AMD Hawaii is 30% smaller than Kepler GK110

Published: 16th Sep 2013, 16:41 GMT

Hawaii GPU is 28nm

Let me start from fabrication process. It’s not a secret that 20nm is not coming sooner than Q2 2014. An exclusive interview with Forbes revealed a long known fact, the new GPU will be made in 28nm (Read here).

Quotes from Matt Skynner – Corporate Vice President & GM, Graphics BU and AMD Canada General Manager

“Another thing I can tell you is about the process node: this GPU is in 28nm. Some have speculated that it was 20nm and it’s not for a specific reason: At 28nm for an enthusiast GPU, we can achieve higher clock speeds and higher absolute performance.”

By reading between the lines it’s more or less confirmed that AMD could already have first 20nm samples, and they know it’s not efficient enough (yet) for mass production.

Hawaii GPU is 30% smaller than Kepler GK110

This is probably the most interesting part of the whole review. According to matt Skynner their new GPU is smaller than TITAN. Considerably smaller. It should be around 423-432mm2. This basically means that it’s 20% bigger than Tahiti.

“It’s also extremely efficient. [Nvidia’s Kepler] GK110 is nearly 30% bigger from a die size point of view. We believe we have the best performance for the die size for the enthusiast GPU.”

Quick calculations of Stream Processors (this is not a proper way of calculating number of components or complexity of the circuity, treat it just for a reference).

Tahiti is 352mm (365mm2 as some of you noted). Kepler GK110 is 551mm2 (561mm2 as some of you noted).

  • 2048 SPS * 15% ~= 2355 => 2304 SPs
  • 2048 SPS * 20% ~= 2457 => 2560 SPs

Hawaii GPU won’t be cheap, but it won’t be expensive either

In the same interview it was said that AMD will focus on enthusiast segment, rather than ultra-enthusiast. What is basically means is that their new flagship will not be cheaper than HD 7970*, but it won’t anywhere near TITAN price point as well:

“I can’t reveal a pricepoint but we’re looking at more traditional enthusiast GPU pricepoints. We’re not targeting a $999 single GPU solution like our competition because we believe not a lot of people have that $999. We normally address what we call the ultra-enthusiast segment with a dual-GPU offering like the 7990. So this next-generation line is targeting more of the enthusiast market versus the ultra-enthusiast one.

* It depends what they actually mean by enthusiast segment. It could cost the same but not less.

AMD is preparing their its own GeForce Experience-like application

Last but not least it was also mentioned that AMD is working on their own gaming optimizing utility. Of course GeForce Experience is much more than that (driver updates or ShadowPlay). Either way, it’s no secret that AMD is copying many things from NVIDIA, rather than making their own ideas.

I’m a proponent of NVIDIA GeForce Experience. I know how to tweak games manually to get the graphics performance I want, but the idea of that software lowering the barrier of entry to PC gaming is exciting. Have you considered anything similar to that for your users?

“The answer is yes. And there could be more on that later…”

Matt Skynner is also talking about drivers, Never Settle program and possible 3rd party cooling solutions on Radeon graphics. It’s nothing worth quoting, but you can read it by following the link below.

Source: Forbes


by WhyCry

Previous Post
AMD Radeon R9 280X to replace HD 7970, features 3GB GDDR5 memory
Next Post
Leadtek joins High-End GTX 780 club with WinFast GTX 780 O.C Triple Fan






Back to Top ↑