- Articles & News
- For IT Pros
- Your Opinion
AMD’s GCN architecture, known for its strong compute and 3D performance, is finally being made available in the company's FirePro workstation graphics card family. Can AMD catch Nvidia? We test the two fastest FirePro cards to answer that question.
GPUs based on AMD’s Graphics Core Next (GCN) architecture have been around in the consumer space since the end of last year. Despite clear advantages in compute performance compared to Nvidia's desktop-oriented Kepler-based cards, AMD took its time transitioning its most modern architecture to the workstation space, though. Professional graphics products require heavily-validated drivers and more robust hardware than their consumer counterparts. Consequently, they’re usually late to the party, though understandably so.
AMD's current GCN-powered workstation line-up looks like this: the W9000 is its ultra high-end offering, the W8000 is considered high-end, the W7000 is a lesser high-end card, and the W5000 brings up the rear as a mid-range board. Everything else in the FirePro family is driven by previous-generation architectures. Naturally, then, as AMD phases out its older products, there's plenty of room here for additional GCN-based cards.
Today we have the two highest-end models in our lab, the FirePro W9000 and W8000, plus a number of existing FirePro cards and several Quadro boards from Nvidia (mainly, the 6000, 5000, 4000, and 2000).
In addition to our usual workstation benchmarks and compute-oriented tests, we also listened to reader feedback asking for us to run parts of our VGA Charts 2012 suite. It's not that we recommend using these workstation-optimized boards for gaming, but this should at least make for interesting reading.
We used our newly-constructed dual Opteron 4284 system sitting on a bright red Lian Li test bench for all our benchmarks. Read on to see how GCN, known for its compelling compute performance, performs against Nvidia's highest-end Quadro cards!