The newly released Athlon II X4 630 and 620 finally bring quad core performance to the $100 level. With respective prices of $120 and $100, the high value X4 630 and X4 630 continue AMD’s sniping of Intel.
September 24, 2009 by Lawrence Lee
| Product | AMD Athlon II X4 630 AM3 Processor | AMD Athlon II X4 620 AM3 Processor |
| Manufacturer | AMD | AMD |
| MSRP | US$120 | US$100 |
The first quad core processors from AMD, the original Phenom line, failed to
top the performance of Intel’s offerings due in part to low clock speeds and
high power consumption. Even though they were modestly priced, they still did
not offer enough value to sway most high-end PC builders to AMD’s side. As a
result, Intel never felt the need to compete with AMD on price, and continued
to attach a price premium to their quad core CPUs. Phenom
II helped AMD get back into the game, but they continued too kept their
quad cores at a distance with $200+ price-tags. The poor, unwashed masses with
only $100 to spare could not dream of having any more than a simple dual core
CPU.
|
|
|
|
The newly released Athlon II X4 630 and 620 finally brings quad core computing
to the $100 level. With respective prices of $120 and $100, the 2.8 GHz X4 630
and 2.6 GHz X4 620 takes a serious stab at undercutting the $140 Q8200, currently
Intel’s cheapest quad core. Like the Athlon II X2, the X4 models lack the ample
L3 cache of the Phenom II’s. While this will undoubtedly have an impact on performance,
it also makes manufacturing these CPUs much cheaper. With prices like these,
even the most frugal PC builders will be compelled to take a look.
| Comparison Table: $100~$150 Desktop Processors | |||||
| Model | Clock Speed | Total L2 Cache | Total L3 Cache | TDP | Street Price* |
| Phenom II X4 810 | 2.60 GHz | 2MB | 4MB | 95W | $145 |
| Core 2 Quad Q8200 | 2.33 GHz | 4MB | N/A | 95W | $140 |
| Athlon II X4 630 | 2.80 GHz | 2MB | N/A | 95W | $123 |
| Phenom II X3 720 BE | 2.80 GHz | 1.5MB | 6MB | 95W | $119 |
| Core 2 Duo E7500 | 2.80 GHz | 3MB | N/A | 65W | $120 |
| Phenom II X2 550 BE | 3.10 GHz | 1MB | 6MB | 80W | $102 |
| Athlon II X4 620 | 2.60 GHz | 2MB | N/A | 95W | $99 |
| Pentium E6500 | 2.93 GHz | 2MB | N/A | 65W | $95 |
| * best prices from Newegg/PriceGrabber as of September 21st, 2009 | |||||
Besides the Q8200, the X4 630’s main competition comes from its own triple
core cousin, the X3 720 Black Edition. The lower clocked X4 620 really has no
equal as its price puts it squarely in dual core country. Its performance against
its closest rivals, the Pentium E6500 and X2 550 Black Edition will depend mainly
on the application, and whether it supports multiple threads.
|
|
|
|
TEST METHODOLOGY
Common Test Platform:
- Zerotherm Zen FZ120
CPU cooler - Corsair
XMS3 memory 2x2GB, DDR3-1600 @ 1333MHz, 9-9-9-24 (1066 MHz,
8-8-8-20 for LGA775) - Asus EN9400GT Silent Edition
graphics card – 512MB - Western Digital VelociRaptor
hard drive – 300 GB, 10,000 RPM, 16MB cache, SATA - Seasonic
SS-400ET ATX power supply - Microsoft
Windows Vista SP1 operating system – Home Premium, 32-bit - nVidia
Forceware 190.26 graphics driver
Intel LGA775:
- Intel
Core 2 Quad Q8200S processor – 2.33 GHz, 45nm, 65W - Intel
Core 2 Quad Q6600 processor – 2.40 GHz, 65nm, 95W, G0 stepping - Asus
P5Q3 motherboard – P45 chipset
AMD AM3:
- AMD
Phenom II X4 810 processor – 2.60 GHz, 45nm, 95W - AMD Athlon II X4 630 processor – 2.80 GHz, 45nm, 95W
- AMD Athlon II X4 620 processor – 2.60 GHz, 45nm, 95W
- AMD
Phenom II X3 720 Black Edition processor – 2.80 GHz, 45nm,
95W - AMD Phenom II X2 550 Black Edition
processor – 3.10 GHz, 45nm, 80W - Asus M4A78T-E motherboard
– 790GX chipset (onboard video disabled)
Measurement and Analysis Tools
- CPU-Z
to monitor CPU frequency and voltage. - CPUBurn
processor stress software. - Prime95
processor stress software. - Cyberlink
PowerDVD to play H.264/VC-1/Blu-ray video. - Eset NOD32 as
an anti-virus benchmark. - WinRAR as an
archiving benchmark. - iTunes
an audio encoding benchmark. - TMPGEnc
Xpress as a video encoding benchmark. - PCMark05
as a general system benchmark. - Seasonic
Power Angel AC power meter, used to measure the power consumption
of the system. - Custom-built, four-channel variable DC power supply, used to power
and regulate the CPU fan.
Benchmark Test Details
- Eset NOD32: In-depth virus scan of a folder containing 32 files of
varying size with many of them being file RAR and ZIP archives. - WinRAR: Archive creation with a folder containing 68 files of varying
size (less than 50MB). - iTunes: Conversion of an MP3 file to AAC.
- TMPGEnc Xpress: Encoding a 1-minute long XVID AVI file to VC-1 (1280×720,
30fps, 20mbps).
Our testing procedure is designed to determine the overall system power consumption
at various states (measured using a Seasonic Power Angel). To stress CPUs we
used Prime95 (large FFTs setting) or CPUBurn (which produced the higher power
draw). We also performed a short series of benchmarks featuring real-world timed
tests and synthetics.
Cool’n’Quiet and/or Intel SpeedStep were enabled (unless otherwise noted).
The following features/services were disabled during testing to prevent spikes
in CPU/HDD usage that are typical of fresh Vista installations:
- Windows Sidebar
- Indexing
- Superfetch
Undervolting
Testing was also conducted with the main processors undervolted as low as possible
while remaining stable enough to survive Prime95’s torture test for 10 minutes.
Power savings from undervolting can be significant and the reduction in thermal
dissipation also makes for easier cooling. We managed to undervolt our X4 630
from 1.325V down to 1.200V. Our X4 620 sample had a higher stock voltage, 1.400V,
which we were able to decrease to 1.1625V.
|
CPU-Z screenshot: X4 620 undervolted to 1.1625V from 1.400V. |
TEST RESULTS
Our test systems consist of 2x2GB of DDR3 memory, a GeForce 9400GT graphics
card, WD VelociRaptor hard drive and an OEM Seasonic power supply. Unfortunately
we did not have a Q8200 at our disposal so instead we a Q8200S (the 45W
version of the Q8200) and a Q6600, an older 95W 65 nm processor using a slower
1066 MHz front bus, but with a higher 2.4 GHz clock speed and twice as much
L2 cache. It is safe to assume an actual Q8200 will fall somewhere in-between
these two processors in terms of power consumption. The ambient temperature
was 22°C.
| Test Results: General System Power Consumption (AC) | ||||
| Processor | Idle | VC-1 Playback | CPU Load (half cores) | CPU (all cores) |
| C2Q Q8200S | 64W | 75W | 99W | 115W |
| X4 620 (UV)* | 66W | 86W | 99W | 124W |
| X2 550 BE | 65W | 85W | N/A | 129W |
| C2Q Q6600 | 69W | 83W | 120W | 139W |
| X4 630 (UV)* | 71W | 89W | 111W | 139W |
| X4 620 | 66W | 87W | 117W | 153W |
| X4 630 | 71W | 89W | 122W | 157W |
| X3 720 BE | 76W | 97W | N/A | 153W |
| X4 810 | 79W | 99W | 127W | 159W |
| *X4 630 undervolted by 0.125V, X4 620 by 0.2375V. | ||||
Overall the X4 630 consistently used about 5W more than the lower clocked X4
620, not enough to be considered significant. Both proved to be more power efficient
than their Phenom II cousins, the X3 720 and X4 810, but only when idle or with
a light load like VC-1 video playback. It seems L3 cache creates a noticeable
energy demand. System power consumption between the two Athlon II X4’s and the
95W Q6600 were also similar except at full load where the Q6600 used 15~20W
less.
Undervolting reduced the X4 630’s full load consumption by 18W, while the X4
620 saved 29W. This was no surprise as we were able to undervolt the 620 by
almost twice the amount as the 630. When undervolted, our X4 620 sample probably
uses about the same amount of energy as a Q8200.
Performance
| Test Results: Benchmarks | |||||
| Processor | NOD32 | WinRAR | iTunes | TMPGEnc | PCMark05 |
| C2Q Q8200S | 2:59 | 3:43 | 4:20 | 3:56 | 7648 |
| C2Q Q6600 | 2:56 | 3:41 | 4:35 | 3:59 | 7740 |
| X2 550 BE | 2:34 | 3:09 | 4:44 | 4:54 | 7217 |
| X4 810 | 3:05 | 3:32 | 5:38 | 3:29 | 7756 |
| X4 630 | 3:10 | 3:58 | 5:21 | 3:19 | 8203 |
| X3 720 BE | 2:47 | 3:16 | 5:13 | 5:08 | 7738 |
| X4 620 | 3:23 | 4:05 | 5:44 | 3:32 | 7660 |
Our brief benchmark suite favored the Q8200S over the X4 620/630. The Q8200S
held a large lead over the two Athlons in our iTunes encoding test, and posted
smaller wins in our anti-virus and file archiving tests. The X4 620 and 630
beat Intel’s offerings in video encoding with TMPGEnc — the type of task
where quad cores typically shine. Overall we’d say the Athlon II X4’s need another
200~300 MHz to truly compete with the Q8200 in our test suite, but if you typically
use more thread-aware applications, the balance will tip in the other direction.
The 2.8 GHz Athlon II X4 630 performed similarly to the 2.6 GHz Phenom X4 810
— an extra 200 MHz in clock speed seems to make up for the 630’s lack of
L3 cache. It should be noted however that the X4 800 series has 4MB of L3 cache
while the 900 series sports 6MB.
Average/Total Benchmark Power Consumption
| Average Benchmark Power Consumption | ||||
| Processor | NOD32 | WinRAR | iTunes | TMPGEnc |
| C2Q Q8200S | 75W | 84W | 78W | 96W |
| X4 620 UV* | 85W | 82W | 88W | 104W |
| C2Q Q6600 | 86W | 99W | 92W | 116W |
| X4 630 UV* | 93W | 90W | 97W | 116W |
| X4 620 | 93W | 89W | 99W | 124W |
| X4 630 | 99W | 95W | 106W | 131W |
| X2 550 BE | 106W | 105W | 108W | 118W |
| X4 810 | 109W | 104W | 112W | 137W |
| X3 720 BE | 112W | 111W | 118W | 130W |
| *X4 630 undervolted by 0.125V, X4 620 by 0.2375V. | ||||
| Total Benchmark Power Consumption (watt-hours) | |||||
| Test State | NOD32 | WinRAR | iTunes | TMPGEnc | Total |
| C2Q Q8200S | 3.73 | 5.20 | 5.63 | 6.29 | 20.85 |
| X4 620 UV* | 4.79 | 5.58 | 8.41 | 6.12 | 24.91 |
| C2Q Q6600 | 4.20 | 6.08 | 7.03 | 7.70 | 25.01 |
| X4 630 UV* | 4.91 | 5.95 | 8.65 | 6.41 | 25.92 |
| X4 620 | 5.24 | 6.06 | 9.46 | 7.30 | 28.06 |
| X4 630 | 5.23 | 6.28 | 9.45 | 7.24 | 28.20 |
| X2 550 BE | 4.53 | 5.51 | 8.52 | 9.64 | 28.20 |
| X4 810 | 5.60 | 6.12 | 10.52 | 7.95 | 30.19 |
| X3 720 BE | 5.20 | 6.04 | 10.26 | 11.12 | 32.62 |
| *X4 630 undervolted by 0.125V, X4 620 by 0.2375V. | |||||
When we calculated the total number of watt-hours used during our tests, we
found that the new Athlon II’s beat their Phenom II cousins — the X4 810
by 7% and the X3 720 by 14%. The dual core X2 550 matched their energy efficiency
due its superior performance in the simpler applications in our test suite,
thanks to its L3 cache and higher clock speed. The Q6600 used about 11% less
while the 65W Q8200S dominated both Athlons by 26%. Undervolting brought the
two processors basically on par with the Q6600.
Thermals
| Temperatures | ||||
| Sensor | X4 630 | X4 620 | ||
| Stock | UV* | Stock | UV* | |
| Core | 35°C | 31°C | 34°C | 28°C |
| CPU | 41°C | 38°C | 41°C | 38°C |
| Zerotherm FZ120 with Nexus fan @ 950 RPM. Ambient temperature: 22°C. *X4 630 undervolted by 0.125V, X4 620 by 0.2375V. | ||||
In our open testbed, cooled by a Zerotherm FZ120 heatsink using a Nexus 120
mm fan running at ~950 RPM, the temperatures reported by SpeedFan were fairly
low. The heatsink body was also cool to the touch. When stressed with CPUBurn,
both processors stabilized at 41°C (CPU sensor) 38°C (motherboard sensor).
Undervolting resulted in a 6-7°C decrease for the X4 630, and slightly more
for the X4 620. The Nexus fan was practically inaudible from one meter — with
a good heatsink keeping the new Athlon II’s cool without much noise pollution
is a relatively simple matter.
FINAL THOUGHTS
Our timed benchmarks showed evidence that the Athlon II X4 630 and 620 perform
slightly slower than Intel’s cheapest quad core, the Q8200, in mundane everyday
tasks like anti-virus scanning and file archiving. However, there is $20~$40
price difference, so from a value standpoint, they’re about the same. On the
other hand, the Q8200 fell to both the 630 and 620 in our most demanding test,
which involved video encoding. This result was echoed by benchmarks performed
by Xbit
Labs and Anandtech
which showed that video encoding and rendering applications heavily favors AMD.
If video is your bag, but you’re hurting for cash, a X4 630 or 620 may be just
what the doctor ordered.
These new processors also complicate AMD’s lineup, with the big loser being
the Phenom II X3 series. The X4 630 is virtually the same price as the X3 720,
but it has an extra core and is more energy efficient. Pinned between the X4
640 and 630 as well as the higher-clocked Phenom II X2’s, there’s really no
place for the X3 720 anywhere in AMD’s lineup, not unless price-cuts are initiated
across the board.
Power consumption has been improved compared to the Phenom II’s, but they’re
still a step behind Intel in this regard. The difference when idle and under
light load is small though and as that’s how most systems operate during the
majority of their lifetime, it’s almost a non-issue. If you prefer to run your
system with as little power as possible, undervolting is an option. Our 630
and 620 samples undervolted well, resulting in a 8% and 11% improvements in
overall energy efficiency respectively during our timed benchmarks. Whether
at stock settings or undervolted, neither chips were difficult to cool quietly.
Removing the L3 cache from their Phenom II’s has resulted in better energy
efficiency and lower manufacturing costs which have been passed onto the consumer.
Best of all, AMD did not sacrifice much in the way of performance. The Athlon
II X4 630 and 620 represent the beginning of a new front in the age-old processor
war. They perform well enough to compete with Intel’s cheapest quad core CPU,
but with a lower price-tag. With the X4 630 and 620 going for $120 and $100
respectively, it seems that they want to put an AMD quad core processor in every
desktop. If Intel wants a piece of the action, they’ll have to lower themselves
to AMD’s level.
If you have an existing compatible AM2+/AM3 motherboard with a single/dual
core processor, or an original Phenom, the X4 630 and 620’s are attractive upgrade
options. For a new system, there are a couple of issues to consider besides
the price and performance of the CPU. The first being upgradeability —
AM3 is AMD’s latest socket, while LGA775, the only socket for which Intel has
sub-$200 processors, is all but a lame duck. Intel has shifted focus to LGA1156
and they are of course pushing their LGA775 user base to upgrade, which can
be a pricey transition given the higher cost of their motherboard chipsets.
AMD motherboards on the other hand continue to deliver more features at lower
cost, and most of them include a superior integrated graphics solution, such
as the HD 4200 incorporated in the 785G
chipset.
Our thanks to AMD
and for
Athlon II X4 samples used in this review.
* * *
Articles of Related Interest
AMD’s
785G Chipset Boards: 780G Evolved
Intel Q9550S: A Greener Quad Core?
AMD Phenom II X2 550 BE & Athlon
II X2 250
AMD Phenom II X4 955 Black Edition for
AM3
Phenom II: AMD pulls closer
Intel Core i7: Nehalem Launched
* * *
