Nvidia (NASDAQ:NVDA) unveiled its new Ampere RTX gaming product line on Tuesday as expected. While this is strong move by Nvidia, as we discuss in this article, it leaves the door open for Advanced Micro Devices (NASDAQ:AMD) to enter the high end of the market with Big Navi.
A key focus of the new product line, unsurprisingly, is to increase performance of Ray Tracing technology that the Company introduced about two years back. Ray Tracing is a challenging task in terms of compute resources, and the Company has had to go past traditional shader technology to achieve the necessary horsepower to enable this functionality (All images in this article from Nvidia Presentation).
Although the Company introduced Ray Tracing in the last generation (Turing), the performance was not adequate at high resolution settings that gamers like to play. In the Ampere generation, Nvidia improved the underlying technology substantially in Shaders, RT cores, and Tensor cores (image below).
The increased performance has a cost. Nvidia’s new RTX chip is large. Made on a custom Samsung 8nm process, the chip is rumored to be over 600mm2. This is a massive die. Advanced Micro Devices’ Navi, for example, measures 251mm2 in a more advanced TSMC (TSM) 7nm process.
To enable this large a chip, in addition to using a more advanced process, one of the things the Company had to do was to improve the overall power efficiency of the design (image below).
The Company is claiming 1.9x performance per watt compared to the Turing generation. Readers should take note that AMD has already claimed that it will improve power efficiency by 1.5x going from current RDNA (Navi) to RDNA2 (Big Navi). It is difficult to read much into the difference between 1.9x and 1.5x improvements given there are many variables at play. However, all else being equal, AMD seems to have lost somewhat in its competitive march against Nvidia.
No doubt Nvidia was aware that AMD was claiming that it would compete in the high-end gaming market with RDNA2. So, in spite of the potential efficiency gains, Nvidia pulled a lever to improve performance of its Ampere line. Note in the image above that Ampere is rated over much higher power generation than Turing. By pushing the power level of the chip and board higher, Nvidia will make Ampere run at a significantly higher performance than Turing.
This higher power dissipation increases the cost of power electronics and cooling making Ampere boards more expensive to build than Turing. On the flip side, Samsung appears to have given a real sweetheart deal to Nvidia to win business from TSMC. The cost savings from Samsung could offset some or all of the board level expenses.
Without knowing the exact costs involved, it is difficult to tell how much more, if any, Ampere boards would cost compared to Turing boards. However, chances are that the new RTX boards will cost more to produce.
Nvidia announced three different RTX products during the unveiling (images below):
Note that the flagship 3080 and the ultra-high-end 3090 will be available later in September and the more mid-range 3070 will be available in October.
Overall, Nvidia claims strong generational gains across the product line (image below).
Effectively, Nvidia is claiming that the new mid-range 3070 is a performance level similar to that of the flagship 2080Ti from the previous generation whereas the 3080 is in a class of its own. 3090 performance is not plotted on the chart, but, of course, will be highest.
This chart can also help investors approximately estimate where AMD Big Navi could slot in. Assuming AMD’s 1.5x performance per watt claim holds, Big Navi should be about 1.5x the performance of current Navi at the same power. Navi is currently at about the same performance level as previous generation Nvidia 2070 Super. The implication here is that Big Navi could slot in between 3070 and 3080 at the same power level as current Navi. However, if AMD pulls the same trick as Nvidia and increases the power envelope, Big Navi could pull up closer to or a bit above 3080. If AMD does achieve that, it will be quite the coup as AMD has not participated at the high end of the market in many years.
While it is difficult to tell where AMD will land due to many variables involved in estimating performance, it is fair to say that AMD now seems to be in the hunt at the high end of the gaming GPU market.
Even if AMD cannot deliver at the 3080 level, it would be adding the key Ray Tracing technology to its products and reducing the functional gap with Nvidia. If AMD can deliver on Big Navi per the current promise, it could be gaining on Nvidia in GPUs as it has gained on Intel (NASDAQ:INTC) in CPUs.
Nvidia, for its part, is keenly aware of this risk and is taking no chances by bumping up the power and also by introducing massive energy burning GPU like the 3090. Nvidia is also pricing 3070 and 3080 very competitively at $499 and $699. This is very attractive pricing, and clearly the Company is looking to drive rapid adoption and big upgrade cycle with such pricing. Nvidia CEO Jen-Hsun Huang made an overt pitch to Pascal generation customers to upgrade.
Currently, we estimate that Big Navi may become available in volume in November – that is barely within two months of Ampere availability. If the assessment holds, this would be the first time in a very long time that AMD would be competitive with Nvidia on the GPU front on the same generation.
Should AMD succeed, Nvidia will have strong competition on the GPU side that will depress the company’s growth and make its valuation even more questionable. On the other hand, success here means that AMD growth can accelerate further.
Subscribers to Beyond The Hype have access to all the linked articles that may otherwise be inaccessible. For timely, cutting-edge insights, analysis and investing ideas in technology, semiconductor, solar, battery, autonomous vehicles, and other emerging technology stocks, check out Beyond the Hype. This Marketplace service gives you early access to my best investing ideas, along with event driven and arbitrage opportunities when they are most edgy and actionable.
Disclosure: I am/we are long AMD, TSM, INTC. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: I am/we are short NVDA.
Any positions disclosed by the Author are part of a diversified portfolio. Author considers lack of diversification to be one of the gravest errors made by many investors.
Shorting is a sophisticated investment strategy that requires superior investment skills and must be avoided by all but experienced investors with the appropriate skill and wherewithal.