As I write this, benchmarks are emerging for Nvidia's (NVDA) newest mobile chip, the Tegra K1. At this point, the benchmarks seem to paint a very favorable picture for Nvidia, with the Tegra K1 showing well ahead of competition. This is especially noticeable in graphics-heavy work, as can be seen in the following benchmark (Source: WCCF tech)
The Tegra K1 is based on the ARM A15 on the CPU side, and Nvidia's Kepler graphics architecture on the GPU side. What this means is that while Nvidia cannot be said to have any kind of exclusive performance lead on the CPU, the GPU architecture is its own and gains from Nvidia's desktop leadership. Nvidia is only challenged in the desktop GPU space by AMD, but AMD doesn't have a horse in the mobile GPU race. What we can infer is that it might be possible that Nvidia's presence in the competitive higher-performance desktop GPU space is giving it the pressure it needs to field a very competitive GPU in the mobile space.
In the benchmark above, it's important to notice that the 3 competitors closest to the Tegra K1 are desktop chips. Apple's iPad Air is the first mobile competitor and it shows less than half the fps (frames per second). Apple (AAPL) uses GPU IP from Imagination Technologies (OTCPK:IGNMF), which up until recently was the most competitive mobile GPU provider.
Here, a little history is warranted. Imagination Technologies's GPU tech was also borne from desktop GPU technology. Back in the day, PowerVR fought tooth and nail against 3dfx interactive's products. But both were wiped out from the market by Nvidia and ATI (now bought by AMD). This thus raises the chance that the same thing might again happen in mobile GPU land, with Nvidia's technology ultimately proving to be superior. The Tegra K1 seems to indicate as much, though statistics in power consumption are still needed.
If, indeed, Nvidia's GPU proves to be superior, this has implications
There are two sets of consequences if Nvidia gains the GPU upper hand:
- First, such a development would help Nvidia sell Tegra chips, which have had a horrendous performance as of late. The Tegra segment saw its revenues drop 54% from year-ago levels and this gave rise to a massive $132.8 million operating loss in the segment. Nvidia thus badly needs this mobile victory (Source: Latest 10-Q);
- Second, Nvidia has said it is open to selling GPU IP much like Imagination Technologies does. If Nvidia's tech proves superior then it wouldn't be too far-fetched to think it could displace Imagination Technologies' own IP. For instance, Apple has never shown itself willing to remain with an inferior solution for long, and presently it uses Imagination Technologies' IP. This could change in Nvidia's favor if its performance leadership is real (barring no bad news on the power consumption front).
Given the likely technological advantage on Nvidia's side, neither of these consequences seems impossible. We still need to see confirmation of these early benchmarks - which are even said to understate the Tegra K1's performance.
The just-released benchmarks on the Nvidia Tegra K1 are somewhat positive for Nvidia, and potentially very negative for Imagination Technologies. If Nvidia's technology proves itself to be superior, and given Nvidia's willingness to license it, this could generate a very ugly development for Imagination Technologies' own business model.
At the same time, if these benchmarks are confirmed, they might lead to a revival of Nvidia's mobile ambitions. There remain doubts about power consumption, though.