Praying For Consoles: AMD Details 2013 Game Plan, Offers Updates on New APU Performance
- The PC market was awful.
- AMD lost money.
- The company has enough cash on hand (thanks to selling its HQ in a lease-back arrangement) to make it through the second quarter.
The Fun Stuff
Let's talk products. According to AMD, Kabini (the 28nm follow-up to Brazos) entered volume production this past quarter. Estimated performance is "2x in CPU, 4x in graphics." The competition, in this case, is almost certainly Intel's 32nm Clover Trail. Given that Brazos is typically 20-25% faster than Atom, the 2x performance figure is reasonable once you calculate the higher number of cores and IPC improvements.
28nm wafers at TSMC
According to AMD, Kabini will target Intel's Pentium, Celeron, and even the low-end Core i3 business, while Temash will principally target Atom. Again, though, the company execs were speaking to 32nm Clover Trail. Intel's own 22nm Atom should ship by the end of the year, with its own quad-core configuration and far higher performance. Still, it looks like AMD is going to get a 4-6 month window of opportunity.
Next up, there's APUs. Read claimed that AMD took back market share in desktop products and won some small amount of share in desktop graphics as well. The net impact of the "Never Settled" bundle was positive on revenue; AMD reported $337M in GPU sales, up from $326M in Q4 2012. It also believes it took some share in workstation graphics, and those cards typically carry higher margins. That proposed gain makes the downward pressure on GPU operating income more troubling. AMD reported $16M in operating income, down from $22M in Q4 2012. When you factor in that the company made $11M more in GPU sales over the same period and claimed a slight average selling price (ASP) increase, it implies that "Never Settled" is costing the company some extra scratch.
"Never Settled," according to AMD executives, is a vital part of the company's long-term plan for gaming and graphics. AMD's executives have stated that they expect developers to leverage the synergy between the Wii U, Xbox Durango, and PS4's graphics hardware to create games that run better on AMD products over on the PC side of the business.
Finally, there's the consoles.
Rory Read really likes the console/embedded business. It's clear that he views this as the future of AMD. The company projects that up to 20% of its revenue will come from embedded/console sales in 2013, with as much as 50% of all revenue coming from these sources by the end of 2014. AMD, in other words, is becoming an embedded processor designer with an x86 business, rather than the other way around.
What Rory Read didn't want to hear was concern from investors that the next generation of consoles might not pan out. When one analyst questioned whether or not historical trends were a good predictor of console uptake, Read told the man to go consult historical trends.
That was rather stupid. The entire PC industry just suffered a historic collapse in sales. The current generation of consoles is seeing a similar plummet. Xbox 360, PS3, and 3DS sales all fell 19-22% in March. Wii sales fell 51%, while the Wii U moved just 55,000 units. The Vita managed 42,500 units (a drop of 80%). Total sales for all consoles were down 22% compared to this time last year. The most recent entries in the Gears of War and God of War series (Gears of War: Judgement and God of War: Ascension) are estimated to have sold 425,000 and 360,000 copies since they launched in late March. That's staggeringly terrible. Gears of War 3 sold over 3 million copies in its first month when it debuted in 2011. God of War 3 sold nearly a million copies within days.
There's a really, really big problem with the way these figures could impact AMD's bottom line, and Read refused to answer whether the company's agreements with Sony, Microsoft, and Nintendo were structured in a way that would insulate it from abysmal console sales.
AMD's upcoming products sound strong. There's definitely some excitement around future console developments. We expect launch sales will be brisk, on both platforms. Nevertheless, AMD's entire strategy is based on the idea that people are going to buy two sorts of things in the next year: Consoles and Windows tablets.
If that happens, AMD could come out of this on the best footing its enjoyed in years. If it doesn't, there won't be an AMD anymore.
There was no mention of Kaveri, by the way. Those of you curious about what AMD has planned for the mainstream desktop market will have to keep on waiting.