All of ATI's Radeon 5000-series videocards draw from a family of GPUs
called 'Evergreen', which is then divided into a smaller GPU lines based on
performance.
The fastest line is code named
'Cypress', which is itself broken up into smaller segments based on speed
tolerances and how many functional shader processors. This is a practice
called "binning" in the tech industry.
Only the 'Cypress' chips that have all
1600 stream processors functioning properly and are completely stable at 850MHz
are used in Radeon HD 5870 graphics cards.
The RV850 'Cypress' GPU is produced on a 40nm production
process that ATI first introduced in the Radeon 4000-series. Getting down to a
40nm process was important because the Radeon HD 5870 has 2.15 billion
transistors on a 334mm2 die, more than double that of last generation's RV770,
which had 956 million transistors.
This wealth of transistors comprise the Radeon HD 5870's
1600 stream processors, 80 texture units, 32 raster operators and the 256-bit
wide memory interface that connects the RV870 GPU to its GDDR5 memory.
Stock Radeon HD 5870 videocards have 1GB
of GDDR5 memory clocked at 1200MHz, however Gigabyte has bumped this speed up
1250MHz, which is then GDDR5-quad-pumped up to an even 5GHz.
All of AMD's Radeon 5000-series of
graphics cards share a number of new technologies, the most important of which
is support for DirectX 11. This latest revision to Microsoft's graphics hardware
standard brings two major additions to the DX11, tessellation and compute
shaders.
Tessellation is a method of adding
density and complexity to a 3D mesh. It can take a simple, low-poly model and
break up the geometry into more complex portions, adding another layer of detail
on top of a mesh. The end result is better looking, more detailed graphics with
a very low cost in performance.
| ATI Radeon HD 5000-series
GPUs |
|
ATI Radeon 4870 |
ATI Radeon HD 5830 |
ATI Radeon HD 5850 |
ATI Radeon HD 5870 |
Gigabyte
GVR587SO-1GD |
| GPU Code Name: |
RV770 |
RV870/Cypress |
RV870/Cypress |
RV870/Cypress |
RV870/Cypress |
| GPU Process: |
55nm |
40nm |
40nm |
40nm |
40nm |
| GPU Transistors: |
956 Million |
2.15 Billion |
2.15 Billion |
2.15 Billion |
2.15 Billion |
| TDP: |
110W |
175W |
151W |
188W |
188W |
| GPU Core Clock Speed: |
750MHz |
800Mhz |
725Mhz |
850Mhz |
950Mhz |
| Memory Speed: |
900MHz (3600MHz
GDDR5) |
1GHz(4GHz GDDR5) |
1GHz(4Ghz GDDR5) |
1.2GHz(4.8Ghz GDDR5) |
1.25GHz(5Ghz GDDR5) |
| Memory Capacity: |
512MB/1GB |
1GB
|
1GB
|
1GB
|
1GB
|
| Memory Bus Width: |
256-bit |
256-bit |
256-bit |
256-bit |
256-bit |
| Stream Processors: |
800 |
1120 |
1440 |
1600 |
1600 |
| Texture units: |
40 |
56 |
72 |
80 |
80 |
| Raster Operators: |
16 |
16 |
32 |
32 |
32 | | |
Compute shaders allow DirectX 11-based GPUs to do tasks
normally reserved for CPUs. This will play a big role in desktop applications
that harness the power of a PCs videocard, as well as enable games programmers
use the GPU to compute things like physics and AI. Again, this opens up the
possibility for more immersed gaming, but with a miniscule performance penalty
when compared to using a CPU. Now let's see just how well ASUS' EAH5850 DirectCU TOP
stands up to a little overclocking... but first, power draw.
Videocard Power Consumption
Here's
how this videocard stacks up in real life against a various other videocards for
power draw. PCSTATS measures total system power draw and compares that in 3D
loaded (max value recorded in 3DMark06) and idle states (at Windows desktop).
The power supply used in each test is an A-PFC compliant PC Power and Cooling 750W model.
 Videocard Power
Consumption (Total System Power Draw) |
| Idle at Desktop |
Watts |
Ranking |
| Palit HD4870 X2 2GB |
177 |
   |
| Asus EAH3870 X2 1GB TOP/G/3DHTI/1G/A |
153 |
|
| Asus EAH3870 X2 1GB TOP/G/3DHTI/1G/A
(Crossfire) |
222 |
|
| MSI R3870X2-T2D1G-OC |
165 |
|
 Gigabyte
GV-R587SO-1GD |
109 |
   |
| ASUS EAH5850 DirectCu Top/2DIS/1GD5 |
93 |
|
| Asus EAH4890 HTDI/1GD5/A (rev.A) |
138 |
|
| Asus EAH4890 HTDI/1GD5/A |
137 |
|
| Asus EAH4850 HTDI/512M |
120 |
|
| MSI R4830 T2D512-OC (Radeon 4830) |
108 |
|
| Gigabyte GV-R477D5-512HB |
135 |
|
| Gigabyte GV-R477D5-512HB Crossfire |
180 |
|
| Asus EAH4770 FML/DI/512MD5/A |
105 |
|
| Asus EAH4770 HTDI/512MD5/A |
101 |
|
| Gigabyte GV-R435OC-512I |
79 |
|
| Diamond Viper Radeon HD 2900XT CrossFire |
195 |
|
| Asus EAX1900XTX 2DHTV/512M/A |
160 |
|
| Asus EAX1950PRO/HTDP/256M/A |
150 |
|
| ASUS ENGTX480 2DI/1536MD5 |
104 |
   |
| Gigabyte GV-N285OC-2G1 |
113 |
|
| Gigabyte GV-N26SO-896I |
105 |
|
| ASUS ENGTX260 MT/HTDI/896MD3-A |
135 |
|
| Gigabyte GV-N26OC896H-GA |
112 |
|
| Sparkle GTX260 Core 216 |
140 |
|
| ASUS ENGTS250 DK |
120 |
|
| ASUS ENGTS250 DK in SLI |
170 |
|
| ASUS EN9600GT DI/512MD3/A |
86 |
|
| Gigabyte GV-N96TSL-1GI |
96 |
|
| nVidia Geforce 9600GTs in SLI |
166 |
|
| Asus EN9600GT Top/HTDI/512M |
152 |
|
| Palit Geforce 9600GT 512 |
151 |
|
| ASUS EN9400GT DI/1GLP |
85 |
|
| MSI NX8800GTX-T2D768E |
196 |
|
| MSI NX7950GX2-T2D1GE |
183 |
|
| MSI NX7900GTX-T2D512E |
165 |
|
| Gigabyte GV-NX76T256D-RH |
140 |
|
| 3D Loaded: |
Watts |
Ranking |
| Palit HD4870 X2 2GB |
421 |
|
| Asus EAH3870 X2 1GB TOP/G/3DHTI/1G/A |
390 |
|
| Asus EAH3870 X2 1GB TOP/G/3DHTI/1G/A
(Crossfire) |
610 |
|
| MSI R3870X2-T2D1G-OC |
330 |
|
| Gigabyte
GV-R587SO-1GD |
217 |
|
| ASUS EAH5850 DirectCu Top/2DIS/1GD5 |
197 |
|
| Asus EAH4890 HTDI/1GD5/A (rev.A) |
244 |
|
| Asus EAH4890 HTDI/1GD5/A |
225 |
|
| Asus EAH4850 HTDI/512M |
202 |
|
| MSI R4830 T2D512-OC (Radeon 4830) |
177 |
|
| Gigabyte GV-R477D5-512HB |
186 |
|
| Gigabyte GV-R477D5-512HB Crossfire |
279 |
|
| Asus EAH4770 FML/DI/512MD5/A |
158 |
|
| Asus EAH4770 HTDI/512MD5/A |
153.3 |
|
| Gigabyte GV-R435OC-512I |
109 |
|
| Diamond Viper Radeon HD 2900XT CrossFire |
549 |
|
| Asus EAX1900XTX 2DHTV/512M/A |
333 |
|
| Asus EAX1950PRO/HTDP/256M/A |
252 |
|
| ASUS ENGTX480 2DI/1536MD5 |
247 |
|
| Gigabyte GV-N285OC-2G1 |
277 |
|
| Gigabyte GV-N26SO-896I |
230 |
|
| ASUS ENGTX260 MT/HTDI/896MD3-A |
277 |
|
| Gigabyte GV-N26OC896H-GA |
215 |
|
| Sparkle GTX260 Core 216 |
262 |
|
| ASUS ENGTS250 DK |
214 |
|
| ASUS ENGTS250 DK in SLI |
316 |
|
| ASUS EN9600GT DI/512MD3/A |
148 |
|
| Gigabyte GV-N96TSL-1GI |
160 |
|
| nVidia Geforce 9600GTs in SLI |
313 |
|
| Asus EN9600GT Top/HTDI/512M |
220 |
|
| Palit Geforce 9600GT 512 |
221 |
|
| ASUS EN9400GT DI/1GLP |
109 |
|
| MSI NX8800GTX-T2D768E |
345 |
|
| MSI NX7950GX2-T2D1GE |
315 |
|
| MSI NX7900GTX-T2D512E |
277 |
|
| Gigabyte GV-NX76T256D-RH |
213 |
| |
Even with factory overclocking, more transistors and
more shader processors than any other single-GPU videocard in this chart,
Gigabyte's GVR587SO-1GD videocard is actually quite energy efficient thanks
to be produced on a 40nm production process. The videocard draws 217W under
load, less than a Radeon HD 4890. At idle power draw is 109W for the entire
test system, which is closer to what you find from economical low wattage
videocards, rather than the power guzzling extremism you'd expect from high end
videocards.
This frugal power consumption bodes well for
overclocking, and its time to see how far PCSTATS can take Gigabyte's
cherry picked RV870 GPU...
RSS Newsletter
