Beating AMD and Analyst Estimates

AMD finished 2017 with a strong quarter and profitable year

January 30th has rolled around and AMD released their Q4 2017 results. The results were positive and somewhat unexpected. I have been curious how the company fared and was waiting for these results to compare them to the relatively strong quarter that Intel experienced. At the Q3 earnings AMD was not entirely bullish about how Q4 would go. The knew that it was going to be a down quarter as compared to an unexpectedly strong third quarter, but they were unsure how that was going to pan out. The primary reason that Q4 was not going to be as strong was due to the known royalty income that AMD was expecting from their Semi-Custom Group. Q4 has traditionally been bad for that group as all of their buildup for the holiday season came from Q1 and Q2 rampings of the physical products that would be integrated into consoles.

The results exceeded AMD’s and analysts’ expectations. They were expecting in the $1.39B range, but their actual revenue came in at a relatively strong $1.48B. Not only was the quarter stronger than expected, but AMD was able to pull out another positive net income of $61M. It has been a while since AMD was able to post back to back profitable quarters. This allowed AMD to have a net positive year to the tune of $43M where in 2016 AMD had a loss of $497M. 2017 as a whole was $1.06B more in revenue over 2016. AMD has been historically lean in terms of expenses for the past few years, and a massive boost in revenue has allowed them to invest in R&D as well as more aggressively ramp up their money making products to compete more adequately with Intel, who is having their own set of issues right now with manufacturing and security.

The strongest area was of course sales of Ryzen processors and Radeon GPUs. Polaris continues to sell strongly as well as the latest Vega units. Semi-Custom is down due its seasonality being about a quarter behind that of typical PC components. Ryzen features strong performance and an advantage in price/performance and price/core. These chips also seem to show strong yields and bins. The combination of gaming and cryptocurrency have allowed AMD to have their strongest quarter in GPUs in the company’s history. Cards are often bought in bulk and there simply has not been enough product to supply current demand. Many fear that another bubble is on the way and partners will be left with excess inventory due to decreased demand and pressure from sales on the secondary market of used cards. For the time being AMD is trying to balance out production with forecasted demand. Currently the market is not entirely happy with the situation. It will improve for consumers eventually, but right now the combination of demand and high prices have kept product flowing and partners and retailers happy with revenue.

EPYC has also made its presence felt with growing sales throughout Q4 and the ramping of product to meet demands of the major server OEMs that have either certified these designs or are very close to it. AMD expects EPYC to do very well and they hope to capture single digit percentage of the server market within the first half of this year. This is actually an impressive number considering that before EPYC AMD had essentially been absent in the enterprise market. Marketshare was around 1% or less with their outdated Opteron offerings based on the 5 year old Piledriver architecture. Getting anywhere over 5% is going to result in significantly higher revenue and more customers willing to look at EPYC offerings. AMD is also heavily leveraging the Spectre/Meltdown situation to tout their higher level security as compared to what Intel is dealing with right now. If AMD can show they have plugged these holes effectively and without the performance hit that Intel is showing especially at the datacenter, then we could see customers changing their long term plans and start rolling out EPYC solutions several quarters from now. Enterprise does move much more slowly than consumer level demands, but when there is some momentum it typically results in consistent and strong results.

Polaris and Vega are also making inroads into the enterprise with professional cards as well as the compute market where NVIDIA has been alone with their Tesla branded cards. AMD has made some significant gains and it has had a positive effect on revenue for that sector. The company is still at a disadvantage when it comes to software support as NVIDIA has been aggressive in pushing CUDA towards universities and companies looking for turnkey solutions for their compute needs.

Q1 is typically the weakest quarter of the year due to seasonality. Namely after the glut of sales from Q4/Holiday Season, the market slows down due to lower consumer demand. This does not look to be the case for AMD. The company expects revenue to be in the $1.55B range. This is due to continued demand for Ryzen processors and Radeon GPUs, the further ramping of the EPYC platform, the ramping of mobile Ryzen parts with Vega graphics, and finally the introduction of the desktop variant of Ryzen with Vega graphics.

The biggest surprise of the past quarter was the announcement that AMD would be providing the graphics portion for the upcoming Kaby Lake G (now known as Core i5/i7 with RX Vega M Graphics). Previously I was unsure how AMD would be handling this semi-custom part. Initial thoughts were that Intel would license the design from AMD and produce the part through either GLOBALFOUNDRIES or TSMC. Today we had confirmation from Dr. Lisa Su that they were actually selling silicon directly to Intel. This means they are continuing to use GF for their 14nm manufacturing which likely helps fulfill their wafer agreement. This will show greater revenue than just licensing and royalties alone. It also opens up future collaborations with Intel until that company can improve their own internal offerings to be on a more even playing field. This also allows AMD to further break into the discrete mobile market where NVIDIA has been dominant.

Past Q1 we will see the introduction of the 12nm Ryzen based on Zen+. AMD did not confirm if any architectural changes have been implemented, but it is a pretty safe guess that it has gone through an optimization process to improve IPC, memory/cache latencies, and clockspeeds. AMD expects this refresh to continue their upward trend in sales with competitive parts as compared to Intel’s lineup at the time. EPYC will continue to be ramped and they expect double digit growth throughout the rest of the year. Vega will also be ported over to TSMC’s upcoming 7nm process and will be released more towards enterprise and machine learning applications. This could very well be a “pipecleaner” design for AMD as it prepares the Navi architecture for 2019.

Zen 2’s design is complete and will be sent off for initial manufacturing. Samples will be provided to partners in late 2018. This timeline suggests that we will see Zen 2 based parts hit the market in late Q1 or early Q2 of 2019. AMD has promised that it will be more aggressive in updating their core features to keep up with Intel. If Intel has underestimated AMD and their design prowess, perhaps we will see another “perfect storm” of trouble for Intel in terms of design combined with process deficiencies. I am not suggesting that another Athlon 64/Pentium 4 situation will occur, just that there is a chance it could with the recent struggles of Intel and their CPUs.

AMD is poised to have their strongest year in at least a decade. Analysts believe revenues to be in the $5.8B range, but if the rest of their quarters scale as Q4 has and Q1 is expected to, it is not inconceivable that AMD would reach $7B or above. Much depends on how the enterprise market reacts to Intel and their Spectre/Meltdown issues, how well Polaris and Vega compete if the coin market shrinks or collapses, and how Intel is able to ramp 10nm production after seemingly quarters of issues trying to get that line up and running. Early indications point to enterprise moving steadily towards AMD solutions, Polaris and Vega will cool down and probably not gain much marketshare against NVIDIA, and the transition to 12nm will provide AMD competitive products against the current known roadmap that Intel has released.