AMD Radeon VII: High-End Chip (Part 3)

Original author: Nate Oh
  • Transfer
Part 1 >> Part 2 >> Part 3

Computing performance

Perhaps enough about the games. Consider the computing aspects of Radeon VII. Although it basically resembles the first generation Vega, there is a lot of noise around it about the improved processing power of Vega 20. It’s time for us to figure this out.

Starting with CompuBench 2.0, the latest iteration of GPU computing tests from Kishonti offers a wide range of different computing loads, and we decided to focus on level set segmentation (LSM), optical flow simulation and N-body simulation.

Next, we look at single-precision floating-point performance with FAHBench, the official benchmark of Folding @ Home. Folding @ Home is a popular research and distributed computing initiative supported by Stanford that spreads across millions of volunteer computers over the Internet and each computer is responsible for a tiny fragment of a protein folding simulation. FAHBench can test single and double precision floating point performance, with single precision being the most important metric for most consumer cards due to the low double precision performance.

Next comes the GPU computing package using Geekbench 4. This is a multifaceted test suite that runs seven different sub-tests of the GPU - from detecting faces to fast Fourier transforms, and then estimating their average geometric mean. As a result, Geekbench 4 does not test any single workload, but is the average of many different basic workloads.

Finally, we have SiSoftware Sandra, with general computational tests with varying degrees of accuracy.

Synthetic performance

Below are the results of our synthetic tests.

Professional visualization and rendering

Since AMD is actively promoting Radeon VII as a semi-professional card for creating content, we should pay attention to rendering performance, CAD and professional visualization. Nevertheless, suitable exact benchmarks for this area are not so easy to find, especially since the performance is highly dependent on the workflow and proprietary licensed ISV software. Given the AnandTech audience, which often includes engineers who use these applications in mission-critical production environments, our goal is to provide the most relevant metrics. However, as Yang previously discussed, the most accurate benchmarks for professional workstations are in the hands of independent software vendors. And these suppliers are at best unhappy,

Given these complexities, the next best thing to gauge the overall performance of the desktop GPU is the venerable SPECviewperf, recently upgraded to version 13. Divided into “viewsets,” which are workload groups for specific applications derived from real data sets, SPECviewperf has long been a package benchmarks for overall desktop / CAD GPU performance. For SPECviewperf 13, test suites are based on:

  • Autodesk 3ds Max 2016 (Nitrous DX11 driver)
  • Dassault Systèmes CATIA V6 R2012
  • PTC Creo 3 & Creo 4
  • Geosurvey software, with workloads based on rendering techniques utilized by the open-source OpendTect seismic visualization application
  • Autodesk Maya 2017
  • Radiological (i.e. CT, MRI scans) rendering, with workloads using the Tuvok rendering core of the ImageVis3D volume visualization application
  • Autodesk Showcase 2013
  • Siemens NX 8.0
  • Dassault Systèmes Solidworks 2013 SP1

Although we did not have time to fully test video editing / production software such as Adobe Premiere Pro CC, we will try to add this in the future.

Test Charts

The results obtained make it clear that some test suites work better on equipment from one manufacturer than on another. Sometimes Radeon VII breaks the trend, although in Siemens NX the reduced performance may be related to the immaturity of the drivers. In reverse scenarios such as Creo-02 or Maya-05, Radeon VII swaps with competitors, naturally ahead of the RX Vega 64, but inferior to the RTX and GTX cards. Although the results highlight the importance of software availability for new hardware, we see signs that the Vega 20 is a powerful workstation board. The disadvantage is that the new one does not change the general picture for applications that traditionally work well on NVIDIA hardware.

The following set of tests evaluates rendering performance. Given the nature of the “rendering wars,” as well as the features of CUDA, it needs to be made clear that the rendered rendering engines do not necessarily show the general GPU rendering capabilities. Since we are looking at Radeon VII, it is impractical to use some of the most popular renderers, such as Redshift and Octane, which are only for CUDA, on the other hand, using the not-so-popular Indigo Renderer gives one more data point.

Note that the official releases of Blender still do not include CUDA 10, so RTX 20 series cards are not officially supported.

V-RAY is the only test using CUDA for NVIDIA cards, while the rest use OpenCL. The results are generally similar to SPECviewperf, where Radeon VII continues to be a leader where AMD hardware has worked well before.

Radeon VII and Radeon RX Vega 64: clock speed

Thanks to the various changes that have taken place from RX Vega 64 based on Vega 10 to the new Radeon VII based on Vega 20, we wanted to take a look at the performance and computing capabilities, controlling their clock speed. This will allow us to look at any significant improvements or differences in pseudo-IPC. There are a couple of refinements here; obviously, since the RX Vega 64 has 64 CUs and the Radeon VII only 60 CUs, the comparison is no longer accurate. Another thing is that “IPC” is not an exact measurable indicator, but in a sense, how much graphic / computational work is performed per clock, and how this can affect performance. Isoclock GPU comparisons are generally not very useful when comparing generations and architectures. For example, Vega developers often design conveyor stages to provide higher clock speeds,

In our case, the incremental nature of the second generation Vega eliminates some of these problems, although, unfortunately, Wattman was not able to lower the memory frequency at the moment, so we could not get data for the case when both cards are configured for comparable memory bandwidth. Although the mechanics of improving the performance of the Vega GPU means that the clock frequency is not fixed, both cards were tuned to a frequency of 1500 MHz, and in both cases the frequency ranged from 1490 to 1500 MHz depending on the workload. All this means that the results obtained should be understood as approximate, and yet they are useful for determining a significant increase or decrease. It also means that interpreting the results is very difficult, but at a high level of abstraction,

Test Charts

Consumption, temperature and noise

As usual, it's time to consider the power consumption, temperature, and noise of the Radeon VII. Although we usually rely on voltage and clock speed, given the changes in the SMU, this approach is not possible for the first time.

At first glance, the noise level of the card looks surprisingly low. Ultimately, what we see here is the result of applying all possible acoustic enhancements to an open triple fan card. Switching to a new type makes the card noise comparable to the RX Vega 64 air duct.

In conclusion

Although there are still many areas to explore, Radeon VII is still the first 7nm gaming GPU, and this is a significant achievement. But besides this, its introduction to the consumer market gives buyers the opportunity to choose among manufacturers; and the more options there are for enthusiasts, the better. The Radeon VII may look like a dual-use product for professional and gaming applications, but it still remains the fastest gaming graphics card in the Radeon stack.

At the risk of being annoying, I still can’t stress how surprised both Ryan and I are that this card came out so early. Currently, 7nm technology is only taking its first steps, and until last month, AMD seemed to be pleased with the use of the Vega 20 GPU on Radeon Instinct server cards. And then there was an unexpected confluence of factors that allowed AMD to introduce a chip that, by their own admission, was originally created for servers, as a flagship in the consumer market. This may be a very good precedent for launching the Radeon VII, but it makes the situation quite interesting for technical enthusiasts.

Summing up, let's talk about performance indicators. From its main competitor, GeForce RTX 2080, Radeon VII is 5-6% behind in our test suite. Unfortunately, the only games in which he showed the best results were Far Cry 5 and Battlefield 1, so Radeon VII was not able to "exchange blows" with a competitor as much as AMD would like to see. At the same time, unlike the RTX 2080, AMD does not seek to shake the price-performance ratio, therefore Radeon VII does not interrupt the price of 2080 in any way. This is a reasonable choice for AMD, given the current market situation. But this also means that when the card does not work fast enough, there is no cost advantage to cover the performance lag.

In terms of performance compared to the original RX Vega 64, the Radeon VII looks even better than we expected. According to recent reports, the new flagship Radeon is 24% faster at 1440p, and 32% faster at 4K resolution than its predecessor. Thus, despite the interesting core configuration, in which Radeon VII comes with fewer CUs than the RX Vega 64, Radeon VII is significantly ahead of its predecessor.
In general, AMD launched the product, faced with a difficult struggle, both in terms of technology and product positioning. And the results for AMD are mixed. Although in modern games it is extremely difficult to capitalize on 16 GB of video memory, I am not ready to write it off as an insignificant factor. In the last two and a half years, the capabilities of VRAM video cards have not changed much, maybe the time has come. However, at the moment, AMD’s additional VRAM will do little good for gamers.

Content creation, on the other hand, looks a lot more interesting. Unlike games, there is no standard workload, so I can only speak in general phrases. Radeon VII is a fast card with 16 GB of video memory; This is a card that has no analogues in the market. And if semi-professional or professional visualization users who want to work with minimal investment have a workload that really requires more than 8-11 gigabytes of video memory, then Radeon VII is worth paying attention to. It will not be superfluous to recall the benefits of professional support, AMD professional drivers and other services that AMD will undoubtedly introduce to professional users of the Radeon Pro-Grade Vega 20 card.

As regards AMD’s technological victories, it’s important to note that the Radeon VII showed a significant improvement to Vega. AMD's downside is that Radeon VII is still Vega. I don’t want to talk a lot about ray tracing or other game moments, because I’m not sure that you can add anything significant to our GeForce reviews. But overall, the Vega 20 is a lot of small, neat additions to the architecture of Vega, even if they are not intended for consumers.

What is most worrying is that AMD’s strategy for configuring its cards has not changed much compared to the RX Vega 64: AMD is still striving for performance above everything else. This makes sense, given the position of AMD, but it also means that Radeon VII is not even trying to eliminate some of the shortcomings of its predecessor, which is especially noticeable against competitors. The Radeon VII has many advantages, but energy efficiency is not one of them.

Overall, Radeon VII does its best to offer itself as a semi-professional high-VRAM card for game content creators. And priced at $ 699, this is a good niche in the market. However, for pure gamers, it’s harder to sell this card instead of the more powerful NVIDIA GeForce RTX 2080.

So where did AMD end up? Fortunately for Radeon rebels, their situation is still improving, even if the overall competitive environment has not undergone significant changes. This is not a victory for AMD, but an opportunity to compete with NVIDIA on a new level. AMD is still competitive. They can compete in performance, and thanks to Vega 20 they have a lot of new computing features. The new product will advance AMD’s business today, and it will help AMD prepare for tomorrow and the next step - Navi. It's still a tough battle, but with Radeon VII and Vega 20 AMD took it one step further.

Thank you for staying with us. Do you like our articles? Want to see more interesting materials? Support us by placing an order or recommending to your friends,30% discount for Habr users on a unique analogue of entry-level servers that was invented by us for you: The whole truth about VPS (KVM) E5-2650 v4 (6 Cores) 10GB DDR4 240GB SSD 1Gbps from $ 20 or how to divide the server correctly? (options are available with RAID1 and RAID10, up to 24 cores and up to 40GB DDR4).

VPS (KVM) E5-2650 v4 (6 Cores) 10GB DDR4 240GB SSD 1Gbps until the summer for free when paying for a period of six months, you can order here .

Dell R730xd 2 times cheaper? Only we have 2 x Intel Dodeca-Core Xeon E5-2650v4 128GB DDR4 6x480GB SSD 1Gbps 100 TV from $ 249 in the Netherlands and the USA! Read about How to Build Infrastructure Bldg. class using Dell R730xd E5-2650 v4 servers costing 9,000 euros for a penny?

Also popular now: