Logo   Banner   TopRight
TopUnder
Transparent
AMD A10 and A8 Trinity APU: Virgo CPU Performance
Transparent
Date: Oct 02, 2012
Section:Processors
Author: Marco Chiappetta and Dave Altavilla
Transparent
Introduction and Related Info

Late last week, we posted a glimpse of what AMD had in store with their latest A-Series APUs for desktop systems. In that article, we were able to officially unveil the Virgo platform and discuss speeds and feeds, GPU performance, and power consumption characteristics. However, we weren’t able to disclose any CPU performance numbers, overclocking data, or pricing. We know some of you aren’t too fond of cliffhangers that leave out important data like that, but thankfully we can give you all the full scoop here today.

In this piece, we’ll fill in the blanks regarding AMD’s latest desktop APUs and cover the rest of the performance characteristics many of you have been eagerly anticipating. We won’t, however, be covering the architectural details again. If you missed our previous coverage of Trinity, Piledriver, and the Virgo platform, please take a moment and peruse these previous HotHardware articles:

In those two articles, we cover all pertinent architectural details and features, spotlight AMD’s latest A-Series mobile platform and cover desktop A-Series GPU performance.


AMD A10-5800K and A8-5600K Socket FM2 A-Series APUs, Side-by-Side

We've had a couple of AMD's latest A-Series APUs in-house for a while now; more specifically the flagship AMD A10-5800K and the A8-5600K. Both of these APUs are quad-core variants, with 100W TDPs, and unlocked CPU and GPU multipliers for easier overclocking. AMD, however is also launching a host of other A-Series APUs, with varying features and specifications. The complete breakdown of the AMD A-Series APUs launching today is as follows:

As you can see, the A10-5800K is the top of the line model. It sports an integrated Radeon HD 7660D GPU with 384 active Radeon Cores, that run at 800MHz. There is 4MB of total cache on-board, and the CPU portion of the chip runs at a base clock of 3.8GHz which can Turbo up to 4.2GHz. The A8-5600K, however, is outfitted with an integrated Radeon HD 7560D, which has only 256 active Radeon Cores and a slightly lower frequency of 760MHz. The A8-5600K's CPU cores run at a base clock of 3.6GHz and Turbo up to 3.9GHz, but it has the same 4MB of total cache of its higher-end counterpart.

The other APUs in the new A-Series vary in their GPU configurations, core counts, clocks, and cache configurations. Also note that the "K" in certain model numbers denotes an unlocked APU, which offers more flexibility for overclocking. We'll have more on the subject of overclocking just a little later. First we have some benchmarks to talk about...
  

Transparent
Test Setup and SiSoft SANDRA

Test System Configuration Notes: When configuring our test systems for this article, we first entered their respective system BIOSes or UEFIs and set each board to its "Optimized" or "High performance Defaults". We then saved the settings, re-entered the BIOS/UEFI and set the memory speed to each platform's maximum, officially supported speed--DDR3-1866 in the case of Virgo. The solid state drives were then formatted, and Windows 7 Ultimate x64 was installed. When the Windows installation was complete, we fully updated the OS, and installed the drivers necessary for our components. Auto-Updating and Windows Defender were then disabled and we installed all of our benchmarking software, performed a disk clean-up, cleared any prefetch and temp data, and ran the tests.

HotHardware's Test Systems
Intel and AMD - Head To Head

System 1:
AMD A10-5800K
(3.8GHz - Quad-Core)
AMD A8-5600K
(3.6GHz - Quad-Core)

Asus F2A85-M Pro
(AMD A85 Chipset)

2x4GB Corsair DDR3-1866

Radeon HD 7660D/7560D
On-Board Ethernet
On-board Audio

OCZ Vertex 3 MaxIOPS

Windows 7 x64
System 2:
Intel Core i5-3470
(3.2GHz - Quad-Core)
Intel Core i3-3220/3225
(3.3GHz - Dual-Core)

MSI Z77A-GD65
(Z77 Express Chipset)

2x4GB G.SKILL DDR3-1866

Intel HD 2500/4000
On-Board Ethernet
On-board Audio

OCZ Vertex 3 MaxIOPS

Windows 7 x64
System 3:
AMD A8-3870K
(3.0GHz - Quad-Core)

Asus F1A75-V Pro
(AMD A75 Chipset)

2x4GB Corsair DDR3-1866
(@ 1866MHz)

Radeon HD 6550D IGP
On-Board Ethernet
On-board Audio

OCZ Vertex 3 MaxIOPS

Windows 7 x64
System 4:
AMD FX 8150
(3.6GHz Eight-Core)

Asus CrossHair V Formula
(AMD 990FX Chipset)

2x4GB G.SKILL DDR3-1866
(@ 1866MHz)

GeForce GTX 280
On-Board Ethernet
On-board Audio

OCZ Vertex 3 MaxIOPS

Windows 7 x64

SiSoftware SANDRA 2012
Synthetic Benchmarks

We began our testing with SiSoftware's SANDRA 2012, the System ANalyzer, Diagnostic and Reporting Assistant. We ran four of the built-in subsystem tests that partially comprise the SANDRA 2012 suite with AMD's latest APUs (CPU Arithmetic, Multimedia, Memory Bandwidth, and Cache and Memory). All of the scores reported below were taken with the processors running at their default clock speeds with 8GB of DDR3-1866 RAM running in dual-channel mode on an Asus F2A85-M Pro motherboard.
 

AMD A10-5800K:
Processor Benchmark
 
 

AMD A10-5800K:
Multimedia Benchmark
 

AMD A10-5800K:
Memory Bandwidth Benchmark
 

AMD A10-5800K:
Cache Latency Benchmark
 

AMD A8-5600K:
Processor Benchmark
 

AMD A8-5600K:
Multimedia Benchmark
 

AMD A8-5600K:
Memory Bandwidth Benchmark
 

AMD A8-5600K:
Cache Latency Benchmark
 

Despite their new architecture and much higher base and Turbo frequencies, the AMD A10-5800K and A8-5600K offer only marginal improvements in performance over previous-gen products according to SANDRA. As the processor and multimedia benchmarks show, the A10-5800K and A8-5600K are only marginally faster than a Phenom II X4 965 and hang with older Core 2 Quads. Memory bandwidth is in the 11.1 - 11.4GB/s range, and like Bulldozer-based FX-Series processors, cache latency is unfortunately much higher than previous-gen products.
 

Transparent
Futuremark PCMark 7

Futuremark's PCMark 7 is the latest version of the PCMark whole-system benchmarking suite. It has updated application performance measurements targeted for a Windows 7 environment and uses newer metrics to gauge relative performance.
Futuremark PCMark 7
System Level Benchmark

Below is what Futuremark says is incorporated into the base PCMark suite and the Entertainment, Creativity, and Productivity suites--the four modules we have benchmark scores for you here.

The PCMark test is a collection of workloads that measure system performance during typical desktop usage. This is the most important test since it returns the official PCMark score for the system
Storage

  • Windows Defender
  • Importing pictures
  • Gaming

Video Playback and transcoding
Graphics

  • DirectX 9

Image manipulation
Web browsing and decrypting

The Entertainment test is a collection of workloads that measure system performance in entertainment scenarios using mostly application workloads. Individual tests include recording, viewing, streaming and transcoding TV shows and movies, importing, organizing and browsing new music and several gaming related workloads. If the target system is not capable of running DirectX 10 workloads then those tests are skipped. At the end of the benchmark run the system is given an Entertainment test score.

The Creativity test contains a collection of workloads to measure the system performance in typical creativity scenarios. Individual tests include viewing, editing, transcoding and storing photos and videos. At the end of the benchmark run the system is given a Creativity test score.

The Productivity test is a collection of workloads that measure system performance in typical productivity scenarios. Individual workloads include loading web pages and using home office applications. At the end of the benchmark run the system is given a Productivity test score.

PCMark7 shows the AMD A10-5800K and A8-5600K outpacing the Llano-based A8-3870K and coming close to the Bulldozer-based FX-8150, but AMD's latest APUs can't quite hang with any of the Intel processors we tested. The Core i3 and i5 processors showed huge gains in a couple of tests, especially the creativity test where their Quick Sync encoding engines show their might, but the AMD APUs actually pulled ahead in the productivity benchmark.

Transparent
LAME MT and SunSpider

In our custom LAME MT MP3 encoding test, we convert a large WAV file to the MP3 format, which is a popular scenario that many end users work with on a day-to-day basis to provide portability and storage of their digital audio content. LAME is an open-source MP3 audio encoder that is used widely in a multitude of third party applications.

LAME MT
Audio Encoding

In this test, we created our own 223MB WAV file (a hallucinogenically-induced Grateful Dead jam) and converted it to the MP3 format using the multi-thread capable LAME MT application, in both single and multi-thread modes. Processing times are recorded below, listed in seconds. Shorter times equate to better performance.

Audio encoding with LAME MT is definitely not one of Piledriver's strong suits. The A10-5800K and A8-5600K outpace the fastest Llano chip here, but Intel's parts simply dominate.

SunSpider JavaScript Benchmark
JavsScript Performance Testing

Next up, we have some numbers from the SunSpider JavaScript benchmark. According to the SunSpider website:

This benchmark tests the core JavaScript language only, not the DOM or other browser APIs. It is designed to compare different versions of the same browser, and different browsers to each other. Unlike many widely available JavaScript benchmarks, this test is:

Real World - This test mostly avoids microbenchmarks, and tries to focus on the kinds of actual problems developers solve with JavaScript today, and the problems they may want to tackle in the future as the language gets faster. This includes tests to generate a tagcloud from JSON input, a 3D raytracer, cryptography tests, code decompression, and many more examples. There are a few microbenchmarkish things, but they mostly represent real performance problems that developers have encountered.

Balanced - This test is balanced between different areas of the language and different types of code. It's not all math, all string processing, or all timing simple loops. In addition to having tests in many categories, the individual tests were balanced to take similar amounts of time on currently shipping versions of popular browsers.

Statistically Sound - One of the challenges of benchmarking is knowing how much noise you have in your measurements. This benchmark runs each test multiple times and determines an error range (technically, a 95% confidence interval). In addition, in comparison mode it tells you if you have enough data to determine if the difference is statistically significant.

All of the systems were tested using the latest version of Internet Explorer 9, with default browser settings, on a clean install of Windows 7 Ultimate x64.

The A10-5800K and A8-5600K were once again somewhat faster than the Llano-based A8-3870K in the SunSpider benchmarks, but every other processor we tested had a huge advantage. Piledriver's lower IPC and higher cache latency versus Intel's offerings certainly hold it back here.

Transparent
Cinebench R11.5 and POV-Ray

Cinebench R11.5 is a 3D rendering performance test based on Cinema 4D from Maxon. Cinema 4D is a 3D rendering and animation suite used by animation houses and producers like Sony Animation and many others. It's very demanding of processor resources and is an excellent gauge of pure computational throughput.

Cinebench R11.5
3D Rendering

This is a multi-threaded, multi-processor aware benchmark that renders a photorealistic 3D scene (from the viral "No Keyframes" animation by AixSponza). This scene makes use of various algorithms to stress all available processor cores. The rate at which each test system was able to render the entire scene is represented in the graph below.

The new A10-5800K and A8-5600K didn't do so well in the Cinebench benchmark. AMD's latest APUs are able to compete well with Llano in terms of single-thread performance, but their shared FPU resources result in lower multi-thread performance versus every other processor we tested in this benchmark.

POV-Ray Performance
Ray Tracing

POV-Ray, or the Persistence of Vision Ray-Tracer, is an open source tool for creating realistically lit 3D graphics artwork. We tested with POV-Ray's standard 'one-CPU' and 'all-CPU' benchmarking tools on all of our test machines, and recorded the scores reported for each. Results are measured in pixels-per-second throughput; higher scores equate to better performance.

POV-Ray tells a similar story to Cinebench. Here, the new A10-5800K and A8-5600K offer somewhat better single-thread performance than Llano, but trail once again in the multi-threaded test. AMD's latest APUs, however, are able to pull ahead of the Core i3's here.

Transparent
Media Encoding and Image Editing
Cyberlink's MediaEspresso is a video conversion tool that imports various video file types and converts them to other formats for publication, portability and/or streaming. In this test, we take a 277MB high definition 1080p AVCHD video clip and convert it to a an H.264 encoded MP4 compatible file designed for use with an iPhone / iPad (or other portable media playback device).

Cyberlink MediaEspresso
Video Transcoding and Compression

Here we're going to look directly at AMD's HD Media Accelerator in the new A10 and A8 series APUs and compare it to Intel's Ivy Bridge Quick Sync engine and competitive CPU-based solutions as well.

AMD's updated media accelerator offers a huge speed improvement over the CPU alone, but it can't come close to matching the performance of Intel's Quick Sync engine.

Musemage
OpenCL Accelerated Image Editing

Musemage is a fully functional GPU-powered photo editing software suite. Thanks to the GPU accelerated filters built into the app, Musemage can leverage the additional performance afforded by a more powerful GPU and offer real-time (or near real-time) visual feedback. The application also offers a built-in benchmark, which output performance as a final "score". The score put up by each test system is represented in the graph below.

When AMD's A-Series APUs are able to leverage the additional performance offered by their relatively powerful integrated GPU's they put up some pretty good scores. Because the Intel processors have to handle the workload on their own, without assistance from a GPU, they lag behind AMD here.

Transparent
Overclocking Desktop Trinity

We took a bit of time to overclock AMD's higher-end A10-5800K APU as it'll likely be one of the more popular chips in AMD's new line-up that performance enthusiasts will want to tinker with.  Overclocking the A10 is somewhat delicate work, perhaps more so than previous AMD architectures. 

Dialing up frequency on the IGP will undoubtedly contribute to additional thermal stress on your cooling solution, affecting the CPU, and vice-versa for that matter. You also do not need a lot of extra voltage in order to overclock AMD's new desktop APU.  Crank VID too high and you'll generate excess heat that will cause the CPU to throttle without high-end cooling.

Overclocking AMD's Latest A-Series APU
Pushing the CPU and GPU to the limit
Our tool of choice for stress testing was Maxon's Cinebench R11.5.  Cinebench allows us to test full, 100% CPU and GPU loads depending on which module of the test is run.  Perform multiple reps with Cinebench and you can pretty much squeeze out an unknown marginality in an overclocking effort.

 

We settled in on a 1.418V VID core voltage on the CPU with a stock 100MHz reference clock and a multiplier of 44X, giving us a top CPU speed of 4.4GHz under load.  This was achieved on standard air cooling and our DDR3 system memory set to 1866MHz. Our GPU speed was able to hit 1GHz with full stability as well, over a stock speed of 800MHz. 

This may seem like a modest overclock but we accomplished this with a run of the mill aftermarket air cooler, the Thermaltake TR2 R1.  Results may vary chip-to-chip and system-to-system but with more robust air cooling, liquid cooling or some other higher-end solution, there is definitely even more headroom available beyond what we're able to show you here.

 

 

The benefits of overclocking are apparent as always, with roughly a 10% kicker in CPU performance realized and a 15 - 17% increase in GPU performance for the overclocked Trinity A10 APU. 

On a side note, overclocking the GPU on the A10 is easy money. With a simple push of the clock slider 1GHz is attainable with virtually no fuss.  Overclocking the CPU can take a bit more finesse with reference clock speeds, multipliers and voltages to balance.

Transparent
Performance Summary and Conclusion

Performance Summary: In last week’s article detailing GPU performance, AMD’s latest Trinity-based APUs put up scores that easily outpaced Intel’s integrated graphics offerings. Today’s look at CPU performance, however, tells an entirely different story. The A10-5800K and A8-5600K generally offer better performance than the previous-generation Llano-based APUs they’ll be supplanting in AMD’s desktop APU line-up. Versus Intel’s similarly priced desktop offerings, however, the picture isn’t as rosy. The dual-core Core i3-3220 and i3-3225 keep pace with and in many cases significantly outpace AMD’s latest APUs.

The competitive landscape doesn’t change all that much with the release of AMD’s latest A-Series APUs. The lead in processor performance Intel has maintained over the last couple of generations remains firmly intact. And the superior integrated graphics performance AMD has offered since the introduction of Llano continues. The deltas have simply shifted a bit. The Trinity-based A10-5800K and A8-5600K are a little more competitive with Intel’s offerings at their respective price points in terms of CPU performance, but AMD has extended their lead a bit in terms of integrated graphics performance.

Speaking of price point, we’re sure many of you are wondering how much these things will cost. Fortunately, AMD is being aggressive on that front. The flagship A10-5800K will be priced around $122 and the A8-5600K will be about $101. The lower-end A6-5400K and A4-5300 will be only $67 and $53, respectively. Considering the kind of graphics performance offered by the A10-5800K and A8-5600K we tested, their somewhat improved CPU performance, and decent overclockability, AMD’s latest APUs represent a good value. If you’re looking to build an entry-level system with good graphics performance, AMD’s Trinity-based A-Series APUs and the Virgo platform are an affordable, no-fuss, solution.

  • Decent Performance
  • Good Overclockability
  • Improved Media Encoding Engine
  • Intel's CPUs still dominant
  • Shared compute resources and high-latency caches hamper performance



Content Property of HotHardware.com