VP8 benchmark measures video compression performance using the Google VP8 (WebM) video codec Version 0.9.5 and stresses the floating point unit. The test encodes 1280x720 resolution video frames in 1-pass mode at a bitrate of 8192 kbps with best quality settings. The content of the frames are then generated by the FPU Julia fractal module. The code behind this benchmark method utilizes MMX, SSE2 or SSSE3 instruction set extensions.

Meanwhile, SinJulia measures the extended precision (also known as 80-bit) floating-point performance through the computation of a single frame of a modified "Julia" fractal. The code behind this benchmark method is written in Assembly, and utilizes trigonometric and exponential x87 instructions.

Kicking off our Ryzen benchmarks with the PhotoWorxx test is perfect since it highlights one of Zen’s challenges against Intel’s competing architecture. While this test may be multi core aware, like many image processing tasks it is lightly threaded at best. This leads to architectures with higher IPC (instructions per clock) rates providing results that often bely their relative positioning. A good example of this is how well the Kaby Lake processors relative to those expensive Broadwell-E chips. This leads me to hypothesize that AMD still has some IPC improvements to do with the Zen cores since despite utilizing a nearly identical clock speed the 1800X loses to the i7-6900K in pretty spectacular fashion.

Moving on to a more parallel processing workload and Ryzen is able to truly shine. ZLib is an important test for any processor since it doesn’t take advantage of instruction set accelerators like AVX or SSE and rather relies on raw horsepower. Meanwhile, it’s evident that AMD has baked some very serious cryptography and hashing optimizations into Zen, likely as a result of its ultra quick caching hierarchy. As a result the Ryzen 7 1800X is able to blow Intel’s most expensive enthusiast-level chips straight out of the water in AIDA’s AES-256 and CPU Hash benchmarks. This architecture looks to be highly potent for bitcoin mining and encryption / decryption.

Finally the Floating Point Unit VP8 and SinJulia tests once again highlight just how far AMD has come between Bulldozer and Zen. In both cases performance is nothing short of astounding for a $500 processor.

 

[SKYMTL]

System Benchmarks: Cinebench / PCMark 8 / WPrime

CineBench R15 64-bit

The latest benchmark from MAXON, Cinebench R15 makes use of all your system's processing power to render a photorealistic 3D scene using various different algorithms to stress all available processor cores. The test scene contains approximately 2,000 objects containing more than 300,000 total polygons and uses sharp and blurred reflections, area lights and shadows, procedural shaders, antialiasing, and much more. This particular benchmarking can measure systems with up to 64 processor threads. The result is given in points (pts). The higher the number, the faster your processor.

PCMark 8

PCMark 8 is the latest iteration of Futuremark’s system benchmark franchise. It generates an overall score based upon system performance with all components being stressed in one way or another. The result is posted as a generalized score. In this case, we didn’t use the Accelerated benchmark but rather just used the standard Computational results which cut out OpenCL from the equation.

WPrime

wPrime is a leading multithreaded benchmark for x86 processors that tests your processor performance by calculating square roots with a recursive call of Newton's method for estimating functions, with f(x)=x2-k, where k is the number we're squaring, until Sgn(f(x)/f'(x)) does not equal that of the previous iteration, starting with an estimation of k/2. It then uses an iterative calling of the estimation method a set amount of times to increase the accuracy of the results. It then confirms that n(k)2=k to ensure the calculation was correct. It repeats this for all numbers from 1 to the requested maximum. This is a highly multi-threaded workload. Below are the scores for the 1024M benchmark.

As the tests move into a little more relatable territory, we have a bit of a yin and yang situation. Ryzen obviously excels in Cinebench’s and WPrime’s multi-threaded workloads but it tends to struggle against Intel’s latest Kaby Lake architecture in the predominantly lightly threaded environments PCMark utilizes. On the other hand, the 1800X competes very well against and even beats the nearly three year old Broadwell-E design.

 

[SKYMTL]

Productivity Benchmarks: 7-Zip / Adobe Premier Pro

7-Zip

At face value, 7-Zip is a simple compression/decompresion tool like popular applications like WinZip and WinRAR but it also has numerous additional functions that can allow encryption, decryption and other options. For this test, we use the standard built-in benchmark which focuses on raw multi-threaded throughput.

Adobe Premier Pro CC

Adobe Premier Pro CC is one of the most recognizable video editing programs on the market today as it is used by videography professionals and YouTubers alike. In this test we take elements of a 60-second 4K video file and render them out into a cohesive MP4 video via Adobe’s Media Encoder. Note that GPU acceleration is turned on.

Moving on to real world tasks highlights just how effective Ryzen can be from a price / performance standpoint, particularly in Adobe Premier Pro’s Media Encoder. Remember this is a $500 processor that’s keeping up blow for blow against competitors that cost twice or even three times as much. With that being said, even with a TITAN X chugging along with background acceleration, there’s still some bottlenecking going on behind the scenes so the differentiation between these processors is relatively minimal.

 

[SKYMTL]

Productivity Benchmarks: Blender / 3ds MAX - Corona

Blender

Blender is a free-to-use 3D content creation program that also features an extremely robust rendering back-end. It boasts extremely good multi core scaling and even incorporates a good amount of GPU acceleration for various higher level tasks. In this benchmark we take a custom 1440P 3D image and render it out using the built-in tool. The results you see below list how long it took each processor to complete the test.

3ds MAX Corona Renderer

Autodesk’s 3ds MAX is currently one of the most-used 3D modeling, animation and rendering programs on the market, providing a creative platform for architects to industrial designers alike. Unfortunately its rendering algorithms leave much to be desired and third party rendering add-ons are quite popular. One of the newest ones is called Corona.

In this test we take a custom 3D scene of a room with global illumination enabled and render it out in 720P using Corona’s built-in renderer.

Here we have two programs that have very similar utilizations but instead of using the standard canned benchmarks, I’m using customized files for both. While the Ryzen 7 1800X doesn’t win against the i7-6900K in either situation, its performance is so close Intel’s $1000 processor that it’s essentially tied. After talking to some developers who are more knowledgeable about CPU utilization in these two applications, I’m convinced that the quad channel memory controller on Broadwell-E has a small but net positive impact on these render times, particularly within Corona’s 3ds Max plugin.