CHIP DESIGNER AMD has said that its Radeon HD branding numbers are not meant to represent the architecture of the chip but rather the year and performance.
AMD announced its Radeon HD 8000 series mobile chips at CES and earlier this week included Radeon HD 8000 series branded GPUs in its latest Richland APUs despite no changes in underlying GPU architecture. Now the firm has told The INQUIRER that its Radeon HD branding is meant to represent the year of release along with the relative performance rating rather than architecture changes.
AMD's Richland APUs sport Radeon HD 8000 series GPUs despite having a core based on the Northern Islands architecture, whereas the firm's discrete GPUs moved onto its newer, more powerful Southern Islands Graphics Core Next (GCN) architecture and are branded as Radeon HD 7000 series and more recently Radeon HD 8000 series parts.
Kevin Lensing, director of AMD's Notebook Products business unit told The INQUIRER that the firm's Radeon HD product numbers are simply trying to convey performance to consumers who don't care about what architecture the GPU is based on.
Lensing said, "For us the important thing about the graphics brand, you've got to put yourself in the shoes of a real end user who is trying to buy a computer rather than a tech journalist who is trying to figure out what the architecture is beneath it.
"Our branding is not about the architecture, end users don't care, they care what is the delivered performance."
Continuing, Lensing broke down AMD's strategy when it comes to Radeon HD product numbers, He said, "For us, the 8 in [Radeon HD] 8000 series doesn't stand for Graphics Core Next, it doesn't [apply] for the discrete graphics either.
"It stands for the product line that we are delivering in the year 2013. With the 8 series you know it is attached to a 2013 GPU or APU, the second number is where it performs in that tiering, regardless of whether it is a Northern Islands, Southern Islands, Sea Islands, Graphics Core Next or whatever, the end user doesn't care about that. The point is how does it perform and how to simply communicate performance."
Lensing said that the firm's GCN architecture is more efficient than the previous generation that is in Richland, but added that the GPU in Richland still held its own against the competition. While Lensing didn't mention the competition by name, it's fairly obvious that AMD's APUs go up against Intel's integrated chips, which despite a GPU update in Ivy Bridge still lag behind AMD's previous generation GPU architecture.
Lensing added that Richland will be the last APU with a non-GCN GPU, with the firm already having said that Kabini will be tipping up later this year.
Given AMD's branding scheme, it will mean that AMD will brand that GPU, with its all new architecture, as a Radeon HD 8000 series graphics core. µ
Uses 20 percent less power than traditional systems
It's becoming more prevalent in car research and development
Sign up for INQbot – a weekly roundup of the best from the INQ