Diamond Radeon HD4850

The Diamond Radeon HD 4850 uses a cooling solution that looks virtually identical to the reference design, which is widely used by many card makers. This suggests that the acoustic, power and thermal performance of the Diamond, which we examine in detail, applies to many models featuring the HD 4850, already considered the best bang-for-the-money video card available today.

July 7, 2008 by Lawrence Lee

Product
Diamond Radeon HD 4850
PCI-E Video Card
Manufacturer
Diamond
Multimedia
Street Price
~US$190

ATI’s Radeon HD 3850/3870
brought a significant improvement to their discrete graphics card lineup. Not
only did they perform well in 3D games, they were the first cards to support
full hardware decoding for H.264 and VC-1 video. In addition, power efficiency
was improved substantially — the HD 3850 for example used literally half
the power of its predecessors on load, and almost four times less when idle.
With the lower power draw, the amount of heat that required dissipating decreased
proportionally, resulting in GPUs that were relatively easy to cool — even
passively if an Accelero
S1
was used to replace the stock cooler.

Fast-forward 8 months, and ATI’s got a whole new batch of cards, this time
the Radeon HD 4850 and 4870, utilizing the new RV770 core. These cards feature
increased memory bandwidth, more stream processors and other improvements to
make them competitive with nVidia’s higher end offerings. Also, UVD has been
slightly improved and the HDMI audio output supports 7.1
sound.


Presenting the Diamond Radeon HD 4850.

Today we’re looking at Diamond Multimedia’s version of the Radeon HD 4850.
Diamond may be a new name to some, but they’ve been around since way back, producing
PCI-E, AGP, PCI, and even ISA video cards based on nVidia or ATI GPUs (depending
on the era). They currently also produce TV tuners and sound cards. The Diamond Radeon HD 4850 uses a cooling solution that looks virtually identical to the reference design, which is widely used by many card makers. This suggests that the acoustic and thermal performance of the Diamond can apply to a large number of Radeon HD 4850 cards in the marketplace.


Diamond’s HD 4850 comes in rather flimsy black box.

 

Diamond Radeon HD 4850: Specifications
(from the
product web page
)
PERFORMANCE
GPU HD 4850
ASIC Core RV 770
Core Clock 625
BUS PCI Express x16
Ports 2 Dual-Link DVI,Video Out [S-Video,
Composite, HDTV]
Stream Processors 800
MFG Process 55nm fabrication process
Transistor Count 666 million transistors on 55nm fabrication
proces
Cooling Active – FanCooler
MEMORY
Memory Clock 993
Memory Configuration 16Mx32
Memory Type GDDR3
Memory Bandwidth 147GB/sec
Memory Size 512MB
Memory Interface 256-bit
RAMDAC 400 Mhz
Video Quality
Cooling Active – FanCooler
Standard Slot Solution SINGLE
CrossFire CrossFire X
QuadFire Support YES
Direct-X 8.0-9.0-10.0-10.1
Open GL OpenGL 2.1
Vide Acceleration MPEG-2, MPEG-4, DivX, WMV9, VC-1, and
H.264
Max Resolution Analog 1600 x 1200
Max Resolution Digital 2560 x 1600
HDR F.Point Rendering 128
HDR Integer 16-bit integer or floating point
Anti Aliasing Modes Smoothvision HD + Adaptive AA.Up to
24x Custom Filter Anti-Aliasing
Antisotropic Modes 2x/4x/8x/16x high quality adaptive anisotropic
filtering modes (up to 128 taps per pixel)
Texture Support Texture filtering features 2x/4x/8x/16x
high quality adaptive anisotropic filtering modes (up to 128 taps per pixel)
128-bit floating point HDR texture filtering Bicubic filtering sRGB filtering
(gamma/degamma) Percentage Closer Filtering (PCF) Depth & stencil texture
(DST) format support Shared exponent HDR (RGBE 9:9:9:5) texture format support
Pixel Operation TBA
Display Support
HDTV YES
HDMI HDMI compliance / HDCP ready
HDMI MODE DONGLE
Native Display Support 10-bit
3D Resolution 2560 x 1600
TV-OUT Resolution 720
Dual Display support Hydravision 4
Holywood Q.VIDEO Yes

PACKAGE CONTENTS & CARD

Upon opening the box, the first thing we noticed was the size of the packaging.
The card itself is very thin, yet it’s surrounded by an enormous plastic clam
shell. While this does keep it safe, it looks wasteful.


Box contents.

Also included are a set
of cables and adapters, and a red cardboard envelope containing documentation
and a support disc.


The accessories.

Diamond supplies a component-out cable, S-Video to composite adapter, HDMI
and VGA dongles, and a CrossFire bridge. There is also a driver CD, short installation
guide, and a voucher for some free music from eMusic.com. Notably absent is
a 6-pin PCI-E power adapter. Diamond’s reasoning may be that any power supply
lacking this connector probably cannot deliver sufficient power for the card.


Look familiar? Sapphire HD 3850 pictured above, Diamond HD 4850 below.

The cooler
is virtually identical to the reference design of ATI’s previous mid-range offering, the HD 3850. The fan has many more blades, however, and they are straighter.
The card itself is very slightly longer: 23.4cm compared to 23.2cm for the HD 3850.


Viewed from above.

The heatsink is decorated with fearsome artwork on top of the
shroud. While this makes for a great visual, you’ll most likely never see it
after installation unless your case has a window, and is upside-down.


The back of the PCB.

The back of the PCB reveals no surprises — it looks just like the HD 3850/3870.
The cooler is held on by several philips head screws, making removal a trivial
process. Around the GPU core, a cushioned metal frame is used to prevent over-tightening.

HD4850 IN CLOSER DETAIL

With the heatsink taking up most of the room on the PCB surface and a sticker
covering it, there isn’t a whole lot to see.


Visible portion of the PCB.

The most notable feature is an extension of the heatsink jutting
out the cover the voltage regulation modules. An odd set of cylindrical fins
protrude upwards like a large array of silos.


The back panel.

Near the rear of the card, the shroud opens up, providing the main points of
exhaust. There are also four claw-shaped holes on top. The design definitely
has problems — what is the point of direct all the exhaust to the rear if
there’s no clear exit path? There’s no room for ventilation on the back panel
as it is filled with connectors, two DVI and one TV-out port.


From the sides.

There is also a small of the cooler open to the elements section
at the top of the card open to the elements.


Under the hood.

The heatsink is a one-piece solution that cools the GPU, and
with the help of some soft thermal pads, the ramsinks, VRMs, and inductors
as well. The card has the same set of mounting holes as the HD 3850/3870,
so a variety of third party coolers can be utilized if desired, though some
extra hardware may be required for the VRMs and inductors. The core is surrounded
by a shim, which can protect it from accidental damage if you decide to replace the stock
heatsink.


Rear of the heatsink.

Removing the stock cooler made the back more accessable. The
fins are fairly thin and the spacing wide.


The tiny fan.

The fan is manufactured by NKT Technologies. It’s a dual ball
bearing fan rated for 0.40A at 12V, or 4.8W, which is quite high for such a tiny fan. At the full 12V input, this fan probably spins at many thousands of RPM and is bound to be very noisy.

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 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

Measurement and Analysis Tools

The anatomy of our test platform is detailed here: Updated
VGA Card/Cooler Test Platform

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 ever produces 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 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. A B&K Sound Meter is employed
to take system noise measurements 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 RESULT

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.

VGA Test Bed: Baseline Results
(no discrete graphics card installed)
System State
CPU Temp
System Power
AC
DC (Est.)
Idle
22°C
73W
Unknown
CPUBurn
39°C
144W
115W
Ambient temperature: 21°C

Dimaond Radeon HD 4850:


HD 4850 installed.

All testing was conducted using the Catalyst 8.5 beta drivers included on the
support CD. There were no official ATI Catalyst drivers for the HD 4xxx series
at time of testing, and the 8.6 hotfix drivers available via ATI’s website proved to be
unstable during video playback, and did not pick up ATI’s HDMI audio device.

VGA Test Bed: Diamond Radeon HD 4850
System State
Fan Speed
System Noise
@1m
GPU
Temp
CPU
Temp
System Power
AC
DC (Est.)
Idle
~21%
18 dBA
80°C
26°C
130W
103W
CPUBurn
~24%
18 dBA
81°C
48°C
203W
165W
CPUBurn + ATITool
~44%
27 dBA
89°C
51°C
264W
216W
Ambient temperature: 21°C, ambient noise level:
14 dBA @1m.
System noise level without graphics card installed: 17 dBA @1m.

When we first installed the card, we were surprised by how quiet it was, after
all it’s supposed to be a more powerful card than the HD 3850, yet it uses more
or less the same cooler. The trade-off is very high
idle temperature. According to GPU-Z, the core was a blazing 80°C sitting
idle at the Windows desktop. Our test system is thermally challenging, but 80°C
is still quite high when the card isn’t even being pushed. On the plus side,
the noise level was only 1 dBA higher than our test system without a graphics
card — we could just barely tell it was there at all. Close up, the fan
generated a very faint, low-pitched hum with a slight clicking.

After the card was put under load, the fan speed reacted immediately. It was
very sensitive to temperature and increased in stages every 1-2°C above
80°C; in other words, the fan speed has a linear, direct relationship to temperature, with no hysteresis. As the GPU temperature never stays the same for more than 30 seconds
or so, this can be fairly annoying, as the fan is constantly, audibly, changing speed. After fifteen minutes of stress-testing,
the temperature finally settled at just under 90°C. By comparison, the HD
3850 achieved the same load temperature without the fan speeding up at all.
No artifacts or other anamolies were detected.

During testing, the card produced a regular high-pitched electrical beeping
when ATITool’s artifact scanner was in use, and a rhythmic whining during 3DView.
Unfortunately, this is a common problem on high-end graphics cards (HD
3850/3870, Geforce 8800/9600, etc.). Forturnately, this isn’t a huge issue as
it is easily masked during game play by the in-game sound blaring out of your
speakers and any fan that increases in speed during load. (Editor’s note: It’s also not clear that 3D demands in gameplay have the constant load of ATITool which causes the electrical noise.)

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.

Power Consumption Comparison
GPU State
ATI X1950XTX
Diamond HD 4850
Palit HD 3870
AC
DC (Est.)
AC
DC (Est.)
AC
DC (Est.)
Idle
+49W
+42W
+59W
+50W
+20W
+17W
Load
+131W
110W
+120W
101W
+84W
72W

The HD 3850/3870 showed us that high performance graphics card could also be
frugal when it came to power consumption. The HD 4850 is a complete reversal,
bringing us back to the days of the X1900 and HD2900 series. Idle power consumption
was horrendous at 50W, higher than any card we’ve tested.
The load power
was a far more acceptable 101W, which is not bad considering the HD 4850’s performance.


Installed.

The high idle power consumption made us wonder whether the card was throttling
properly in 2D mode, but GPU-Z confirmed the core and memory speeds were decreased
when idle. It was a very minor underclock, however, and judging by the power consumption
figures, it seems that ATI PowerPlay — tasked with throttling
the clocks and voltage to save power when the GPU is not being stressed — is
either not working or not fully implemented at this time. It’s possible that
the PowerPlay settings for the HD 4000 series are controlled on the driver-level
instead of BIOS-level,
as was the case of the HD 3000 series. As mentioned earlier, there were no offical Catalyst drivers for the card at time of testing.

There is an ongoing discussion about the matter in our
forums
, with some users reporting success using modified BIOSs to change
how the card throttles down. As a disclaimer, we should note that altering your
video card’s BIOS can be dangerous. Proceed with caution if you dare.

VIDEO PLAYBACK

The HD 4000 series utilizes UVD 2, an updated version of ATI’s Unified Video
Decoder to offload video decoding to the GPU. The HD 4850 handled all our playback
clips with ease. Despite high incidences of CPU usage during the Drag Race clip,
it played smooth without any skips or anamolies in either the video or audio.

Video Playback Results
Video Clip
Mean CPU Usage
Peak CPU Usage
AC Power
Rush Hour (H.264)
3%
8%
~136W
Coral Reef (WMV3)
28%
39%
~151W
Flight Sim. (WVC1)
55%
76%
~168W
Drag Race (WVC1)
73%
88%
~176W

UVD 2 though does not seem to provide a significant improvement in offloading
though, posting virtually identical CPU utilization numbers as the HD 3870 and
UVD 1. The HD 4850’s high power consumption is also evident during HD playback,
pulling between 20-40W more from the wall than the HD 3870 or nVidia’s Geforce
9600GT. As a pure playback device, it’s not very efficient, at least with the
current set of beta drivers.

Video Playback Comparison
Video Clip
Diamond HD 4850
Palit HD 3870
Asus EN9600GT
Mean CPU
Peak CPU
Mean CPU
Mean CPU
Peak CPU
AC Power
Mean CPU
Peak CPU
AC Power
Rush Hour (H.264)
3%
8%
~136W
3%
9%
~105W
2%
7%
~113W
Coral Reef (WMV3)
28%
39%
~151W
27%
41%
~118W
15%
30%
~121W
Flight Sim. (WVC1)
55%
76%
~168W
50%
78%
~130W
28%
46%
~131W

 

HDMI OUTPUT

The HD 4850 is the first ATI card that we’ve been able to get working properly with HDMI audio
in our test setup. We weren’t able to test its full 7.1 functionality
with our BenQ FP94VW monitor as it outputs sound via a stereo headphone jack.
The graphics driver wasn’t able to detect/deliver
our monitor’s native resolution of 1440×900. 1280×768 was the closest it would
allow, and thus the resulting image was slightly degraded due to scaling. The
available resolutions were more ideal for a HDTV rather than a LCD monitor equipped
with HDMI, which is probably appropriate.


Available resolutions via HDMI.

MP3 SOUND RECORDINGS

The recording begins with the ambient noise of the test
room. Please set your playback volume so that the ambient noise is almost inaudible,
then don’t adjust the volume control again. For best results, save the sound
file to your own PC, then listen.

VGA
test system with Diamond Radeon HD 4850
(at 21%/18 dBA and 44%/27
dBA @ 1m)
The MP3 is broken into four second sections beginning with the ambient in
the room (14 dBA), followed by the VGA test system without a video card installed
(17 dBA), and then idle and load with the video card installed.

Reference Comparatives

VGA
test system with Palit HD 3870
(at 30%/21 dBA, 40%/22 dBA, and 50%/26
dBA @1m)

 

FINAL THOUGHTS

Gaming: This is not our forté, so for an overall sense of how
well the HD 4850 performs in games, check out the reviews at X-bit
Labs
, Anandtech,
techPowerUp,
and HardwareCanucks.
The general consensus is that the HD 4850 is a great value in the $200 level,
defeating the 9800GTX by a slim margin most of the time. It represents a sizable
increase in performance compared to the HD 3870 or Geforce 8800GT, and is capable
of playing most games smoothly at 1680×1050 or lower with high image quality
settings. It can also make less demanding games playable at 1920×1200 and higher.

Video Playback: Video playback was good, but UVD 2 is not any more efficient
than the original UVD, at least in terms if CPU utilization/efficiency.

Cooling: The stock cooler is very quiet when the card is idle, and
as an unfortunate consequence, the idle GPU temperature is very high. In fact
it is almost as high as the temperature during load, when the fan is spinning
rapidly and producing an exorbitant amount of noise. ATI cut it a little close
in the cooling department — it should have been equipped with a dual slot
cooler like the HD 3870/4870.

Power Consumption: By our estimates, the HD 4850 requires approximately
50W when idle and up to 101W when stressed to the limit. 50W is incredibly high
for a graphics card, especially if all its doing is rendering the Windows XP
desktop. As most graphics cards lay idle in 2D mode for the majority of operation,
this is very wasteful — an entire system with well-selected components
can easily use less than 50W when idle. Hopefully this is not a permanent disadvantage
— our theory is that PowerPlay isn’t fully properly configured in the current
batch of beta drivers, and that once we have official Catalysts to work with,
this figure will drop substantially.

Overall, the HD 4850 is a very strong gaming card, and as such is hot and power
hungry. Once the games are turned off its only efficient as an energy leech,
sapping far too much power than is necessary, at least for the moment. The reference
cooler is a bare minimum solution that is just good enough to cool the GPU properly.
It is however, just poor enough to annoy users with fan noise required to cool
the card when it is put through its paces. We highly recommend it be replaced;
HardwareCanucks
installed an Accelero S1
on the HD 4850 and it ran much cooler passively than the stock cooler — it
is our top choice.

Diamond Radeon HD 4850
PROS* Good high definition playback
* Fan very quiet when idle
* Excellent 3D performance
* HDMI functionality
CONS* Heatsink barely adequate
* Fan too loud on load
* High idle power consumption?
* Drivers not quite ready?

Our thanks to Diamond
Multimedia

for the video card sample.

* * *

Articles of Related Interest
Asus EN3650 Silent Graphics
Card

Xigmatek Battle-Axe: First
Direct-Touch Heatpipe VGA Cooler

Asus EN9600GT Silent Edition
Graphics Card

ATI HD 3850 & HD 3870:
Improved Acoustics & Power Efficiency

Arctic Cooling Accelero S1 VGA
Cooler

Updated VGA Card/Cooler Test
Platform


* * *

Discuss this article in the SPCR forums.

Leave a Comment

Your email address will not be published. Required fields are marked *