Aren't AMD's resources being stretched too thin?

[Jumper]

I was thinking about this for longer time. AMD is investing in all of the right things. Just to mention few MI400, ZEN5, RDNA4. These are all great products but AMD's revenue right now isn't growing so fast and they will need the best engineers to pull of these things. Nvidia for example has more resources and they seem to be focusing on less complex (mostly monolithic) solutions. Also there are rumors from MLID and others that suggest RDNA4 being similar to RDNA1 in having only mid to low-end offerings.

So my question is. Does AMD have enough money and people to invest in all of these products at the same time? If not do you think they will prioritize server and AI over client GPU's and CPU's. They still seem to be lot more efficient with their resources than Intel for example. But it really shows that they're saving lots of money on their marketing for example.

 

[Daniel Nenni]

I think AMD getting rid of their fabs was a key pivot. Had they not done that I don't believe they would be in business much less an industry leader. The same can be said about AMD's partnership with TSMC. That was a HUGE pivot and is the foundation of their current success. I also believe the acquisition of Xilinx was an important part of AMD's success with the TSMC partnership. Xilinx had one of the best foundry groups and a very close relationship with TSMC. Xilinx and TSMC co-developed CoWaS, right?

 

[Jumper]

I think AMD getting rid of their fabs was a key pivot. Had they not done that I don't believe they would be in business much less an industry leader. The same can be said about AMD's partnership with TSMC. That was a HUGE pivot and is the foundation of their current success. I also believe the acquisition of Xilinx was an important part of AMD's success with the TSMC partnership. Xilinx had one of the best foundry groups and a very close relationship with TSMC. Xilinx and TSMC co-developed CoWaS, right?

That's all great points Daniel. But the question still stands. Given all of the investments and capital - personal or financial needed to acquire those goals isn't AMD being stretched too thin at this moment. As an example I give the marketing which I think is underinvested and therefore I think Nvidia, Intel can just "outPR" them in some areas. In terms of software AMD's still lacking behind Nvidia.

 

[delong.height]

That's all great points Daniel. But the question still stands. Given all of the investments and capital - personal or financial needed to acquire those goals isn't AMD being stretched too thin at this moment. As an example I give the marketing which I think is underinvested and therefore I think Nvidia, Intel can just "outPR" them in some areas. In terms of software AMD's still lacking behind Nvidia.

I think they can still sustain this pace for a few more quarter. The problem is if Intel does execute, which is still a question until the products are officially out of the door. How can AMD pivot to its customers when Intel can launch a full-scale price war against AMD? I think I might overestimate this problem when the market is so big. But Pat G has already stated that once his process and foundry are back on track, he'll be doing any of these three things:

1. Find more foundry customers
2. Compete more aggressively with AMD by lowering its prices.
3. Pull products from external sources to IFS

Consider the fact that AMD and Intel are competing in multiple fronts, consumer CPU, data center CPU, various GPU segments, FPGA (Agilex and Xilinx), accelerators, a strong Intel that has steady backbone (IFS) does present many challenges to AMD. Are they going to give up on low to medium range to focus only on high end in regard to that? I see things happening already with ZEN x3d series that can really compete with Intel and win in some benchmarks at same price point.

And if Intel does execute (which is a probability guess, but highly likely at this point, 1 year away), will AMD negotiate with TSMC to become the first wave (TSMC's lead customer, like Apple) to embrace N2 in 2025 or are they doing so in 2026?

As you can see, AMD really cannot compete with Intel because they are not the same size (they are in terms of market share, and AMD can beat Intel in that). Despite both have full category of chips (CPU, GPU, FPGA, ASIC), Intel is far more vertically integrated.

AMD really have to find a way out of this, but it won't be easy. And I'm sure, with Pat G in charge, Intel is going to do whatever they can to replicate or come up with a better way than AMD's

 

[Daniel Nenni]

That's all great points Daniel. But the question still stands. Given all of the investments and capital - personal or financial needed to acquire those goals isn't AMD being stretched too thin at this moment. As an example I give the marketing which I think is underinvested and therefore I think Nvidia, Intel can just "outPR" them in some areas. In terms of software AMD's still lacking behind Nvidia.

Samsung and Intel can both outspend TSMC, right? The same thing with Arm vs RISV-V.

AMD will always have the "Not Intel" market, people who want an alternative just as Samsung has the "Not TSMC" business. Today, however, AMD has a much larger following due to technology and Lisa Su's leadership. It really is like a religion and not going to be easy to erase unless AMD does something really stupid.

I grew up with Intel and AMD. It has been a legendary competition. I have not seen AMD this strong in a very long time if ever. I do not think money is everything in this case. I think AMD technology is compelling and the TSMC/AMD partnership is very strong. TSMC is also in the strongest position I have ever seen. The new fabs in Japan and Germany show great promise/position for TSMC.

 

[hist78]

I was thinking about this for longer time. AMD is investing in all of the right things. Just to mention few MI400, ZEN5, RDNA4. These are all great products but AMD's revenue right now isn't growing so fast and they will need the best engineers to pull of these things. Nvidia for example has more resources and they seem to be focusing on less complex (mostly monolithic) solutions. Also there are rumors from MLID and others that suggest RDNA4 being similar to RDNA1 in having only mid to low-end offerings.

So my question is. Does AMD have enough money and people to invest in all of these products at the same time? If not do you think they will prioritize server and AI over client GPU's and CPU's. They still seem to be lot more efficient with their resources than Intel for example. But it really shows that they're saving lots of money on their marketing for example.

In terms of revenue, AMD has grown to the size probably a lot of people haven't paid attention to.

For the first half of 2023, AMD revenue is $10.71 billion vs Intel's $24.67 billion. That means AMD's revenue is Intel's 43.41%. Looking back:

Revenue - AMD vs Intel, %

2022 37.43%
2021 20.79%
2020 12.54%
2019 9.35%
2018 9.14%
2017 8.37%

I don't think most people (including me) can foresee AMD would grow so fast to near half the size of Intel. My feeling is that AMD is very focused, strategical, and disciplined. They have the patience to grow their business in long term without scarifying profit and financial health.

 

[count]

Not having fabs saves AMD $20b a year in capex and a lot of R&D around process development.

As was pointed out AMD is about 40% of Intel revenue. They also have higher gross margins. Surely they can keep up with Intel in the 4-5 areas they compete? Desktop, Laptop, Server, GPU, FPGA.

Intel has all kinds of divisions like IOT, Mobileeye, IDM 2.0, ect

To me it looks like Intel is stretched much thinner trying to compete in different areas with AMD, TSMC, NVidia, Apple, Qualcomm at the same time. And the fact that they are divesting so many divisions is a tacit admission of this.

 

[cliff]

Is it possible that Intel hires lots of dummies?
Is it also possible that marketing doesn't matter anymore? There are so many good bloggers providing quality opinions. Perhaps the focus should be on stronger engineers?

Note: I have no knowledge of their fab nor packaging, just many dummies hired into their design groups.

 

[cliff]

TJ Rogers famously noted that Bob Swan should be fired. He claimed that Intel was trying to compete through diversity. Every incompetent female MSEE that worked for me as volunteers got a job at Intel. Cute female => Intel. No knowledge of how a transistor works. Actually most of the MSEEs I get from the universities don't know how a transistor works, but 65% are trainable. The pretty untrainable ones go to Intel. Perhaps HR should be called the matchmaking department.

 

[blueone]

TJ Rogers famously noted that Bob Swan should be fired. He claimed that Intel was trying to compete through diversity. Every incompetent female MSEE that worked for me as volunteers got a job at Intel. Cute female => Intel. No knowledge of how a transistor works. Actually most of the MSEEs I get from the universities don't know how a transistor works, but 65% are trainable. The pretty untrainable ones go to Intel. Perhaps HR should be called the matchmaking department.

Cliff, when you say "no knowledge of how a transistor works", are you referring to the physics of how a transistor functions, or are you referring to how to apply transistors in a circuit?

 

[cliff]

Both. My buddy calls them "tool jockeys". They believe Cadence invented the electron. They cannot apply FETs in a circuit, even for creation of standard cells (logic). For example, many of the students take 2 semesters of standard cell design. Most don't know that nmos is more efficient pulling signals to ground, while pmos pulls the signal up to the supply. They cannot lay out AND gates properly. They cannot make test benches to test them. I can go on and on. Many universities just want their $60K for 2 years and pass them through.

This past summer I brought on 24 volunteer interns. I had to terminate 10 of them because they didn't care to learn their occupations, even though they paid $60K to learn. Many want the signature, not the knowledge. Terminating volunteers... think about that. On the other hand, we had 14 good ones who picked up quickly.

Most US MSEE universities don't filter. Intel does a good job at filtering. They filter the pretty ones.

 

[TICin852]

Most US MSEE universities don't filter. Intel does a good job at filtering. They filter the pretty ones.

I'm curious to know which universities have MSEEs that are actually competent.

 

[hist78]

Not having fabs saves AMD $20b a year in capex and a lot of R&D around process development.

As was pointed out AMD is about 40% of Intel revenue. They also have higher gross margins. Surely they can keep up with Intel in the 4-5 areas they compete? Desktop, Laptop, Server, GPU, FPGA.

Intel has all kinds of divisions like IOT, Mobileeye, IDM 2.0, ect

To me it looks like Intel is stretched much thinner trying to compete in different areas with AMD, TSMC, NVidia, Apple, Qualcomm at the same time. And the fact that they are divesting so many divisions is a tacit admission of this.

Another major issue with Intel is that its revenue is declining fast. This large and fast revenue shrinkage is not something IFS can fill. Intel must improve its product lineup to sell what people need with attractive value propositions.

 

[cliff]

Sorry TICin, I will withhold that. It is possible that some of the professors are doing a good job at teaching the concept, but the students are only interested in knowing what is on the test. After 20 years of bringing on MSEEs, I still haven't completely figured that out. I will tell you that the newbies are getting worse every year. It makes no sense. Maybe there is too many funny videos out there to distract the yuts. Dunno.

 

[clhende]

Hi cliff, at Semitracks we notice that a lot of people that participate in our training Bootcamps just want to "check the box" like you say. They are usually being forced to do the work to completion by the company they work for. We see a similar patter in our Online Training System as well. A lot of people sign up with good intentions (I suppose) but then just fall away because they don't have the discipline to stick with it. We do have some that go through our self-paced Online Training in detail though. Perhaps that is a good way to pre-screen people before brining them on.

 

[milesgehm]

TJ Rogers famously noted that Bob Swan should be fired. He claimed that Intel was trying to compete through diversity. Every incompetent female MSEE that worked for me as volunteers got a job at Intel. Cute female => Intel. No knowledge of how a transistor works. Actually most of the MSEEs I get from the universities don't know how a transistor works, but 65% are trainable. The pretty untrainable ones go to Intel. Perhaps HR should be called the matchmaking department.

I just watched that 2021 interview where TJ said Swan should be replaced. He mentioned diversity and sustainability (very briefly) as examples of how Swan wasn't focused on the core issues. That is long way from trying to compete through diversity. You are saying Intel has problems because they hire too many pretty stupid women engineers. That is hard to believe.

 

[blueone]

I just watched that 2021 interview where TJ said Swan should be replaced. He mentioned diversity and sustainability (very briefly) as examples of how Swan wasn't focused on the core issues. That is long way from trying to compete through diversity. You are saying Intel has problems because they hire too many pretty stupid women engineers. That is hard to believe.

That wasn't my experience from 10 years ago or so, not even once, but I didn't work in analog groups. Cliff is very demanding, and he also likes to turn up the contrast to eleven.

 

[mdsomuncu]

I think AMD is at big risk of going out of business in client sector. AMD dgpu market share is getting HAMMERED by nVidia. They have massively cut prices of their gpus and they are still not selling. They are now at 12% market share and intel is also chipping on them now in the budget sector.

From the reports you can see that they are also losing market share on CPUs both mobile and desktop. It literally took 2 good Intel cpu generations to kill off the momentum of Zen and turn the tides. Intel is now at around 83%+ market share and this is while being a full node behind. What do you think happen when meteor lake launches whose fabrication density is similar to that of tsmc 3 hp. Given it will also have a great igpu and will actually be available unlike amd laptop apus, I wouldn't be surprised to see Intel at 90% market share at laptops next year.

We can see from people's comments here that it is basically riding the wave of TSMC at this point and their node edge disappears this winter with intel 4.

 

[Xebec]

AMD is absolutely resource limited - they’re still a much smaller company than Intel or Nvidia, despite years of growth. However, they’re definitely pivoting their investments to where it makes the most sense. Back in 2013-2016 era, they chose to starve the GPU division to focus on CPUs, and that’s what resulted in Zen 1 coming out and being greatly successful. It takes 4-6 years to get a new CPU or GPU from idea to production for reference.

That “Radeon division diet” clearly persisted where AMD Polaris (Rx 400, Rx 500 series) had to compete against multiple generations of Nvidia, and essentially stayed entry level (or low midrange. Rx 5700XT was also limited to mid range despite being a true next gen architecture from them. That they’re in the game at all is a testament to their engineers - I don’t have the numbers, but I believe they’re investing like 10% of the engineering in their GPUs that Nvidia is, so the fact that they have something competing in the $700-1000 range at all is impressive.

I do worry they’re a little too focused on bean counting and not going for more growth overall - though you can see they’re taking the server and enterprise market by storm. AMD made a mistake of not second sourcing capacity (they talked about it but did nothing) in the Athlon 64/X2 days (2003-2006) and they capped at like 30% market share for desktop and mobile. When Core 2 came out, they had no options for reducing capacity that weren’t expensive, and also less mindshare than they might have had otherwise.

I think overall AMD is in the strongest position it’s ever been in it’s entire history - through a combination of smart investment, excellent engineering discipline, and some luck too. (Intel mis-stepping on nodes and products, COVID shortages providing opportunity for “sell every GPU at a good price” helped but weren’t the sole reason for AMD’s rise).

 

[cliff]

That is long way from trying to compete through diversity.

Watch it again. Listen carefully. He said they were competent under tough leaders like Grove, then became soft under Swan. Others tell me they went woke. They were no nonsense under Grove.