|Introduction & Specifications|
AMD has been attacking the mainstream PC segment lately with a slew of modestly priced processors and chipsets. The attack started with the new Athlon II X2 value processors back in June, which were a much needed replacement for the out-dated Athlon X2 series. Then a few weeks later, AMD released their new mainstream chipset, the 785G, which we found to be an excellent complement to the new Athlon II's. Today AMD is adding the final piece to their new mainstream desktop platform with the release of a line-up of value-priced quad-cores; the Athlon II X4 series.
At first the announcement of Athlon II X4s doesn't seem too interesting. Especially since they have been rumored for some time and it wasn't much of a stretch to imagine that AMD would get around to releasing quad-core Athlon II processors eventually. Not to mention the Athlon IIs are in many ways just Phenom IIs with the L3 cache removed, so there isn't a ton of brand new tech under the hood to ogle at. However, things get a lot more interesting when you hear the list price.
Consider for a moment the prospect of full, native quad-core capabilities at a palatable mainstream price normally associated with dual-cores. Pleasant thought, isn't it? Well now AMD is making it a reality. The new Athlon II X4s will start at just $99, making them the cheapest quad-core processors on the market. A quad-core with a two digit price tag, interested yet?
AMD is making two models available immediately for the Athlon II X4 launch, the 2.6GHz 620 and the 2.8GHz 630. The higher clocked Athlon II X4 630 will have an initial MSRP of $122 while the Athlon II X4 620 will come in just under the magic $100 price point, at $99. This puts them in direct price competition with Intel's high-end dual-cores, low-end quad-cores, and even AMD's own Phenom II X3 triple-cores.
While the new Athlon II X4 620 undercuts every currently available quad and triple core processor on the market in cost, the 630 will be butting heads with the Intel Core 2 Quad Q8200 and the Phenom II X3 720. This makes for some interesting benchmark match-ups as you'll see in the coming pages. If the new Athlon II X4 processors can stand up against Intel's cheapest quad, the Q8200, AMD could have a winner on its hands.
AMD Athlon II X4 front & back
The Athlon II X4s are based on the new Propus processor die which differs in a several ways from the Deneb dies used in existing Phenom IIs. However, the new Propus die is still very similar to Deneb in most respects. We've covered the details of Phenom II processors and supporting chipsets a number of times in the past, so we won't do the same again here. We would, however, recommend taking a look at a few of our previous articles if you'd like a refresher on all of the pertinent details regarding the Phenom II. Here is a list of recommended reading:
The Athlon II processor series is AMD's mainstream processor offering and they are well complimented by AMD's own mainstream chipsets, the latest of which is the recently released 785G. We will be pairing up the new Athlon II X4s with the 785G chipset in our benchmarks, so be sure to check out our AMD 785G chipset review if you'd like a refresher.
|A Closer Look At The Athlon II X4|
The Athlon II X4 series is based on a new spin on the familiar Deneb core used in existing Phenom II processors, called Propus. The new Propus core is essentially a trimmed down Deneb with the L3 cache removed.
A quick peek with the latest version of CPU-Z shows us that unlike the dual-core Athlon II X2s which benefit from increased L2 cache to compensate for the lack of L3 cache, quad-core Athlon II X4s have to make do with the same amount of L1 and L2 cache as Deneb. Just like Deneb-based Phenom II processors, the Athlon II X4 630 and 620 are equipped with 64KB of L1 instruction and 64KB of L1 data cache per core, for a total of 512KB of L1. Each core also has 512KB of L2 data cache, for a grand total of 2MB L2 cache. Unlike Deneb which has 6MB of L3 cache, we see that Propus has none at all.
While the removal of Deneb's 6MB of L3 cache results in drastic die size savings, it comes at the cost of performance. As we saw in our Athlon II X2 review, cutting out the L3 cache isn't debilitating, but it does hurt performance in certain applications.
Other than the removal of L3 cache, the Propus core is otherwise identical to Deneb. As you can see in the images of the Propus and Deneb dies above, the Propus appears to literally be a Deneb with the L3 cache chopped off.
In fact, some of the early Athlon II X4 engineering samples that are currently floating around are actually 258mm2 Denebs cores with the L3 cache disabled and not true 169mm2 Propus cores. Some users who have gotten their hands on these chips have even claimed to have been able to unlock the L3 cache. However, if you have hopes of doing the same, you'll have to hurry as all future Athlon II X4 stock will be made from Propus cores with the L3 cache physically removed.
|Our Test Systems & SANDRA|
To assess the performance of the new Athlon II X4 processors, we put the two launch models through a gauntlet of benchmarks. Starting with some preliminary testing with SiSoftware's SANDRA benchmark suite, we then move to PC Mark Vantage followed by 3D rendering tests with Cinebench and Kribibench, and finally mp3 encoding with LAME MT. We also put the Athlon II X4's through Crysis' CPU benchmark.
For these tests the Athlon II X4 series was represented by the 2.6GHz 620 and the 2.8GHz 630. They were both equipped with an AMD 785G chipset based motherboard and DDR3 memory. For reference, we pitted them against Intel's G41 chipset equipped with the Intel Core 2 Quad Q8200 processor. While the Q8200 costs more than both the Athlon II X4 620 and 630, it's Intel's cheapest quad-core processor and should provide for some interesting comparisons. We also threw an AMD Phenom II X3 720 into the mix, considering its competing in the same price range and it will represent AMD's Phenom II's.
We began our testing with SiSoftware's SANDRA XII, the System ANalyzer, Diagnostic and Reporting Assistant. We ran two of the built-in subsystem tests that partially comprise the SANDRA 2009 suite (CPU Arithmetic & CPU Multimedia). All of the scores reported below were taken with the hardware running at default speeds and settings.
Overall, all of the various SANDRA CPU benchmarks we ran reported scores in line with expectations. The performance results in both the CPU arithmetic and CPU multi-media tests were similar to the scores an Intel Core 2 Q8200 can achieve. In the rest of the benchmarks, we expand our scope and put the two Athlon II X4 processors up against some real competition.
Next we ran a number of processor and platforms, including both of the new Athlon II X4 models, through Futuremark’s latest system performance metric built especially for Windows Vista, PCMark Vantage. PCMark Vantage runs through a host of different usage scenarios to simulate different types of workloads including High Definition TV and movie playback and manipulation, gaming, image editing and manipulation, music compression, communications, and productivity. Most of the tests are multi-threaded as well, so the tests can exploit the additional resources offered by a quad-core CPU.
The PCMark Vantage benchmark results are pleasantly surprising. Not only do both the new Athlon II X4's perform exceptionally well, but they keep right up with the more expensive Phenom II X3 720. The new quad-core Athlons also pull ahead of Intel's cheapest quad. Most surprising is the performance of the $99 2.6GHz 620, which competes favorably with the much more expensive Core 2 Quad Q8200.
|LAME MT & Kribibench|
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 mid to high bit-rate and VBR (variable bit rate) MP3 audio encoder that is used widely around the world in a multitude of third party applications.
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 single and multi-thread modes. Processing times are recorded below, listed in seconds. Once again, shorter times equate to better performance.
In the LAME MT test, all three of the AMD processoes tested fall behind the Core 2 Quad. However, the Athlon II X4 remain competitive and they even manage to post better scores than the Phenom II X3.
For this next batch of tests, we ran Kribibench v1.1, a 3D rendering benchmark produced by the folks at Adept Development. Kribibench is an SSE aware software renderer in which a 3D model is rendered and animated by the host CPU and the average frame rate is reported. We used two of the included models with this benchmark: a "Sponge Explode" model consisting of over 19.2 million polygons and the test suite's "Ultra" model that is comprised of over 16 billion polys.
Kribibench provides some interesting results. The two quad Athlon II X4 processors mop the floor with the more expensive Phenom II X3, thanks to the extra core. Unlike LAME MT which only uses up to two cores, Kribibench can utilize all the cores available, hence the advantage the Athlon quads have.
Cinebench R10 is an OpenGL 3D rendering performance test based on Cinema 4D from Maxon. Cinema 4D is a 3D rendering and animation tool suite used by 3D animation houses and producers like Sony Animation and many others. It's very demanding of system processor resources and is an excellent gauge of pure computational throughput.
This is a multi-threaded, multi-processor aware benchmark that renders a single 3D scene and tracks the length of the entire process. The rate at which each test system was able to render the entire scene is represented in the graph below.
Like Kribibench, Cinebench is another rendering test that can take full advantage of all the cores available in the new Athlon quads. Once again we see the cheaper Athlons beating out the Phenom II. They also remain very competitive with the Intel Core 2 Quad Q8200. The higher clocked 630 actually nudges ahead of the Q8200 for first place.
|Low-Res Gaming: Crysis|
For our next set of tests, we moved on to some in-game benchmarking with Crysis. When testing processors with Crysis, we dropped the resolution to 800x600, and reduce all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible. However, the in-game effects, which control the level of detail for the games' physics engines and particle systems, are left at their maximum values, since these actually do place some load on the CPU rather than GPU. All platforms in this test were equipped with a NVIDIA GTX 280 to eliminate the graphics card as a possible bottleneck, in order to isolate the CPU's performance as much as possible.
For the Crysis gaming test, we pit the new Athlon II X4s against a larger field of competitors including a dual-core Intel CPU and a Phenom II quad-core. We can see here that the Athlon II X4's lack of L3 cache really hurts it in games where cache memory is frequently used, such as Crysis. Unlike in all of the previous tests where both of the Athlon II X4 processors were very competitive with the more expensive Core 2 Quad Q8200, in Crysis the Q8200's vastly superior cache configuration gives it a huge edge. While the rest of the field posted scores in the 110+ area, the two Athlon II X4 processors are stuck around 80. It's clear that the new Athlons aren't the best choices around for gaming chips.
Considering the very low price of the new Athlon II X4 processors, we wondered how well they'd do with a bit of overclocking. We were especially interested in the 2.6GHz 620, which now claims the title of the cheapest quad-core available. Rather than loading the little 620 into a full blown overclocking rig with all the bells and whistles, we decided to try something more typical of an actual build someone might use with the new Athlon II quads. For our overclocking romp, we used a cheap Arctic Cooling Freezer heatsink and fan combo.
It's worth noting that all Athlon IIs are multiplier locked and Black Edition models with unlocked multipliers are currently not available. Therefore, for our quick overclocking experiment, we relied primarily on increasing the base clock speed, which by default runs at 200MHz.
Starting off with the 620, we effortlessly increased the base clock up to 265MHz, bringing the overall clock speed to 3.45Ghz. We found we could push the system beyond 265MHz, but stability becomes unreliable and pumping in more voltage didn't seem to have any effect. We could boot into windows up to 280MHz, but benchmarks would lock the system. We then tried the same method with the 630 which was also able to reach 265MHz, but due to its higher clock multiplier, it was able to achieve 3.71Ghz overall.
Athlon II X4 620 overclocked to 3.45GHz
Athlon II X4 630 overclocked to 3.71GHz
In the end, our quick overclocking experiment shows that the Athlon II X4 processors have quite a bit of overclocking room. Even with our crude methods and cheap cooling, we were able to bump up the clock speed by nearly a full gigahertz. That's on par with what we've seen from Phenom IIs and is just another reminder that these Propus cores are basically just cut-down Denebs.Overall we are pretty impressed with the overclocking potential in the Athlon II X4 quads. The 620 was able to reach 3.45GHz completely stable, which is extremely impressive considering its $99 price tag. The 630 was able to reach 3.71GHz which is certainly very respectable for its $122 price tag. For those of you who like to tweak and overclock their systems, the new Athlon II X4s are looking like an excellent choice for some cheap overclocking fun.
|Our Summary & Conclusion|
Performance Summary: The new quad-core AMD Athlon II X4 processors performed very well, especially given their relatively low price points. Despite its lack of L3 cache, the new Propus core used in the Athlon II X4s perform extremely well compared to their full-fledged Deneb-based siblings. Comparatively, the new Athlon II X4 processors are also very competitive with Intel's low-end quad-core offerings. However, gaming performance suffered due to the lack of L3 cache. Overall though, the Athlon II X4 processors are surprisingly good performers.
It's often hard to get excited about low-end and mainstream hardware. While they can offer good value for the dollar, they don't often deliver much to get excited about. However, AMD's new quad-core Athlon II X4 processors are something we can definitely get excited about.
While quad-core processors have been around in the consumer space for several years now and they've gotten progressively cheaper, they have yet to hit the sub-$100 price range. Until now, that is. However, as we saw in our benchmarks, AMD didn't just deliver on price, they also managed to produce quite a competitive product that was able to keep up with more expensive processors like Intel's Q8200 and AMD's own Phenom IIs.
While the new quad-core Athlon IIs are essentially just cut-down Phenom IIs, they don't suffer too much from having all of their L3 cache removed. For general usage scenarios, the Athlon II X4 processors compared well to their fully endowed Phenom II stablemates. The only area where the lack of L3 cache was especially noticeable was with games like Crysis which make heavy use of cache memory. However, for casual or non-gamers, the Athlon II X4s will definitely be a good pick. They might even be a hit with the overclocking crowd, thanks to their low price and decent overclocking performance. With simple air cooling and no voltage adjustments, we were able to boost the $99 2.6GHz 620 up to a healthy 3.45GHz.
The new quad-core Athlon IIs complete AMD's mainstream desktop platform. Partnered with the excellent AMD 785G chipset, AMD's new Athlon II X4 quad-cores provide a feature-packed platform at a very low price. While there may be cheaper dual-core solutions out there, a quad-core for a few more bucks will be hard to pass up. We think the new Athlon II X4's are an excellent value and they will be a great pick for typical productivity and internet builds, as well as light multimedia duty.