Redefining Budget Gaming Graphics: ATI’s HD 4670

The recently released HD 4670 graphics card offers a modern gaming experience with good eye candy support even at high resolution. We examine a reference card straight from ATI, one equipped with dual DisplayPorts.

October 13, 2008 by Lawrence Lee

PC gaming promises the potential for constantly improved gaming experience due to the
inherent upgradeability of the hardware. It can be an expensive hobby as games and hardware push and shove each other to greater performance. To play the latest games smoothly and consistantly at high resolutions
requires a gamer to stay on top of current technology, always eyeing the next
big upgrade.

But what about the rest of us? With the demands of work, family, and what have
you, many people just don’t have time for PC gaming, or if they do, only for
brief intervals to unwind. There are also those who don’t have high resolution
monitors, and casual gamers who prefer to play games at lower resolution in
windowed mode while multitasking. For these people, the cost of a high
performance video card isn’t justified. There’s also power consumption,
and noise to deal with — high performance graphics cards are usually pigs in both regards.

What is ideal for the casual gamer is a cheap, video card with enough
oomph for them to enjoy the occasional gaming session, while staying quiet
and energy efficient when not in heavy use. With the Radeon HD 4670, ATI hopes
to fill that need. It is perhaps the fastest and most powerful budget graphics
card released, with a high clock speed, fast GDDR3 memory, and a whopping 320
stream processors — all in a nice tidy package. Our sample also supports
DisplayPort, a new digital interface standard that is license and royalty-free. While
the standard has many technical advantages, few devices support DisplayPort
currently.

(Editor’s Note: This entry on the Direct2Dell blog site provide a succint summary of The Truth About DisplayPort vs. HDMI. The author notes, in essence, that HDMI is intended as an external consumer electronics connection for HD TVs, while DisplayPort is the digital interface for connecting flat-panel displays to computer systems. The latter supports higher performance as a standard feature — every 6-foot cable supports 10.8 Gbps. Please read the linked article for more details.)

PHYSICAL DETAILS & INSTALLATION

Our HD 4670 sample came straight from ATI with no packaging of any kind. When
shopping for a HD 4670, check the specifications carefully — manufacturers
usually do not follow the reference design so different HD 4670’s will have
different video outputs, clock/memory speeds, memory type, and coolers.

TEST METHODOLOGY

Our test procedure is an in-system test, designed to:

1. Determine whether the card’s cooler is adequate for use in a low-noise system.
By adequately cooled, we mean cooled well enough that no misbehavior
related to thermal overload is exhibited. Thermal misbehavior in a graphics
card can show up in a variety of ways, including:

• Sudden system shutdown, bluescreen or reboot without warning.
• Jaggies and other visual artifacts on the screen.
• Motion slowing and/or screen freezing.

Any of these misbehaviors are annoying at best and dangerous at worst —
dangerous to the health and lifespan of the graphics card, and sometimes to
the system OS.

2. Estimate the card’s power consumption. This is a good indicator of how efficient
the card is and will have an effect on how hot the stock cooler becomes due
to power lost in the form of heat. The lower the better.

3. Determine the card’s ability to play back high definition video, to see
if whether it is a suitable choice for a home theater PC.

Test Platform

• Intel
  Pentium D 930 Presler core processor. Official TDP of 95W.
• AOpen
  i945Ga-PHS motherboard – Intel i945Ga Chipset; built-in VGA.
• Gigabyte G-Power 2 Pro
  heatsink, modified with a Scythe
  Slip Stream 500RPM 120mm fan.
• Corsair
  DDR2 RAM, PC2-6400, 1024 MB
• Seagate Momentus 5400.3
  160 GB, 5400RPM, ATA/100, 2-platter notebook hard drive, suspended
• Seasonic S12-600
  ATX12V v2.0 compliant power supply, modified with a Scythe
  Slip Stream 800RPM 120mm fan @ 5V.
• Antec P180B case,
  modified in detail below.
• Nexus 120mm
  fan connected to a variable fan speed controller.
• Microsoft
  Windows XP Professional operating system
• QuickTime
  Alternative 1.81 codec package

Measurement and Analysis Tools

• ATITool
  version 0.27 Beta 4 as a tool for stressing the GPU
• FurMark 1.4 as
  a tool for stressing the GPU
• GPU-Z 0.2.8
  to show the GPU temperature and fan speed
• CPUBurn
  P6 to stress the CPU
• SpeedFan
  version 4.33 to show CPU temperature
• Cyberlink
  PowerDVD 7.3 to play video.
• Seasonic
  Power Angel AC power meter, used to measure the power consumption
  of the system
• A custom-built variable fan speed controller to power the system
  fan
• PC-based spectrum analyzer: SpectraPlus with ACO Pacific mic and M-Audio digital audio interfaces.
• Anechoic chamber with ambient level of 11 dBA or lower

Testing Procedures

Our first test involves recording the system power consumption using a Seasonic
Power Angel as well as CPU and GPU temperatures using SpeedFan and ATITool (or
just SpeedFan if a nVidia based card is used) during different states: Idle,
with CPUBurn running to stress the processor, and with CPUBurn and ATITool’s
artifact scanner (or 3D View, which produces even higher power consumption)
running to stress both the CPU and GPU simultaneously. This last state mimics
the stress on the CPU and GPU produced by a modern video game. The software
is left running until the GPU temperature stabilizes for at least 10 minutes.
If artifacts are detected in ATITool or other instability is noted, the heatsink
is deemed inadequate to cool the video card in our test system.

If the heatsink has a fan, the load state tests are repeated at various fan
speeds while the system case fan is left at its lowest setting of 7V. If the card
utilizes a passive cooler, the system fan is varied instead to study the effect
of system airflow on the heatsink’s performance. System noise measurements are made at each fan speed.

Video Playback Testing

For our second test, we play a variety of video clips with PowerDVD. A CPU
usage graph is created via the Windows Task Manger for analysis to determine
the approximate mean and peak CPU usage. If the card (in conjunction with the
processor) is unable to properly decompress the clip, the video will skip or
freeze, often with instances of extremely high CPU usage as the system struggles
to play it back. High CPU usage is undesirable as it increases power consumption,
and leaves fewer resources for background tasks and other applications that
happen to be running during playback. Power draw is also recorded during playback.

Video Test Suite

1920×816 | 24fps | ~10mbps

H.264: Rush Hour 3 Trailer 1 is encoded with H.264. It has a good mixture of light and dark scenes, interspersed with fast-motion action and cutaways.

1440×1080 | 24fps | ~8mbps

WMV3: Coral Reef Adventure trailer is encoded in VC-1 using the WMV3 codec (commonly recognized by the moniker, “HD WMV”). It features multiple outdoor landscape and dark underwater scenes.

1280×720 | 60fps | ~12mbps

WVC1: Microsoft Flight Simulator X trailer is encoded in VC-1. It’s a compilation of in-game action from a third person point of view. It is encoded using the Windows Media Video 9 Advanced Profile (aka WVC1) codec — a much more demanding implementation of VC-1.

 

1920×1080 | 24fps | ~19mbps

WVC1: Drag Race is a recording of a scene from network television re-encoded with TMPGEnc using the WVC1 codec. It features a high-paced drag race. It is the most demanding clip in our test suite.

Estimating DC Power

The following power efficiency figures were obtained for the Seasonic
S12-600 used in our test system:

Seasonic S12-500 / 600 TEST RESULTS
DC Output (W)	65.3	89.7	148.7	198.5	249.5	300.2
AC Input (W)	87.0	115.0	183.1	242.1	305.0	370.2
Efficiency	75.1%	78.0%	81.2%	82.0%	81.8%	81.1%

This data is enough to give us a very good estimate of DC demand in our test system. We extrapolate the DC power output from the measured AC power input based on this data. We won’t go through the math; it’s easy enough to figure out for yourself if you really want to.

TEST RESULTS

BASELINE, with Integrated Graphics: First, here are the results of
our baseline results of the system with just its integrated graphics, without
a discrete video card. We’ll also need the power consumption reading during
CPUBurn to estimate the actual power draw of discrete card later.

Note: In our semi-anechoic chamber, our VGA test bed now
measures 12 dBA@1m. Ambient noise is only 11 dBA.

ATI RADEON HD 4670:

At idle, the fan spun at only 970 RPM according to GPU-Z. The GPU temperature
was very low and the system AC power draw was also excellent
— only 6W higher than our baseline system without a video card installed.
When the system was stressed with both CPUBurn and ATITool, the GPU temperature
gradually increased, until at 75°C, the fan speed finally began to increase.
Eventually after fifteen minutes the fan speed settled at 2500 RPM and the GPU
temperature stayed steady at 82°C, which is acceptable.

At the suggestion of some of our readers, we also tested with the FurMark,
a 3D benchmarking and stability application, in place of ATITool. While this
did not increase the GPU temperature, it did manage to squeeze a few extra watts
out of the system’s power draw.

NOISE

When the test system was first fired up, the GPU fan was very loud, but it quickly
ramped down to acceptable levels by the time the POST screen appeared. From
a meter away it became barely audible — indeed the measured SPL was only 1 dBA
higher than without the card installed. Up close however, the fan had an unpleasant, clickysound quality, though much of it was muffled once the side panel of our case was put back in
place.

Under load, the noise was
conspicious and unpleasant even though it only registered 16 dBA. The fan
exhibited a high level of tonality, which was not masked in any way due to the low level of wind turbulence noise. The sound quality generated by the fan on our sample card was impossible to
ignore. This is typical of graphics card coolers with small fans.

Using RivaTuner,
we were able to manipulate the fan speed from 25% to 100%. 2500 RPM, the speed
which the fan settled at during load, correlated to approximately 37% in RivaTuner.
At 60%, the fan speed was 4020 RPM, and the noise level was 21 dBA. Again, the
sound was much worse subjectively than the SPL suggests. In the anechoic
chamber, with its lack of ambient noise, it was painful
to listen to the fan at this level. Strangely, at 100%, the broadband sound of airflow turbulence masked a lot of the buzzy, clicky, tonal noise, resulting in a smoother, more benign
effect even though the SPL measured 5 dB higher.

(Editor’s Note: It’s often difficult to know exactly why a fan sounds bad. There are two basic reasons, and sometimes both apply: It’s intrinsically bad sounding, ie, integral to its design, or it was damaged in transit, and thus off-balance, with notches in the bearing, etc. In this particular sample, we suspect design more than damage. But note that this may not apply to all iterations of the HD 4670, which is offered by many brands, some of whom are surely using different fan / heatsink suppliers.)

POWER

The power consumption of an add-on video card can be estimated by comparing
the total system power draw with and without the card installed on our test system. Our results
were derived thus:

1. Power consumption of the graphics card at idle – When CPUBurn is run on a system, the video card is not stressed at all, and stays in idle mode. This is true whether the video card is integrated or an add-on PCIe 16X device. Hence, when the system power under CPUBurn with just the integrated graphics is subtracted from the system power under CPUBurn with the add-on card, we obtain the increase in idle power of the add-on card. (The actual idle power of the add-on card cannot be derived, because the integrated graphics does draw some power — we’d guess no more than a watt or two.)

2. Power consumption of the graphics card under load – The power draw
of the system is measured with the add-on video card, with CPUBurn and ATITool
running simultaneously. Then the power of the baseline system (with integrated
graphics) running just CPUBurn is subtracted. The difference is the load power
of the add-on card. (If you want to nitpick, the 1~2W power of the integrated
graphics at idle should be added to this number.) Any load on the CPU from ATITool
should not skew the results, since the CPU was running at full load in both
systems

By our measurements the HD 4670 used a surprisingly low amount of power at
idle — only 3W. We’ve never encountered a card with such a low power
draw before. During load with ATITool, the card used about 38W, more
or less matching the Asus
HD 3650, the last Radeon we reviewed. Using FurMark, the power draw
was additional 2W. The HD 3650 did not match the low idle power of
the previous generation HD 3850, but
the HD 4670 has clearly leapfrogged ahead in this regard..

PowerPlay, ATI’s power management technology seemed to be very attentive with
the HD 4670. According to GPU-Z, when idle the clocks stayed at 165/250Mhz.
During video playback the core clock varied between 165Mhz and 300Mhz, and the
memory speed varied between 250Mhz and 500Mhz. At full load, clock speeds rose
up to the maximum 750/999Mhz.

Video Playback

The HD 4670 handled our video playback testbed
with ease. CPU usage during playback failry low except for our more demanding
VC-1 clips. The H.264 Rush Hour 3 trailer had a trivial effect on system resources.

In terms of CPU usage, video playback was very similar to the previous Radeons.
The HD 4670 ,however, used less power during video playback than both the
HD 3650 and 3850.

MP3 SOUND RECORDINGS

These recordings were made with a high resolution, lab quality,
digital recording system inside SPCR’s own 11 dBA ambient anechoic chamber, then converted to LAME 128kbps encoded MP3s. We’ve
listened long and hard to ensure there is no audible degradation from the original
WAV files to these MP3s. They represent a quick snapshot of what we heard during
the review.

These recordings are intended to give you an idea of how the
product sounds in actual use — one meter is a reasonable
typical distance between a computer or computer component and your ear. The
recording contains stretches of ambient noise that you can use to judge the
relative loudness of the subject. Be aware that very quiet subjects may not
be audible — if we couldn’t hear it from one meter, chances are we couldn’t
record it either!

The recording starts with 10 seconds of room ambience, followed by 10 seconds
of the VGA test system without a video card installed, and then the actual product’s
noise at various levels. For the most realistic results, set the volume
so that the starting ambient level is just barely audible, then don’t change
the volume setting again.

• ATI Radeon HD 4670 stock fan at one
  meter meter
  — idle (13 dBA@1m)
  — load (16 dBA@1m)
  — 60% (21 dBA@1m)
  — 100% (26 dBA@1m)

Comparison (with the only other VGA card recorded in the new anechoic chamber with the new microphone)

• Asus ENGTX260 stock fan
  — idle (40%/12 dBA@1m)
  — load (50%/17 dBA@1m)

FINAL THOUGHTS

Gaming: Please check out the gaming-oriented reviews at X-bit Labs, The
Guru of 3D, NeoSeeker,
and HardwareCanucks.
The general consensus is that the HD 4670 is the fastest budget graphics card
released in some time, easily defeating nVidia’s 9500GT and the previous generation ATI HD 3850, and at times challenging the more expensive nVidia 9600GT. It makes most games
playable at 1680×1050 or 1600×1200, though excels at lower resolutions.

Video Playback: Video playback was very good. The decoding hardware has
not changed from the previous HD 4xxx series cards.

Cooling: The stock cooler was effective, but really only quiet when
idle. The fan has lousy acoustic properties that become more apparent as the
fan speeds up. At full load the noise it generates is too poor to be considered
for a silent PC, though this may depend on how high you set the volume of your
speakers during gameplay. As the GPU does not run very hot, almost any quiet
third party heatsink can be used to to cool the HD 4670 GPU sufficiently.

Power Consumption: By our estimates, the HD 4670 requires approximately
3W when idle and up to 40W when stressed to the limit. The idle figure is extremely
impressive — 3W is just a bit more than adding an extra idle optical drive.
40W on load is not much either — the power supply inside even a brand name
computer should be enough to accomodate the HD 4670.

Overall, the HD 4670 is a great budget level graphics card. It performs surprisingly
well in gamings, offers good high definition playback, and its power efficiency
is second to none. The fan on the cooler is quite
poor at load, though at least it is inconspicous when idle, and easily swapped for a silent cooler like the Arctic Cooling Accelero S2. The sub-$100 price for most iterations of the HD4670 is quite good for the performance. If space is an
issue, the HD 4670’s small footprint cannot be challenged.

Our thanks to ATI
for the video card sample.

* * *

Articles of Related Interest
Asus ENGTX260: A Quiet Graphics
Card for Gamers?
Diamond Radeon HD4850
Asus EN3650 Silent Graphics
Card
Asus EN9600GT Silent Edition
Graphics Card
ATI HD 3850 & HD 3870:
Improved Acoustics & Power Efficiency
Updated VGA Card/Cooler Test
Platform

* * *

Discuss this article in the SPCR forums.