AMD has the best chance of competing against Intel, but patience still needs to be in the server CPU market

AMD’s revenue rose 40.4 percent to $1.79 billion in the first quarter of 2020, while net income more than doubled to $162 million, or 9.1 percent of revenue, according to amds. AMD has been putting increasing pressure on Intel in recent years to eat into the market share of its server CPUs, but Intel’s long-standing success is hard to shake. More interestingly, AMD and Intel are hoping to shake up Nvidia’s position in the server GPU market, an epic battle.

AMD has the best chance of competing against Intel, but patience still needs to be in the server CPU market

In recent weeks, we’ve said many times that recessions will accelerate technological transformation, but they’re not fundamental to it. Now we can see the competition between AMD and Intel for server CPUs and GPUs, and amd and Intel’s desire to change Nvidia’s dominance in the data center GPU market.

To put it bluntly, AMD has never had a better chance of trying to compete with Intel in the CPU market and trying to weaken the attack it will bring to the GPU market. But recessions, both in the long and short term, make enforcement more difficult because the process can be volatile.

Surprisingly, AMD in general was unable to take more shipments and shares from Intel, especially as the company reviewed its server spending plans in the first quarter and forecasts for 2020 and 2021. This allows Intel to stay in touch with customers to ensure that the data center division performs at a new high, as Intel did in the first quarter of this year, and has repeatedly occurred in the second half of 2019.

As far as we know, hyperscale service providers have made rigorous adjustments to Xeon processor designs, and cloud service providers are concerned that users may experience similar problems when they bring their applications to run on the cloud they sell. We don’t think that’s the case, but we’re not sure. It’s hard to believe that server buyers still doubt AMD’s commitment to server CPUs and server GPUs. AMD has been involved for a long time and has billions of supercomputing capabilities to prove it.

AMD has been growing its data center business for the past three years since the launch of the “Naples” Epyc 7001 processor in June 2017 and the “Rome” Epyc 7002 processor in August 2019, but Intel’s long-standing success has been unmatched.

After Sledge Hammer opterons in 2003 and 2004, AMD has been putting pressure on Intel in a way that has never been seen before. Rome Epycs can play alongside “Cascade Lake” Intel Xeon SP. In terms of core and total throughput, AMD more than doubled the number of cores, increased the number of memory controllers by a third, and the PCI-Express I/O channel was 60 percent faster. At the same time, AMD has a price advantage.

That’s all we can think of. Intel’s 14nm technology for making Cascade Lake and Cascade Lake-R chips must be very mature and highly productive, with price and cost advantages. It also means that the 14nm production of the “Broadwell” Xeon E5/E7 and “Skylake” Xeon SP chips a few years ago is not as good as many of us think, because today’s sharp price cuts still maintain profit margins.

We don’t know how much the price will be, but if Intel can make up for the huge performance and price/performance differences at AMD Epycs, the price cut is bound to be very steep. If so, we put forward a strange idea in the article. If that does happen, we can assure you that this is not Intel’s planned strategy, but an inevitable strategy, as it has been working for years to mass-produce 10nm manufacturing technology. It runs counter to what Intel has believed in since its inception.

Obviously, Intel has taken an absolute advantage in very bad circumstances. To be honest, it’s amazing.

In this case, it is not clear how much AMD Epyc can reduce the price range because the company has reduced costs by switching to small chips and has proposed aggressive pricing to gain market share today. It uses TSMC’s advanced 7nm technology for small chips in the core of Rome, but still uses a very mature 14nm process from Globalfoundries to manufacture small chips of I/O and memory hubs in the Roman packaging center.

Small chip packages also increase costs, but the net cost of processor sockets is lower than for a single-chip design. If this chip package is of a certain size, it will become easier and cheaper over time, giving AMD greater computing power and higher profit margins in the field.

AMD has the best chance of competing against Intel, but patience still needs to be in the server CPU market

AMD’s revenue rose 40.4 percent to $1.79 billion in the quarter ended March, and net profit more than doubled to $162 million, or 9.1 percent of revenue. The company ended the quarter with $1.39 billion in cash and investments at the end of the third quarter.

Sales in the computing and graphics division rose 73 percent to $1.44 billion, while revenue jumped 16.4-fold to $262 million. Obviously, the client CPU and GPU business is performing well. Sales of game consoles have been a concern and must begin to increase as soon as possible to meet demand for the holiday shopping season.

Depending on the extent of the recession, the new console selling may be a big selling point for household spending, but it’s hard to say. In the short term, sales of the corporate, embedded and semi-customized console chip business fell 21.1 percent to $348 million in the first quarter of 2020, when the division reported an operating loss of $26 million, compared with $68 million a year earlier.

AMD earned $60 million in one-time licensing revenue in the same period last year, so the profit was lower than expected, meaning the core business’s operating losses were not as bad as they looked.

AMD CEO Lisa Su talked about the quarter’s data on a conference call with Wall Street analysts that data center sales, known as Epyc CPU and Radeon Instinct GPU sales, contributed to the company’s overall sales. Therefore, we will set this at 18.5 per cent, compared with about 15 per cent in previous quarters.

AMD has the best chance of competing against Intel, but patience still needs to be in the server CPU market

Based on the streamlined data, past trends and forecasts, we think epyc CPU generated $252 million in sales in the first quarter, up 2.2 times from the same period last year and 9% from the previous quarter. We believe that data center GPU sales, including Radeon and Radeon Instinct cards manufactured in the computing and graphics divisions, reached $78 million in the first quarter, up a bit from the fourth quarter of 2019.

Su said on a conference call that server shipments grew by double digits from the fourth quarter of 2019, more than three times a year earlier, meaning AMD has the upper hand in price/performance to take a share of Intel.

“We’re driving growth from cloud, business, and HPC customers. “

“We saw the power of cloud providers, introduced new instances and accelerated current deployments,” Explains Su. Microsoft Azure, Google and IBM have all announced new products powered by second-generation Epyc processors, notably Google,”” which has released several common VMs and Microsoft’s all-AMD virtual desktop offerings, including the Radeon Instinct GPU. Several cloud providers have accelerated their infrastructure deployments to meet the growing demands of an increasing number of users working and studying from home. For example, one of our large cloud customers was able to deploy 10,000 second-generation Epyc servers in less than ten days to support the surge in demand for collaboration services. “

The impact of large sales on hyperscale producers and cloud builders is a decline in average sales prices. And we think there are some price cuts even for corporate customers. The good news is that AMD’s Epyc server chip’s SMB (Server Message Block, a protocol name) business isn’t big, and the Radeon Instinct GPU accelerator’s SMB business isn’t, so the companies most adversely affected by the coronavirus pandemic won’t buy these data center products anyway.

In another interesting twist, Su added that the growth in Epyc CPU sales by hyperscale producers and cloud computing builders was largely for internal use rather than building a public cloud. Su said sales of HPC workloads were particularly strong across the enterprise.

Many of us are impatient with the immediate and quick adoption of the Epyc CPU because it shows that competition is extremely intense. But from 2003 to 2008, this was not opteron’s confrontation with Xeon. That may not seem like this, but Intel is not as vulnerable as it was then. Su is advising people to expect stable and positive returns, not substantial gains.

“From my point of view, this is very similar to the growth we’ve seen in the PC business. Su explains: “As the server evolves, the server has been running steadily. Each quarter, there are more qualified platforms, and they all increase. So, in relation to our expectations, it is actually going well. Because it’s about cloud acceleration, I think we’re happy with that. We don’t have the numbers ready for growth at the moment. I think we’ve got very aggressive growth assumptions. “

AMD may be patient, but that doesn’t mean the rest of us will be. AMD is still expected to launch the “Milan” Epyc 7003 processor with the Zen3 core by the end of 2020, and at the same time will launch the new RDNA GPU (hopefully with the Radeon Instinct version).

We’ll soon see what the third round of battle brings, but for at least the other two quarters, we’re still in the second round, it’s a very complicated battle, it’s complicated, it’s an epic battle.