AMD Athlon II X4 620 and 630 - BeHardware
>> Processors

Written by Marc Prieur

Published on September 18, 2009


Page 1

The Propus core

AMD has launched two new quad core processors, the Athlon II X4 620 and 630. Announced at $99 and $122, they are clocked at 2.6 and 2.8 GHz respectively. How has AMD managed to bring them out so cheaply and how do the processors perform?

A third die at 45nm

Following on from the Phenom II X4 and the Athlon II X2, the Athlon II X4 enjoys a third die engraved at 45 nm. The new chip then is not an assembly of two Athlon II X2s or a Phenom II X4 with the L3 cache disactivated (even if in practice, that may also be the case): the Athlon II X4 integrates 4 cores on its silicon die, each with 128 KB of L1 as on all recent AMD processors. The L2 cache is 512 KB per core, as with the Phenom IIs, which is half the L2 cache on the Athlon II X2. Otherwise you’ll find the HyperTransport 3 bus and DDR2/DDR3 memory controller used by all AM3 CPUs.

Of course, it is much smaller as it is made up of 300 millions transistors that fit on 169 mm² of silicon, as against 258 mm² for a Phenom II X4. This means that AMD can engrave more processors per wafer and therefore drastically reduce production costs.

Compared to the Intel offer, we can see that the Athlon II X4 is around the same size as a Core 2 Quad from the Q9x00 (Q9300/9400) range. These Core 2 Quads are based on the assembly of two Penryn dies with 3 MB of cache that are used on their own on the Core 2 E7xxx but also on the Pentium E6xxx and E5xxx (with 1 and 2 MB of L2 cache deactivated respectively). The same die is also used for the Q8x00 (Q8200/8300/8400), with just 2x2 MB of L2.

Page 2
Power consumption: overclocking

The processors
For this test, we managed to get our hands on both the Athlon II X4 620 and X4 630.

It is interesting to note that the voltages of the two processors differ by quite a bit: the Athlon II X4 620 works at 1.125V in idle and 1.4V in load, as against 1.025V and 1.325V for the Athlon II X4 630. This can vary from one processor to another depending on the quality of the die, AMD simply saying that the nominal voltage is between 0.925 and 1.425V.
Energy consumption
Energy consumption is measured in two places: processor consumption alone at the ATX12V socket and with a clip-on ammeter and total consumption with a power meter at the wall socket, always loading the machine with Prime 95. Watch out, then, for your other components such as the graphics card or hard drive, which are in idle when these readings are taken.

In the case of the reading at the ATX12V we only compare readings within the same platform so as to be able to compare apple to apple; on AM3 for example, the memory controller is supplied by this power source but on the LGA1156/1366 platform, the uncore part of the processor is supplied by the standard ATX connector.

The Athlon II X4 620 remains quite economical both in idle and in charge, a long way down on the X4 965. Strangely, in spite of its lower voltages, the 630 is lot more demanding both in idle and load. What this probably means is that although the Athlon II X4 620 is a “Propus”, the 630 is actually a “Deneb”, namily a chip with 6 MB of L3 but which are disactivated.

The comparison with socket 775 is not necessarily in favour of AMD as the configuration based on the Q9400 proves more economical both in idle and in charge. The scores are nevertheless reasonable.
Naturally, we tried to overclock the processors. The Athlon II X4 620 managed 3.2 GHz without increasing its voltage of 1.4V, which leads you to think that the voltage is somewhat overestimated. At 1.5V, we stabilized it at 3.4 GHz in 4 instances of Prime 95.

Version 630 was also stable up to 3.2 GHz without us having to change its voltage of 1.325V. We stabilized it at 3.4 Ghz at 1.4V and 3.6 GHz at 1.5V. Although overclocking is by its nature a bit random, the results for the Athlon II X4 620 were a little disappointing. Things are a bit better for the 630 but nothing exceptional.

Page 3
The test, influence of the L3

The test
For this test, we used the same protocol as the one we inaugurated for the test of the Core i5s. As you’ll remember if you’ve had a look at that test, we took advantage of the availability of the final version of Windows 7 to revamp the protocol. The OS first then: we’re now using a 64-bit version of Windows 7, which means that all software available in 64-bit mode is tested in this mode.

We have taken the opportunity to update the software, which means 3ds max is now tested in Version 2010, Min GW and WinRAR (3.8 up to 3.9) have been updated, as have After Effects (CS3 up to CS4) and Nuendo (4.2 up to 4.3). The VirtualDub/DiVX combos and AutoMKV/x264 have been replaced by Avidemux/x264 and MainConcept Reference/H.264, while the test files of virtually all the tests have changed or been modified (higher rendering resolution for example).

In terms of games, we have decided to keep Crysis 1.2 and its ultra-heavy CPU test but to retire World In Conflict and replace it with more recent and demanding games: Arma 2, Grand Theft Auto IV and Anno 1404 join the protocol. So as to show up processor differences to a maximum, we set all graphics options to a max so as to load right up, at the same time as limiting resolution to 800*600 so as to eliminate any smoothing due to the power of the mono-GPU solution used on the test configuration.

The configurations are as follows:

- Intel DP55KG (LGA1156)
- Intel DX58SO (LGA1366)
- ASUSTeK M4A79-T (AM3)
- 2x2 GB DDR3-1333 7-7-7
- GeForce GTX 280 + GeForce 190.62
- Raptor 74 GB + Raptor 150 GB
- Creative Audigy
- Windows 7 64 bit
Impact of the L3 cache
What impact does the 6 MB of L3 cache have on performance? To answer this question, we tested the Athlon II X4 630 and the Phenom II X4 clocked at 2.8 GHz. Here are our results:

There is a significant impact on performance, 8.3% on average. Of course this varies from one application to another, with less than a 2% difference in Avidemux, MainConcept Reference and Nuendo but over 16% in Crysis and GTA IV.

It remains to be seen how these new Athlon II X4s do in comparison with the rest of the range in spite of this handicap!

Page 4
3ds Max 2010 & C4D R11

3D Studio Max 20010

We begin with the famous image rendering software, now in its x64 and 2010 version. The test scene used is from SPECapc for 3ds max 9 (space_flyby_mentalray) which employs the MentalRay rendering machine.

The Phenom II X4 965 is on a par with the Intel Q8200 but can do nothing to combat the Core i7s that are far out in front.
Cinema 4D R11

The rendering software Maxon is well known in the overclocker community through Cinebench, which allows you to compare processor performance easily. Cinebench however uses version R10 of the Cinema 4D rendering machine, while version R11 doubles performance. We use this latest version in 64 bit mode with the scene from Cinebench R10 rendered at a higher resolution so as to prolong rendering time.

The same as in Cinema 4D, where our two Athlon II X4s outperform the Q8200.

Page 5
MinGW & WinRAR 3.9


This is an application test with the compilation of MAME source code using GCC under the MinGW development environment. We are now using version 5.1.4 of MinGW and compile the source code of Mame 0.133.

Compilation is also in favour of the two Athlon II X4s, which are faster than the Q8200.
WinRAR 3.9

We are now using vesion 3.9 64 bit WinRAR, which introduces new multithreading optimisations for the compression of a group of files.

Here things are different as, even though WinRAR has been optimised for multithreading, going from 2 to 4 cores only has a very limited impact on performance, in the order of 5 to 10%. The small cache handicaps the Athlon II X4s here and they come in at the back of the field, behind even the E6300!

Page 6
H.264: Avidemux and MainConcept

Avidemux + x264

We have abandonned VirtualDub and DiVX encoding and replaced it with H.264 encoding for video processing. To start with, we use Avidemux version 2.5.1 to compress a 1920x1080 HD video file via the x264 codec at intermediary quality.

Here, our two quad cores are more at ease, once again outdoing the Q8200.
MainConcept Reference + H.264/AVC Pro

For this second H.264 encoding we use MainConcept Reference and its H.264/AVC Pro codec on “High”, still with the same video.

The Athlon IIs do well with H.264 encoding in MainConcept Reference, with the 620 on a par with the Q9550. The 630 is on an equal footing with the Q9650.

Page 7
After Effects CS3, Nuendo 4.3

After Effects CS4

In our new protocol we use version CS4 of Adobe After Effects. We are using a new composition using various effects so as to render 3D animation. Multiprocecessing is activated so as to make the most of the available number of cores.

After Effects CS4 is less to the liking of AMD processors and the Athlon IIs finish behind the Q8200.
Nuendo 4

We are now using version 4 of Nuendo, with patch 4.3, all in 64-bit. A new music project using various native plugins as well as 2 HalionOne virtual instruments was exported as a wav file (thanks to Draculax). As Nuendo doesn’t react well to HT (negative impact of about 6%), we forced the affinity on the physical processors by hand.

The same trend observed in After Effects CS4 is also visible in Nuendo 4.

Page 8
Crysis & Arma 2

Crysis 1.2

With patch 1.2, Crysis has a very heavy CPU bench (to be found in the Bin32/Bin64 directory). The test was carried out at very high settings but at 800x600 so as to limit dependence on the graphics card. Affinity was forced on the physical processors, as Crysis is down by around 3% otherwise.

Penalised by their clocks and L2 cache size, the Athlon II X4s don’t really turn up in Crysis and are behind even the E6300. Indeed, they can’t even outdo the Athlon II X2 250, the gain given by the presence of 4 cores being completely undermined by their clocks and the lack of L2.

Arma 2

New to our test protocol, Arma 2 is configured with all settings at a max including max visibility (10 km), which brings the configurations to their knees. Resolution stays at 800x600 to avoid the graphics card impacting on performance. Note that you must disactivate HT or force the affinity for this game, as you’ll lose 15 to 20% of fps otherwise and it will suffer from rather disagreeable stuttering.

In Arma 2 things are different with the Athlon II X4s doing better than the Athlon II X2 250. Even better, the 620 is on a level with the Q9400 and the 630 on a par with the Q9550

Page 9
GTA IV & Anno 1404

Grand Theft Auto IV

GTA IV is also new to our protocol, chosen for its weight and multi-threading optimisations. Once again all the settings were pushed to a maximum, with the exception of the textures so as not to exceed available video memory, all at a res of 800x600.

GTA IV takes advantage of the presence of four cores quite well, allowing the Athlon II X4s to do well, the 620 and 630 sandwiching the Q9400.

Anno 1404

Our last test game is Anno 1404, a strategy game tested at max settings but with resolution still at 800x600. We use a saved game with a city of 46,600 inhabitants that we partly visualize from a distance.

Here, the Athlon II 620 and 630 do better than the Core 2 Q8200.

Page 10
Average, Athlon II X2 vs Phenom II X2

Although individual app results are worth looking at, we calculated a performance index based on all tests with the same weight for each test. An index of 100 was set to the Intel Core 2 Q6600.

Overall, the Athlon II X4 620 is situated at the same level as the Core 2 Q8200: not bad seeing as it is less expensive! Dual core processors have been included in our test protocol for the first time and you can see that the Athlon II X2 250 is not far behind the Core 2 E7500. What’s more, with applications increasingly multithreaded and in spite of its clock, the Core 2 E8600 is a little further behind the entry level quad cores than it was in our old test protocol.

Page 11

It’s been clear for a while that AMD can’t compete with Intel in terms of pure performance. Although The Phenom II X4 965 did make up some ground for AMD against Socket 775, with a significant increase in energy consumption, the LGA1156/1366 processors are still at a significantly higher level. What’s more, as the Phenom II is fairly expensive to make because of its die size, it was difficult for AMD to compete at quad-core entry level.

As a result, even though it has fewer available production lines than Intel, AMD chose to develop a third 45 nm die, at the same time as leaving itself the option of selling Phenom IIs with the L3 disactivated (the rest of the spec identical) in the Athlon II X4 range. On the standard versions, the lack of L3 means AMD can go from 758 to 300 million transistors and from 258 to 169 mm²… as against 164 mm² for the entry level Intel quad cores. Coupled with an aggressive pricing policy, this allows AMD to offer the first quad core for under $100, the Athlon II X4 620.

This makes it a good deal cheaper than the Intel Core 2 Q8200 that you won’t find for under $140 but which gives a similar level of performance! The AMD offer is by far the best here then. Of course in some rare cases, such as WinRAR or Crysis in our protocol, the cache and clock of this processor bring performance down, but this is generally outweighed by performance elsewhere (even in games such as GTA IV and Anno 1404) given by the presence of 4 cores.

At a time when the number of applications that can fully use all four cores is increasing, AMD has scored a hit with the Athlon II X4 620! It remains to be seen if Intel will react by reducing its margin on the Core 2 Quads, or if we’ll have to wait for the 32 nm Core i3s that are planned for the beginning of 2010.

Copyright © 1997-2014 BeHardware. All rights reserved.