AMD's Radeon series 4800 chipset

As you guys know by now ATI's Radeon HD 4850/4870 are both using the same GPU (graphics processor). The codename for these chips (ASIC) is RV770. The Radeon HD 4890 uses the RV790 GPU. Honestly, we feel it's the exact same GPU yet a better yielded version which can take a slightly higher voltage and clock much higher.

Taking too long to download? Order a driver CD containing all of these installation files or buy high-speed priority download access. Installation Instructions. To install the ATI Radeon HD 4890 Series Video Card driver, download the version of the driver that corresponds to your operating system by clicking on the appropriate link above.

AMD put nearly a billion transistors into the GPU, it is still built upon a 55nm (260 mm2 Die size) production, and not 40nm as rumored earlier. The chip is 16 mm wide and high. Actually quite large, for a 55nm product.

Being based off that RV770 graphics processor the Radeon 4890 series graphics processors also have 800 scalar processors and forty texture units (16 in last-gen architecture). The stream/compute/shader processors (can we please just name them all shader processors?) definitely had a good number of changes; if you are into this geek talk, you'll spot 10 SIMD clusters each carrying 80 32-bit Shader processors (this accumulates to 800). If I remember correctly, one SIMD unit can handle double precision.

Much like the NVIDIA GTX 200 architecture, the scalar stream processors per SIMD unit have 16KB of local data cache/buffer that is shared among the shader processors. Next to the hefty shader processor increase you probably already notice the massive amount of texture units. In the last generation product we noticed 16 units, the 4800 series has 40 units.

It's a bit inaccurate to do but divide the number of ATI's scalar shader processors with the number 5 and you'll roughly equal the performance to NVIDIA's stream processor. You could (in an abstract way) say that the 4800 series have 160 shader units, if that helps you compare it towards NVIDIA's scaling. Again there's nothing scientific or objective about that explanation.

Anyhow. With 800 shader processors combined with some new higher clock frequencies it can produce the raw power of 1360 GigaFLOPs in simple precision. Well, depending on how that is measured of course.

Let's make a comparison chart for you to understand the differences between the Radeon HD 4870 and 4890:

ATI Radeon HD 4850	ATI Radeon HD 4870	ATI Radeon HD 4890
# of transistors	956 million	956 million	959 million
Stream Processing Units	800	800	800
Clock speed	625 MHz	750 MHz	850 (and higher)
Memory Clock	1980 MHz (effective)	3600 MHz (effective)	3900 MHz (effective)
Math processing rate (Multiply Add)	1000 GigaFLOPS	1200 GigaFLOPS	1360 GigaFLOPS
Texture Units	40	40	40
Render back-ends	16	16	16
Memory & type	512MB GDDR3	512/1024MB GDDR5	1024MB GDDR5
Memory Bandwith	64 GB/s	115 GB/s	125 GB/s
Memory interface	256-bit	256-bit	256-bit
Fabrication process	55nm	55nm	55nm
Power Consumption (peak)	110W	160W	190W
Power Consumption (Idle)	30W	90W	60W

So comparing apples to apples, the Radeon HD 4890 simply put is a faster clocked product, a graphics update. The core GPU frequency will be clocked 100 MHz faster at minimum (some board partners will clock at 900 MHz), and then there's an increase of the GDDR5 memory clock giving the product ~125 GB/s of framebuffer bandwidth.