If AMD continues like this, it could really, I mean REALLY, challenge Intel as a company in 2-3 years time. Not just in a market with a product, but as a company.

This is also the first time I see AMD becoming much more optimistic about it's future income. Probably they managed to secure much more wafers from TSMC and also know that Intel wouldn't be able to really respond to Ryzen and EPYC in the next 12-18 months.

I don't quite know how this part of the industry reflects the more directly consumer-oriented parts, but higher Q1 than Q4 revenue sounds really surprising to me. Pretty much underscores that AMD is selling every single thing they are able to produce. As @john_ said above, this could lead to a very interesting competitive positioning if it keeps up. Let's hope they feed a lot of those profits back into R&D and don't just pay them out to shareholders.

ValantarLet's hope they feed a lot of those profits back into R&D and don't just pay them out to shareholders.

THIS !!!

z1n0xTHIS !!!

How is this going to happen without dividends? AMD had the last five years on lockdown and they've added the next five.

AleksandarKRevenue was $3.45 billion, up 93 percent

There goes "NV is 80%+ of the GPU market" bazinga.

thesmokingmanHow is this going to happen without dividends?

Most US companies tend not to pay dividends, you get value from company gaining in value.

medi01There goes "NV is 80%+ of the GPU market" bazinga.

How so? They haven't published anything on the revenue split within the Computing and Graphics segment, so it's impossible to tell how the $2.1B revenue of that segment divides between its various parts (laptop CPUs, desktop CPUs (DIY/OEM), laptop GPUs, desktop GPUs (DIY/OEM)). They say RX 6000 revenue has doubled since Q4 2020, but we don't know any precise numbers.

And to think that just 5 years ago AMD didn't look like they could avoid bankruptcy to several Financial Analysts. No way to repay their mountainous debt. Losses in the hundreds of millions yearly. Their stock trading for under $2 a share. Today it trades for $85 a share.

Amazing.

john_This is also the first time I see AMD becoming much more optimistic about it's future income. Probably they managed to secure much more wafers from TSMC and also know that Intel wouldn't be able to really respond to Ryzen and EPYC in the next 12-18 months.

I think, looking at it from TSMC perspective, AMD is a monster consumer.
Huge GPUs, huge APUs for tens of millions of consoles (interestingly, CPUs are laughably small, compared to those, which raises "why on earth do they cost so much then").

So I would not be surprised if they get preferential treatment.

ValantarLet's hope they feed a lot of those profits back into R&D and don't just pay them out to shareholders.

Welp, let's recap:
CPUs - trounced Intel on all fronts, the only reason Intel is still going so strong, are closed clubs called OEMs
GPUs - RDNA2, just one year after RDNA2, topples NV's pricier, power hungrier GPU equipped with faster memory in the 7 most recent games (computerbase)



JUST one year after RDNA1, when expectations were AMD would merely wrestle with 2080Ti (which even smaller AMD chip appears capable of).
And I mean, not only fanboi expectations, but look at mem configs and pricing on NV side, Huang clearly hasn't expected it.

ValantarAs @john_ said above, this could lead to a very interesting competitive positioning if it keeps up

If it keeps up, I don't see a place for GPU only or CPU only companies.
Whether it would be Intel making its GPU branch much stronger, of green-blue merger, I think in a couple of years they'll feel the heat.

ValantarHow so? They haven't published anything on the revenue split within the Computing and Graphics segment, so it's impossible to tell how the $2.1B revenue of that segment divides between its various parts (laptop CPUs, desktop CPUs (DIY/OEM), laptop GPUs, desktop GPUs (DIY/OEM)).

You mean "despite AMD rolling out pricey high end GPUs that they sell like hotcakes, maybe revenue jump is largely caused by CPUs"?
Welp, maybe. Although, I'd dare say, unlikely.

R&D :Intel~ 13.5 billion/year; AMD~ $3 billion now....(2017-1.8;2018-2.4;2019-2.9)

amd will need a decade to reach intel levels only if they manage to have better products....and is a big if as intel is working hard to fix their production issues ...

medi01Welp, let's recap:
CPUs - trounced Intel on all fronts, the only reason Intel is still going so strong, are closed clubs called OEMs
GPUs - RDNA2, just one year after RDNA2, topples NV's pricier, power hungrier GPU equipped with faster memory in the 7 most recent games (computerbase)



JUST one year after RDNA1, when expectations were AMD would merely wrestle with 2080Ti (which even smaller AMD chip appears capable of).
And I mean, not only fanboi expectations, but look at mem configs and pricing on NV side, Huang clearly hasn't expected it.

Previous R&D expenditures don't tell us anything about current and future R&D expenditures. And many, many companies throughout history have become lazy and lost their edge through gaining competitive advantage - just look at Intel.

medi01If it keeps up, I don't see a place for GPU only or CPU only companies.
Whether it would be Intel making its GPU branch much stronger, of green-blue merger, I think in a couple of years they'll feel the heat.

An Intel+Nvidia merger would never pass regulatory approval. Period.

medi01You mean "despite AMD rolling out pricey high end GPUs that they sell like hotcakes, maybe revenue jump is largely caused by CPUs"?
Welp, maybe. Although, I'd dare say, unlikely.

The "revenue jump" you describe didn't start this quarter, but rather in Q3 2020. That was before the launch of RDNA2, and directly after the launch of Zen3. So there's a pretty strong corellation there. There has of course been a notable increase since then, but nowhere near the jump that came in the quarter containing the Zen3 launch. Hence me questioning the revenue split. In parallel to the GPUs, AMD has also launched a laptop lineup that by their own words has ramped in sales far faster than Renoir. I know they're selling every RDNA2 GPU they can make. But we have no idea how that compares to the other sales in this segment, other than indications that the segment is still dominated by CPU sales given that the revenue jump mainly started before the GPUs launched. Please stop making poorly founded assumptions.

There is a way to see if AMD is throwing enough cash in R&D for CPUs and GPUs. In the past AMD's products where in a type of Tick-Tock schedule where you either getting from the newest product better efficiency or better performance but rarely both. Usually it was better performance because AMD needed to factory overclock it's products to remain competitive with Intel's and Nvidia's products. These last years this had changed. We get products that offer much better performance and efficiency as a package. Not necessarily thanks to a new manufacturing node. We saw that with RDNA2. Better efficiency AND performance at the same node. AMD is throwing much more money and manpower on it's CPUs and GPUs compared to the past.

64KAnd to think that just 5 years ago AMD didn't look like they could avoid bankruptcy to several Financial Analysts. No way to repay their mountainous debt. Losses in the hundreds of millions yearly. Their stock trading for under $2 a share. Today it trades for $85 a share.

Amazing.

In my opinion, Ryzen offered very competitive performance from the abysmal Bulldozer. However the main contributor to their success is all thanks to the sleeping, slow and complacent Intel that offered them a window of opportunity to shine. As Intel never managed to offer any competitive or exciting products subsequently, AMD's popularity continues to grow with each passing generation or Ryzen. If Intel had released their 10nm processor with a Cypress Cove on schedule, it would have opened a fairly wide performance gap when comparing with early Ryzen chips. But because Intel failed to do so, and coupled with AMD's aggressive release of higher core count processors at a lower price point which became the undoing of Intel. Now that Intel has effectively lost their fab advantage, it is going to be an uphill battle for them going forward leaning purely on chip architecture to try and fend off competition.

john_There is a way to see if AMD is throwing enough cash in R&D for CPUs and GPUs. In the past AMD's products where in a type of Tick-Tock schedule where you either getting from the newest product better efficiency or better performance but rarely both. Usually it was better performance because AMD needed to factory overclock it's products to remain competitive with Intel's and Nvidia's products. These last years this had changed. We get products that offer much better performance and efficiency as a package. Not necessarily thanks to a new manufacturing node. We saw that with RDNA2. Better efficiency AND performance at the same node. AMD is throwing much more money and manpower on it's CPUs and GPUs compared to the past.

I am sure they will reinvest their money to R&D. Historically being a significantly smaller CPU and GPU company as compared to Intel and Nvidia, they had little to spend on R&D especially with the losses they are incurring each quarter. I think kudos to Lisa to navigate the firm out of bankruptcy over the last few years. The strategy to put GPU design as secondary then and focus most of their resources on CPU development is paying off. With CPU pulling them out of the red and into a healthy profit, they are now reinvesting money into GPU design to also compete with Nvidia, evident with the release of RDNA2 that allowed them to compete at the high end graphic solution for the first time in many years.

ValantarPrevious R&D expenditures don't tell us anything about current and future R&D expenditures. And many, many companies throughout history have become lazy and lost their edge through gaining competitive advantage - just look at Intel.

If anything, AMD is a living example of "you can't measure innovation with money" (SpaceX vs BlueOrigin is another)
CEO has strategy in mind, they were hiring in the previous years, they are still FAR FAR behind competiotrs on R&D spending (largely offset by Intel maintaining superexpensive Fabs, that must also be said, but even with that point in mind).

Why would they suddenly sharply drop R&D spending? The idea feels weird.

ValantarAn Intel+Nvidia merger would never pass regulatory approval. Period.

I don't recall any specially strict regulatory approval requirements for one industry + other industry merger.
If you'd said that about AMD merging with Intel, I'd agree.

ValantarThe "revenue jump" you describe didn't start this quarter, but rather in Q3 2020. That was before the launch of RDNA2, and directly after the launch of Zen3. So there's a pretty strong corellation there.

Fair enough, however:

1) isn't Q4 the strongest quarter? Yet Q1 next year beats it.
2) I simply cannot see how one could start pumping expensive products out and selling them as hotcakes, without boosting graphics department earnings substantially. AMD used to sell mid range card at best. Of course it's an assumption, not a fact.

watzupkenIn my opinion, Ryzen offered very competitive performance from the abysmal Bulldozer. However the main contributor to their success is all thanks to the sleeping, slow and complacent Intel that offered them a window of opportunity to shine. As Intel never managed to offer any competitive or exciting products subsequently, AMD's popularity continues to grow with each passing generation or Ryzen. If Intel had released their 10nm processor with a Cypress Cove on schedule, it would have opened a fairly wide performance gap when comparing with early Ryzen chips. But because Intel failed to do so, and coupled with AMD's aggressive release of higher core count processors at a lower price point which became the undoing of Intel. Now that Intel has effectively lost their fab advantage, it is going to be an uphill battle for them going forward leaning purely on chip architecture to try and fend off competition.


I am sure they will reinvest their money to R&D. Historically being a significantly smaller CPU and GPU company as compared to Intel and Nvidia, they had little to spend on R&D especially with the losses they are incurring each quarter. I think kudos to Lisa to navigate the firm out of bankruptcy over the last few years. The strategy to put GPU design as secondary then and focus most of their resources on CPU development is paying off. With CPU pulling them out of the red and into a healthy profit, they are now reinvesting money into GPU design to also compete with Nvidia, evident with the release of RDNA2 that allowed them to compete at the high end graphic solution for the first time in many years.

The fact that they can rise from ashes, battling two Eastern Front ( CPU) and Pacific Theaters ( GPU) with resources equivalent for just musket and mortar is nothing sort of amazing. I hope they always strive their momentum, never stop innovating, or they will have a hard time finding it again.

medi01If anything, AMD is a living example of "you can't measure innovation with money" (SpaceX vs BlueOrigin is another)
CEO has strategy in mind, they were hiring in the previous years, they are still FAR FAR behind competiotrs on R&D spending (largely offset by Intel maintaining superexpensive Fabs, that must also be said, but even with that point in mind).

Why would they suddenly sharply drop R&D spending? The idea feels weird.

I never said they did. I just said I hope they keep reinvesting their profits into R&D. As a company grows more successful and more profitable, investors are much more likely to start demanding greater returns on their investments. Which is part of why so many large corporations turn into midless drone husks producing nothing of value and losing everything that made them big.

medi01I don't recall any specially strict regulatory approval requirements for one industry + other industry merger.
If you'd said that about AMD merging with Intel, I'd agree.

They are both major players in compute in general. Nvidia competes with Intel in portions of the HPC and datacenter markets, as well as Intel moving into GPU-based compute accelerators. But more than that, they're both in the general computing space, and hold massive market share leads. Two such companies merging to create a monolith with massive market share advantages across the major components for an entire industry is unlikely to be accepted even if they weren't direct competitors, as the merger would open the door for a type of bundle sales that would be anticompetitive across all relevant parts of the industry. It's the same reason Nvidia's purchase of ARM isn't likely to go through.

medi01Fair enough, however:

1) isn't Q4 the strongest quarter? Yet Q1 next year beats it.
2) I simply cannot see how one could start pumping expensive products out and selling them as hotcakes, without boosting graphics department earnings substantially. AMD used to sell mid range card at best. Of course it's an assumption, not a fact.

None of that is wrong, and it is indeed highly unusual for a Q1 to be a record high (and frankly it's unusual for a Q1 to not be significantly lower than Q4). But their CPUs are still in relatively short supply too, so there's little indication of sales of those dropping off, even if supply does seem to be improving somewhat. And no doubt GPU sales, and especially GPU ASPs, are increasing markedly. But it's impossible to tell from these numbers whether this is sufficient to represent a major part of these numbers. We do know AMD has up until this time sold drastically more CPUs than GPUs, and that the RX 6000 series is still extremely rare on for example Steam Hardware Surveys. That of course doesn't mean they aren't producing and selling significant numbers of them, and many might be going to other uses such as cryptomining. But we can't tell, so the safe assumption is to think that GPUs represent a relatively minor increase on top of the already massive one from Zen3 CPU sales. I'd be very happy to see AMD GPUs gain more traction with consumers, but so far there isn't much evidence of that.

watzupkenIn my opinion, Ryzen offered very competitive performance from the abysmal Bulldozer. However the main contributor to their success is all thanks to the sleeping, slow and complacent Intel that offered them a window of opportunity to shine. As Intel never managed to offer any competitive or exciting products subsequently, AMD's popularity continues to grow with each passing generation or Ryzen. If Intel had released their 10nm processor with a Cypress Cove on schedule, it would have opened a fairly wide performance gap when comparing with early Ryzen chips. But because Intel failed to do so, and coupled with AMD's aggressive release of higher core count processors at a lower price point which became the undoing of Intel. Now that Intel has effectively lost their fab advantage, it is going to be an uphill battle for them going forward leaning purely on chip architecture to try and fend off competition.


I am sure they will reinvest their money to R&D. Historically being a significantly smaller CPU and GPU company as compared to Intel and Nvidia, they had little to spend on R&D especially with the losses they are incurring each quarter. I think kudos to Lisa to navigate the firm out of bankruptcy over the last few years. The strategy to put GPU design as secondary then and focus most of their resources on CPU development is paying off. With CPU pulling them out of the red and into a healthy profit, they are now reinvesting money into GPU design to also compete with Nvidia, evident with the release of RDNA2 that allowed them to compete at the high end graphic solution for the first time in many years.

Regardless of Intel's actions or lack there of, what AMD has managed to do while not only competing, but besting Intel who has an R&D budget 685% greater (2020 figures) is amazing, and perhaps one of the most impressive feats in the entirity of the x86 industry.

Let's not forget that AMD is not only competing, but trading blows (besting them in some circumstances) with Nvidia who has an R&D budget 198% greater than AMD's and a 159% larger staff than AMD.... can anyone name another hardware company, or a company from any other industry even, that is not only keeping up with, but exceeding its entrenched competitors who also happen to have staffs and budgets magnitudes larger? It's quite impressive once your factor in the financial reality, which, as anyone knows, generally determines how competitive a company is capable of being. Even if AMD was 10% behind Intel and Nvidia in performance, based on the much smaller financial resources at AMD's disposal, it'd still be a seriously impressive feat. Thats why when anyone tries to diminish AMD's success by pointing out that Intel is still on 14nm or anything of the like, I just refer to the resource and financial gulf that exists between these three companies to demonstrate why AMD deserves as much praise as they receive, if not more.

The Pope should consider anointing Lisa Su as a Saint because she has really performed a miracle.

AMD could well treble their sales by the end of 2024, from the current level, especially after acquiring Xilinx. At this point in time, assuming the merger is complete by the end of the year, they are by far the most well positioned in the computing space after ARM. Did I mention Nvidia's acquisition is unlikely to go through :D

64KAnd to think that just 5 years ago AMD didn't look like they could avoid bankruptcy to several Financial Analysts. No way to repay their mountainous debt. Losses in the hundreds of millions yearly. Their stock trading for under $2 a share. Today it trades for $85 a share.

Amazing.

Just needed the right kind of leadership with Lisa Su and engineering team.

medi01I think, looking at it from TSMC perspective, AMD is a monster consumer.
Huge GPUs, huge APUs for tens of millions of consoles (interestingly, CPUs are laughably small, compared to those, which raises "why on earth do they cost so much then").

So I would not be surprised if they get preferential treatment.




Welp, let's recap:
CPUs - trounced Intel on all fronts, the only reason Intel is still going so strong, are closed clubs called OEMs
GPUs - RDNA2, just one year after RDNA2, topples NV's pricier, power hungrier GPU equipped with faster memory in the 7 most recent games (computerbase)



JUST one year after RDNA1, when expectations were AMD would merely wrestle with 2080Ti (which even smaller AMD chip appears capable of).
And I mean, not only fanboi expectations, but look at mem configs and pricing on NV side, Huang clearly hasn't expected it.



If it keeps up, I don't see a place for GPU only or CPU only companies.
Whether it would be Intel making its GPU branch much stronger, of green-blue merger, I think in a couple of years they'll feel the heat.


You mean "despite AMD rolling out pricey high end GPUs that they sell like hotcakes, maybe revenue jump is largely caused by CPUs"?
Welp, maybe. Although, I'd dare say, unlikely.

Look at the line that says “Gross Margin %”, which increased from 45% to 46%. The prices exceeding MSRP is a supply and demand issue, not necessarily AMD easing prices, but businesses who may be paying more to get product and or abusing prices for their benefit. AMD is selling a lot more product, which has added to their bottom line.

medi01There goes "NV is 80%+ of the GPU market" bazinga.


Most US companies tend not to pay dividends, you get value from company gaining in value.

I know how equities work. That's why I worded my posted the way I did. Clearly the two above me don't know that AMD doesn't pay out dividends. Hell at least know wtf you are talking about before dissing on the company, which judging from the direction of this thread.... yea more of the same bullshit.

ValantarPrevious R&D expenditures don't tell us anything about current and future R&D expenditures. And many, many companies throughout history have become lazy and lost their edge through gaining competitive advantage - just look at Intel.

An Intel+Nvidia merger would never pass regulatory approval. Period.

The "revenue jump" you describe didn't start this quarter, but rather in Q3 2020. That was before the launch of RDNA2, and directly after the launch of Zen3. So there's a pretty strong corellation there. There has of course been a notable increase since then, but nowhere near the jump that came in the quarter containing the Zen3 launch. Hence me questioning the revenue split. In parallel to the GPUs, AMD has also launched a laptop lineup that by their own words has ramped in sales far faster than Renoir. I know they're selling every RDNA2 GPU they can make. But we have no idea how that compares to the other sales in this segment, other than indications that the segment is still dominated by CPU sales given that the revenue jump mainly started before the GPUs launched. Please stop making poorly founded assumptions.

Uh, do you know when Zen 3 launched?! 3rd quarter profit would be strictly from Zen 2 chips, not zen 3 chips. 4th quarter 2020 would include sales of Zen 3 chips.

agentnathan009Uh, do you know when Zen 3 launched?! 3rd quarter profit would be strictly from Zen 2 chips, not zen 3 chips. 4th quarter 2020 would include sales of Zen 3 chips.

Damn, you're right, I completely forgot just how late Zen3 was compared to previous launches. Makes me wonder what exactly changed in Q3, given that there weren't any new product launches then. I guess Renoir adoption had finally ramped, which undoubtedly amounts to something, but it can't be that much overall. Might be the effect of the production ramp for the XSX and PS5 I guess (combines with continued enterprise/server growth), as the Enterprise, Emedded and Semi-custom segment had huge growth that quarter. The Computing and Graphics segment had relatively modest growth in comparison. Ultimately that makes it even more impossible to tell whether the growth mainly comes from CPUs or GPUs, though given die sizes and thus how many are produced per wafer, plus overall availability, it's still safest to assume CPUs are in a clear majority.

agentnathan009Look at the line that says “Gross Margin %”, which increased from 45% to 46%. The prices exceeding MSRP is a supply and demand issue, not necessarily AMD easing prices, but businesses who may be paying more to get product and or abusing prices for their benefit. AMD is selling a lot more product, which has added to their bottom line.

Fair enough, but let us keep in mind that margins are diluted by console business, which, I'd estimate, would be margins much lower than 47% reported.

medi01Fair enough, but let us keep in mind that margins are diluted by console business, which, I'd estimate, would be margins much lower than 47% reported.

AMD did bump prices on 5000 series chips and probably boosted price for enterprise chips which would offset the console margins to some degree. Something their bean counters would know.

ValantarDamn, you're right, I completely forgot just how late Zen3 was compared to previous launches. Makes me wonder what exactly changed in Q3, given that there weren't any new product launches then. I guess Renoir adoption had finally ramped, which undoubtedly amounts to something, but it can't be that much overall. Might be the effect of the production ramp for the XSX and PS5 I guess (combines with continued enterprise/server growth), as the Enterprise, Emedded and Semi-custom segment had huge growth that quarter. The Computing and Graphics segment had relatively modest growth in comparison. Ultimately that makes it even more impossible to tell whether the growth mainly comes from CPUs or GPUs, though given die sizes and thus how many are produced per wafer, plus overall availability, it's still safest to assume CPUs are in a clear majority.

I think the stay at home let’s start working from home and now we need more computers to do so part of the global pandemic known as COVID-19 probably contributed a lot to the sudden and sustained growth.

agentnathan009I think the stay at home let’s start working from home and now we need more computers to do so part of the global pandemic known as COVID-19 probably contributed a lot to the sudden and sustained growth.

That's no doubt true, though I would have expected that to at least start happening in Q2. And as mentioned, the growth in the Computing and Graphics (i.e. all consumer/oem/business desktop+laptop CPUs and GPUs) group didn't grow that much in Q3 2020 - 391M YoY and 300M QoQ, compared to 609M YoY and 569M QoQ for the Enterprise, Embedded and Semi-custom group. So it seems like consoles + EPYC (and possibly Radeon Instinct?) started the growth in Q3, with CPUs and GPUs launched in Q4 maintaining that level and ensuring further growth. I really wish AMD would publish the CPU-GPU split though.