The Rise of AMD
How One Woman Changed The CPU Industry
Just a few short years ago, the once great semiconductor company, Advanced Micro Devices (AMD) was on the verge of bankruptcy.
The company had a sustained net loss of $7 Billion and it was downgraded by just about every financial analyst. It was deemed ‘un-investable’ and its share price fell proportionally to this sentiment to under $2 a share by 2015.
To quote Stacy Rasgon of Bernstein Research in 2012:
“We had held out hope that some sort of upside might be forthcoming; however after the report we have no further confidence that any aspect of our prior structural thesis (margin accretion, cash flow, and balance sheet deleveraging) will play out in the foreseeable future”
Today, AMD has a share price of over $90. Its share price has risen almost 5000% in just 5 years and it has cemented itself as one of the most successful CPU corporations in the world taking large market share from titans like Intel.
Before we dive into what led AMD to its current success, we will look at its rocky history, by analyzing management decisions, financials and the competitive environment that led AMD to its near demise.
The Call to Adventure
In the early 2000’s AMD was a formidable semiconductor corporation. Their CEO, Jerry Sanders, led the path to $1 Billion profits and their Athlon chips were named CPU of the Year in 1999. The company seemed like it could not fail.
AMD began as a second-source supplier for companies using Intel processors. Larger computing corporations did not want to solely rely on Intel for computer components, so they licensed AMD to produce versions of various Intel processors.
While these CPUs were manufactured by AMD, most of their designs came from Intel.
Eventually, Intel began viewing AMD as a potential threat and they stopped giving AMD access to its designs.
The time began for AMD to create their own CPUs and they did so, first, by reverse-engineering existing Intel products.
As you can imagine, this method did not sell well as the complexity and time it took AMD to reverse-engineer the chips would make them largely noncompetitive when the time came for them to be sold.
As an example, Intel’s original Pentium chip had 3.1 million transistors in 1993 but within 2yrs the successor Pentium Pro chip had already grown to 5.5 million transistors.
AMD released their first original CPU called the K5 in 1996, that matched the performance of the Pentium of 1993.
By 1999, they had purchased NexGen, an incredible up-and-comer in the microprocessor space, and together AMD and NexGen produced the remarkable K7 Chip.
AMD’s net sales skyrocketed from $2.5 billion in 1998 to $4.6 billion by 2000, and the company rented out the San Jose arena to celebrate the success.
Within the next few years, their chips sold very well. Their new K8 Chip had a great desktop compatible architecture and AMD had control of over 25% of the server market.
AMD continued to iterate successfully on the K8 chip for a few years, performing several architecture updates and manufacturing improvements.
The Trials
Despite these successes, growing internal conflict and financial issues brought on by over supply of chips, the unnecessarily expensive construction of their new Dresden Fab 30 manufacturing plant, poor margins and looming debt began a difficult road ahead for AMD.
AMD would began to spend $100s of millions funding various plants.
“We currently estimate that the construction and facilitation costs of Fab 30 will be $2.6 billion when it is fully equipped by the end of 2005. As of December 29, 2002, we had invested $2.1 billion in AMD Saxony.”
These crazy investments relied on AMD’s ability to sell chips which were mostly dependent on the failure of Intel to stay competitive in the server market. This all changed when the giant, Intel, finally woke up.
Intel gathered a small but brilliant team in Israel to create the chip known as the Pentium M. After ramping up clock speed, adding a second core and adapting it for server compatibly, the chip was a success!
The chip was faster and more power-efficient than AMD’s. It made Intel gain rapid market share in the server market and it took back the desktop market. AMD could not compete.
Intel established an incredible rhythm with their new line. Each year they would slightly modify their existing lineup with slightly faster, smaller, more energy efficient processors.
While AMD’s CPUs continued to improve, they were eventually shut out of the high-end market and forced to compete on price, further growing the existing issues.
Financial struggles further exacerbated internal issues. When they acquired a company named ATI, a GPU company, for $2.5 Billion, an internal feud began.
Differences in company culture, outlook and a great CPU vs GPU divide ensued. The employees of one side of the divide would prioritize their needs over others leading to shortcomings in product quality which led to extensive delays.
AMD’s 2007 and 2011 processor releases went very poorly. They delivered late and didn’t meet performance expectations.
Intel dominated the mid-2000s with great execution, and large market share. Intel, grew to be over 60 times larger than AMD by market capitalization.
By 2014, all had seemed lost. AMD was drowning in billions of debt they accumulated and to top it all off, the demand for PCs seemed to be slowly diminishing. The share price was heading to $0 fast.
Rebirth — Enter Lisa Su
Lisa Su was appointed president and CEO of AMD in October 2014 after holding a C-Suite position for 2 years prior.
When Lisa entered, everything changed. A business plan was adapted for a long term outlook. She instituted an effective strategy to minimize debt and focused processor design and manufacturing on computer gaming, cloud and data center compatibility.
Who is Lisa Su?
Lisa Su is arguably one of the most impressive CEOs of our time, and certainly in the high performance computing space.
She is a Taiwanese-born American immigrant who received her BS, MS and PhD in Electrical Engineer from the prestigious M.I.T.
Before saving AMD, she worked extensively with semiconductors. She was part of the technical staff at Texas Instruments, the General Manager at Free-scale Semiconductor and led global research and development for semiconductors at IBM.
Of her many notable accolades, she was named Executive of the Year by E.E times, World’s Greatest Leaders by Fortune and 26th Best CEO of 2019 by the Harvard Business Review and she received the highest compensation of any CEO in 2019 of $58.5 Million.
Business Model — How AMD Operates Now
When Lisa Su entered as the CEO, she made several decisive long term changes. These can be summarized as:
- Create competitive central and graphic processors primarily targeted at the computer gaming sector.
- Provide enterprise data and cloud solutions with high performance and embedded processors.
- Reduce debt.
“It’s not what you do today but it’s really what you do year after year, after year that people pay attention to.” — Lisa Su
AMD had a history of trying everything and anything under the sun in the semiconductor space. They acquired many corporations, spent billions on manufacturing and divided the internal culture in the process.
To rectify the mistakes of the past, Lisa Su strictly focused AMD’s efforts on developments they were great at, and to get rid of any areas that were dunce.
“It is really important when you’re a technology company to decide what you are really, really good at because you have to be the best, number one or number two… …It’s all about focusing on, ‘Hey, this is the DNA of the company, let us make it as great as possible in terms of what we can bring to the market.’” — Lisa Su
The strategy, going forward, was to execute on creating robust high performance chips. And that’s what they did.
Today, AMD manufactures chipsets used in desktops and notebooks. They manufacture embedded processors used in products that require high-to-moderate levels of performance as well as System-on-Chip (SoC) products and technology for game consoles.
With this strategy and new product line AMD added $2.5 billion in revenues between 2015 and 2018.
Under Lisa Su’s tutelage, AMD’s total revenue grew from $4.0 billion in 2015 to $6.5 billion in 2018. This growth was largely driven by higher demand for its Ryzen and Radeon product lines.
AMD’s Segments and their Contribution to Total Revenue:
AMD has been gaining share from Intel in the processors market. AMD’s market share gained from 18% in 2016 to 23% in 2018. A staggering growth of 5% in a market that was dominated by Intel within just two years.
In fact, as of late 2019, AMD’s market share stands at over 32%. That is an almost 100% increase from 3yrs ago and a 1/3rd of the entire market.
Looking at the Enterprise segment, revenue grew from $2.2 billion in 2015 to $2.4 billion in 2018.
This can be attributed to the company’s EPYC processors providing high performance, at a relatively lower price to its competitors.
These products have helped AMD make a dent in the market share from Intel and Nvidia, and have helped drive revenue growth of around 50%!!
AMD’s products have proven so efficient that the US Department of Energy chose AMD’s chips to power its supercomputer at the Oak Ridge National Laboratory in Tennessee.
The Road Back — Taking Down Intel
Although Lisa Su played a pivotal role in the recent success of AMD, it stands to mention that Intel has struggled to stay competitive.
Notably, to move from a 14 nanometer chip to the more powerful 10 nanometer chip.
AMD brought a processor equivalent in power to the 10 nm chip to market ahead of Intel, and developed a product architecture that allows it to keep costs low.
“(Intel’s) lack of execution and AMD’s flawless execution has yet to play out in terms of the market share AMD is going to gain in the coming years,” Mosesmann said. “I wouldn’t be surprised if they captured half the data center market.”
The data center and cloud market has become a priority for AMD. As companies transition to the cloud and 5G, AMD has a product line competitive take the incoming demand from Intel.
More recently, Intel announced a 6 month delay in 7 nanometer chips for their product line. Chips with greater transistor density, translate to improved performance and efficiency. With AMD, and competitor TSMC already selling 7nm products, this will put Intel behind 1-2yrs for server and laptop chip technology. This announcement led to a stark decline of 16% in share price in a single day.
Intel is struggling to stay competitive and AMD is playing with all time stock market highs, growing market demand and sales to match.
Here is great video by LinusTechTips that breaks down the topic well:
Among the Stars — Looking Ahead
If history can tell us anything between these two competitors, it is that Intel can wake back up and take back lost market share.
Thankfully, AMD has released a product roadmap and future financial targets at this years financial analyst day (F.A.D).
Lisa announced that the strategy over the next five years is a commitment to high performance computing and to bring disruptive solutions that can combine the company’s CPU and GPU architectures.
With growing demand, AMD is still prioritizing data centers and communications with a $35 billion market size.
PCs have a strong $32 billion market size and consumer graphics and game consoles have a $12 billion market size.
AMD will also reach the 5G services market as these will require processors to handle extensive data processing.
On GPUs, AMD already introduced its RDNA graphics architecture. This will last 5 years and will be used for both consoles and PCs. RDNA architecture was released with 7nm last year as the Navi chip.
AMD will divide the GPU roadmap by adding a separate path for computing products while maintaining the Radeon graphics roadmap for consoles and PCs.
AMD expects to ship the next 7nm RDNA 2 graphics chip by the end of 2020 that will include advanced capabilities such as 4K performance, ray tracing, and variable rate shading.
AMD also plans to continue to improve performance and battery life for its Ryzen processor family. With notebooks making up 64% of the market, AMD is making progress on battery life with up to an 18-hour battery life.
The CPU roadmap is on track to ship 7nm Zen 3 cores in the “Milan” server processor late this year.
The 5nm Zen 4-based “Genoa” CPU is part of the El Capitan Exascale supercomputer that will start production in late 2021.
The company estimates that the data center GPU market has an $11 billion untapped market size, which includes AI, and cloud gaming.
In terms of projected financials, AMD is optimistic. They updated their forecast with a new 20% compounded annual growth rate. The company expects significantly improved gross margins, operating margins, and improved free cash flow.
AMD will look to improve it’s exposure to software as this has been somewhat of a weak point and its competitors have solid products. They are relying on an open source strategy, offering an alternative to NVIDIA’s CUDA and Intel’s OneAPI. However, they are already making headways on that front with machine learning frameworks Pytorch and TensorFlow now being compatible with Radeon Instinct GPUs.
Overall, the semiconductor space is tilted in AMD’s favor for the first time in a long time, now it is just a matter of execution. With Lisa Su at the realm, this is more possible than ever!
Conclusion
Lisa Su saved AMD!
Thank you for reading! Please give me many claps for Lisa Su bae :)