Unpopular opinion: Rocket Lake Is Fine

[fonsui]

I'll intro this with pointing out that I find the "Team Blue vs. Team Red" argument to be unproductive, if I consider myself on any "side" its "Team Consumer". I have a healthy respect and appreciation for the strengths and accomplishments of both Intel and AMD, and I am able to recognize their individual failings.

 

That established, I believe that the discussion surrounding Rocket Lake has been somewhat dishonest/misleading, absolutely not by intent, but in impact. I will explain:

 

Obviously, benchmarks are science and properly executed science has shown that Intel is absolutely behind the ball in per-core performance, and the product stack itself shows how behind Intel is in core count - AM4 Zen processors scale to 16c/32t while Intel is maxing out on mainstream desktop at 10c/20t last gen, and goes down to 8c/16t this generation. We are also within our rights to not love the Rocket Lake pricing, Intel went higher when we were not-so-secretly hoping they'd go lower.

 

That's the bad, now let's look at the "fine" - Intel has landed a product line that is:

 

- Running on a new chipset which allowed a feature refresh and includes some previously "premium" functionality filtered down from the Zx90 tier

- At least in the same ballpark as Ryzen 5000 overall performance at every tier, even if it doesn't trash it like customers have come to expect

- Not completely ridiculous pricewise, and downright competitive in some tiers/use cases (see 600K, all the -F models)

- Using a brand-new memory controller that provides comparable performance and functionality as Ryzen 5000

- Operating PCIe at 4.0, and has an additional x4 link for NVMe in addition to the standard 16 lanes that have previously been available

 

Does it trounce Ryzen, unqualified? Absolutely not. It does, however, put Intel parts in the market that are competitive choices that provide a lot more feature parity with Ryzen 5000 than Comet Lake or prior gens did. This was a parity move to keep them in the game, not a run at the crown. Intel's money in the desktop consumer market is on Alder Lake and Xe HPG, that's what they really want to push hard.

 

Now, if we look at the HEDT space Intel has slipped even further and Threadripper simply dominates, workstation Xeons haven't been keeping up for quite some time now, and the non-workstation Core X-Series CPU (latest being Cascade Lake AKA Yet Another Skylake) and chipset (X299) are so far behind they're barely worth talking about, and I've seen no plans to change that.

 

TL;DR: Rocket Lake seems to be suffering from an expectation management problem more than actually being a "Bad Product". It doesn't leave customers with poor performance or non-working systems or functions, it's not wildly overpriced, and if people can snag a working current-gen CPU in this climate that in and of itself is a win.

[Fasauceome]

5 minutes ago, fonsui said:

Running on a new chipset which allowed a feature refresh and includes some previously "premium" functionality filtered down from the Zx90 tier

This is just them catching up to AMD, who has offered ram and CPU overclocking on non-premium chipsets since ryzen dropped. AMD went on to get boards that incorporate features like thunderbold (when that became an open standard) and other premium aspects when board partners started taking them seriously. There was only really a lull in features at first gen.

 

7 minutes ago, fonsui said:

At least in the same ballpark as Ryzen 5000 overall performance at every tier, even if it doesn't trash it like customers have come to expect

Intel made some big promises for 11th gen. Ultimately it was wishful thinking, leading to no reason for 10th gen owners to upgrade.

 

7 minutes ago, fonsui said:

Not completely ridiculous pricewise, and downright competitive in some tiers/use cases (see 600K, all the -F models)

Availability is a strong point in Intel's favor. We will see if this continues after the pandemic's effects pass, though. 

 

8 minutes ago, fonsui said:

Using a brand-new memory controller that provides comparable performance and functionality as Ryzen 5000

- Operating PCIe at 4.0, and has an additional x4 link for NVMe in addition to the standard 16 lanes that have previously been available

Once again, them playing catch up to AMD. They're new features, but only new to Intel.

 

9 minutes ago, fonsui said:

This was a parity move to keep them in the game, not a run at the crown. Intel's money in the desktop consumer market is on Alder Lake and Xe HPG, that's what they really want to push hard

Intel used the "fastest gaming CPU ever" line in their previous marketing, clearly having the best gaming performance is super important to them, and if they would have been able to get the crown back, they would have absolutely done so.

 

10 minutes ago, fonsui said:

TL;DR: Rocket Lake seems to be suffering from an expectation management problem more than actually being a "Bad Product". It doesn't leave customers with poor performance or non-working systems or functions, it's not wildly overpriced, and if people can snag a working current-gen CPU in this climate that in and of itself is a win.

Bad, good, the product quality is relative. Relative to 10th gen, the product is actually the same. But expectations exist, and I find it pretty reasonable to decry 11th gen's release because it doesn't bring anything of real value to the table. If anything it muddied the water (m.2 slots inactive when using 10th gen CPUs on 500 series boards, weird memory frequency locks on 400 series boards, uncertainty continuing on 10nm, etc.) and it can be seen as a step backwards in some ways because of the difficulty in marketing.

 

I don't think Intel is being unfairly judged. AMD keeps moving forwards, intel keeps moving sideways. Will this last forever? Probably not, so people shouldn't latch on to one brand or the other with their loyalty. But criticizing Intel has looked to be pretty par for the course this generation.

[freeagent]

If I was running an older gen Intel, heck yeah I would consider the upgrade. I left AMD over a decade ago, and I took a real leap of faith by ignoring my gut and not getting an Intel. Lots of guys wouldn't consider an AMD no matter how good they are doing. I feel that way about some hardware manufacturers.. but I wont name names as much as I would like to.. just cant let that old grudge go.. 

 

But really, there is no need to pledge allegiance to any manufacturer. They are only worthy of your money if they have something to offer you at a price you think is fair. And in return we get insanely high performance machines at a decent price.

[fonsui]

We agree for the most part, but I'd like to restate a few things:

 

13 minutes ago, Fasauceome said:

I don't think Intel is being unfairly judged.

I am not addressing whether or not Intel is being judged fairly, that's irrelevant to me; I am not in a parasocial relationship with a semiconductor corporation. I am targeting purchasing-related information being provided to consumers, which has been largely negative despite an acceptable if admittedly not very inspiring product.

 

13 minutes ago, Fasauceome said:

Relative to 10th gen, the product is actually the same

 

13 minutes ago, Fasauceome said:

intel keeps moving sideways

I listed off, and we both agree, that Rocket Lake contains a good number of catch-ups with Ryzen compared to Comet Lake, those being memory controller, PCIe lanes and spec version, and an additional improvement I allowed "chipset" to cover, that being that the chipset link is now PCIe 3.0 x8, equivalent in bandwidth to the AM4 PCIe 4.0 x4. And, again, this isn't about track record, this is about examining the market of available products and pricing at a point in time, that point being now.

 

They've largely caught up in performance and functionality, the pricing is "acceptable", and the products are in stock. Let the people compute!

[bellabichon]

I think I agree with you that Rocket Lake is 'fine'. But 'fine' doesn't get consumers excited. 'Fine' doesn't create a line of people wanting to throw their money at you. Part of what was so exciting about AMD's (relatively) recent resurgence is that it brought some real competition to the CPU market. For the first time in a while, it felt like Intel would stop resting on their laurels and actually try to innovate in a major way. But aside from a few select products (8700K comes to mind), Intel has only budged when AMD has forced them to. So I don't think that the frustration you're seeing is solely directed at Rocket Lake. Rather, I think it's the overall negative reaction to a long trend of incremental performance increases. 

[fonsui]

Just now, bellabichon said:

I think it's the overall negative reaction to a long trend of incremental performance increases

Agreed, and that is why I felt it important to try to look at the slice of time we exist in right this moment, and make decisions based on what reality looks like now, rather than allowing frustration to prevent us from considering a viable option.

[freeagent]

Actually.. Intel's line does look compelling, even though its not quite as strong, chews a bit more power, has thermals mere mortals complain about. 

 

But from a tweakability standpoint, as someone who enjoys overclocking, Intel does look pretty good. If I had the cash right now I would buy one just to play with, like I do with my kinda boring AMD

[fonsui]

For sure, basically my whole thing is that everyone can buy their preference/what is available right now, and not feel like they had to settle - both product lines are acceptable.

[dean_5234]

34 minutes ago, fonsui said:

TL;DR: Rocket Lake seems to be suffering from an expectation management problem more than actually being a "Bad Product". It doesn't leave customers with poor performance or non-working systems or functions, it's not wildly overpriced, and if people can snag a working current-gen CPU in this climate that in and of itself is a win.

I personally believe that 11th gen i5s look great, especially with no Zen 3 Ryzen 3s or Non-X R5s. The 11400 and 11600 look like great value chips and if I was planning on an upgrade I would certainly consider them. However, I am absolutely stunned by how bad the 11900K and 11700K were, especially with the 11900K being barely an upgrade in performance vs the 10900K and losing 2c/4t. As Gamers Nexus put it, those two are "wastes of sand"

 

7 minutes ago, bellabichon said:

I think I agree with you that Rocket Lake is 'fine'. But 'fine' doesn't get consumers excited. 'Fine' doesn't create a line of people wanting to throw their money at you. Part of what was so exciting about AMD's (relatively) recent resurgence is that it brought some real competition to the CPU market. For the first time in a while, it felt like Intel would stop resting on their laurels and actually try to innovate in a major way. But aside from a few select products (8700K comes to mind), Intel has only budged when AMD has forced them to. So I don't think that the frustration you're seeing is solely directed at Rocket Lake. Rather, I think it's the overall negative reaction to a long trend of incremental performance increases. 

I couldn't agree more. Legitimately only one processor from Rocket Lake is better than AMDs option. Some are similar or equal (111600k) but the 11400 seems like the only processor that outdoes AMD's competition (3600 at the moment, HOWEVER AMD might drop a much cheaper Zen 3 R5 that equals or betters the 11400)

Rocket lake is so underwhelming and boring, I wish Intel and AMD would fight for the lead, not Intel fighting to be equal to AMD

[fonsui]

Just now, davemc said:

I wish Intel and AMD would fight for the lead, not Intel fighting to be equal to AMD

Oh, they are. Stay tuned for the Gelsinger show. 7nm (which is likely to be comparable in density to competitors 5nm) is on-target, Alder Lake will debut 10nm (comparable in density to competitors 7nm) on desktop, and then there's all the other fun stuff like getting into the fab-sharing business and of more direct importance to us, the GPU business.

[Entropy.]

51 minutes ago, fonsui said:

- Running on a new chipset which allowed a feature refresh and includes some previously "premium" functionality filtered down from the Zx90 tier

- At least in the same ballpark as Ryzen 5000 overall performance at every tier, even if it doesn't trash it like customers have come to expect

- Not completely ridiculous pricewise, and downright competitive in some tiers/use cases (see 600K, all the -F models)

- Using a brand-new memory controller that provides comparable performance and functionality as Ryzen 5000

- Operating PCIe at 4.0, and has an additional x4 link for NVMe in addition to the standard 16 lanes that have previously been available

- Kinda? But in contrast to AMD it's still not great. Faster ram support on non-z series boards, for example. AMD has allowed for ram overclocking and cpu overclocking on their non-flagship chipsets since the inception of the ryzen series of processors.

-Overall performance? No. The 11900k does not stack up to the 5900x in many different situations favorably. Intel's only really competing favorably right now in comparison to AMD with their i5 chips, namely the i5-11400/f, but if AMD drops lower-tier 5000 series chips, this may change

- Basically same as above. The i9 lineup is the worst, with high power draw, bad temps, and performance that might be considered somewhat competitive to AMD's competing chips if you ignore certain situations and obviously any workloads that heavily rely on multiple threads. i7's are okay for now, with more reasonable temps and prices for performance that isn't too far off from the i9's, but that's basically only because the 5800x is a hot mess right now (literally and metaphorically). The price of that chip is too high for what it brings and it's temps are unreasonable. The 11700k is basically competing with the 10700k. i5s are where Intel really shines this generation, those chips are excellent all-round. i3's are basically a refresh of last-gen, worth buying but not worth commenting on. Better than AMD's chips for now, but they're pretty boring. And again, If AMD brings out midrange and lower-end 5000-series chips, that could change. All these chips are helped by the fact that Ryzen 5000 series chips are somewhat difficult to aquire (Specifically the 5900x and especially the 5950x, the 5600x is somewhat hard to acquire, usually, at least for around MSRP). 

-No comment, not really familiar with the memory controllers on the chipsets.

- Fair enough, but this isn't new stuff period, just new from intel. This shows progress but shouldn't really be a point made to show how well intel is doing in contrast to AMD. AMD has already had this implemented for over a year and a half. 

51 minutes ago, fonsui said:

Does it trounce Ryzen, unqualified? Absolutely not. It does, however, put Intel parts in the market that are competitive choices that provide a lot more feature parity with Ryzen 5000 than Comet Lake or prior gens did. This was a parity move to keep them in the game, not a run at the crown. Intel's money in the desktop consumer market is on Alder Lake and Xe HPG, that's what they really want to push hard.

Competitive? Maybe sometimes. At this point at least 1/2 of their lineup has been made irrelevant by AMD chips. Intel isn't doing so hot right now. 

Quote

Now, if we look at the HEDT space Intel has slipped even further and Threadripper simply dominates, workstation Xeons haven't been keeping up for quite some time now, and the non-workstation Core X-Series CPU (latest being Cascade Lake AKA Yet Another Skylake) and chipset (X299) are so far behind they're barely worth talking about, and I've seen no plans to change that.

Not much to say here, AMD's basically kicking the dead body of Intel's HEDT chips. Oh well, new architecture, new hopes. 

[Falkentyne]

I said in my review (which most people like more than any of the tech tuber reviews except Skatterbencher, who was a world champion competitor for LN2 overclocking once) these things already.  Rocket Lake is necessary and is a stop gap for Alder Lake.  ADL will have two IMC's.  That's why RKL has gear 2.  But removing one of the IMC's and maintaining Z590 compatibility with CML really hurt RKL here.  

 

Intel is going to be fiercely competitive with ADL.  Whether they will beat AMD completely is under NDA.  Just imagine RKL as "Pentium M" before Core architecture. 

[freeagent]

[dean_5234]

16 minutes ago, fonsui said:

Oh, they are. Stay tuned for the Gelsinger show. 7nm (which is likely to be comparable in density to competitors 5nm) is on-target, Alder Lake will debut 10nm (comparable in density to competitors 7nm) on desktop, and then there's all the other fun stuff like getting into the fab-sharing business and of more direct importance to us, the GPU business.

Yea, I worded mine wrong. What I shouldve said is that Rocket Lake was fighting for the lead.

 

Im excited for Intel 7nm (finally!!!)

I hope we as consumers just keep getting more and more great CPUs and 12th gen is great

[fonsui]

20 minutes ago, Falkentyne said:

Rocket Lake is necessary and is a stop gap for Alder Lake

 

20 minutes ago, Falkentyne said:

Just imagine RKL as "Pentium M" before Core architecture. 

This is the feel I have been getting

[Mister Woof]

I'm what you would call an Intel fanboy on this forum (although I do find value in AMD CPUs) - but I don't think 11th generation should have ever existed.

 

Despite the 11400 and 11600k being "fine", they're still not really compelling enough on their own. They only are meaningful because AMD has nothing in that bracket.

 

10th generation is only noteworthy because its available and inexpensive, and still keeps up with Ryzen 5000 in most gaming scenarios when configured properly.

 

They should have just slashed 10th gen prices, skipped Rocket Lake, and went all in on Alder Lake.

[Bombastinator]

6 minutes ago, Mister Woof said:

I'm what you would call an Intel fanboy on this forum (although I do find value in AMD CPUs) - but I don't think 11th generation should have ever existed.

 

Despite the 11400 and 11600k being "fine", they're still not really compelling enough on their own. They only are meaningful because AMD has nothing in that bracket.

 

10th generation is only noteworthy because its available and inexpensive, and still keeps up with Ryzen 5000 in most gaming scenarios when configured properly.

 

They should have just slashed 10th gen prices, skipped Rocket Lake, and went all in on Alder Lake.

It may have more to do with how their marketing and distribution system works than anything else.  It might have been less bad for the company to do what they did instead of your path because of reasons that have nothing to do with the product they sell.

[Mister Woof]

1 hour ago, Bombastinator said:

It may have more to do with how their marketing and distribution system works than anything else.  It might have been less bad for the company to do what they did instead of your path because of reasons that have nothing to do with the product they sell.

Obviously that's from a consumer standpoint - intel as a company still needs to satisfy their shareholders so they likely just need to have something new to sell or look like they are failing.

 

AMD was able to hibernate and get Ryzen ready, but a large world leading company like Intel doesn't have the same luxury.

 

They likely know their product sucks but have worked their mind share loses on this into their greater strategy.

 

It's just unfortunate they didn't have anything really good this time.

[yenic]

I like Rocket Lake a lot, because my 5900X system isn't stable after recent AGESA updates. It doesn't matter how fast or cool something runs, if it's not reliable. Mine freezes immediately when starting up P95. And frankly, I moved on Ryzen with the 1800X, this is my 4th one, while I've been able to stabilize them, I've had to fight them all. Disabling DRAM power down, disabling C-States, it's a joke. I did not have to do so with any of my Intel rigs. I think I'm done. What do you think Fine Wine is? Hacked products that are stabilized a year or two after you buy them.

 

Strongly considering selling my 5900X and board locally, and moving to a 10850K on Z490, or 11900K on Z590. I'm willing to bet I go 5+ years without having to think about if my PC is still-stable. I can just install BIOS updates and software from Intel, and not worry that it'll start crashing right after. I ran a Q9450 for 7+ years, daily, and hard. Sandy Bridge came, and I didn't care, I was running reliably every single day. Loved it.

 

Between the 10850K and 11900K, both are slower than my current Ryzen, the 10-core sounds like it'd be the closest. I lose two cores, gain Intel's platform engineering, quality assurance, and reliability. AMD has proven they can make a CPU, but IMO they have not proven that they can make a platform. Their chipsets were trash in the slot A / socket A Athlon days, Nvidia had to rescue them with Nforce.. and they're still not quite there yet on AM4.

 

I should never have betrayed the brands I fanboy on, so much wasted time over the past few years just for a few more cores. I'm open minded in general but increasingly closed minded. I've long preferred Intel-Nvidia-Asus, due to my experiences with them over the course of the past 25 years building, and 10 more years of Intel prebuilts before that.

 

That's my two cents. I've given AMD plenty of money, 4 Ryzen CPUs, 2 boards, I know how to build and work a PC, so if you love your Ryzen setup and it's stable, remember there's millions of hardware/software combinations for PCs.. so don't let your butt get hurt over my anecdotes and opinion. AMD just can't afford or doesn't do enough QA to match Intel. And/or, other companies just don't use them to QA against either. I paid everyone's current-favorite and precious AMD good money, and I don't have to like them or what they gave me in return. Intel for life, from here on out. Rocket Lake kicks ass, it includes Intel stability and support, so it's the current market leader for anyone trying to run a business like me, and not just play games. Half the gamers on these forums don't have stable systems, and just don't know it, "weird stuff happens" occasionally to them but their games generally work so they shrug it off, but have never ran memtest86 all night once in their life.

 

So in other words without the inflammatory language, my point is that a CPU is more than its performance, it also has a platform. AMD may have better performing CPUs today (in nearly every way), but they don't have Intel beat on platform. Platform is more important since that's everything that isn't crunching 0s and 1s, it ensures if you have a widely compatible and thus stable system, or not.

[freeagent]

Dunno what to tell you man.. my current AMD is just as, if not more stable than any of my Intel systems

 

Im not an AMD fanboy either. This will probably be my last one for a bit. I like to tinker and this has some but not enough tinkering built in.

[Alan G]

Why has AMD stock dropped 10% since January and Intel's price risen by 30%?  Does the investing community know something that most do not?  Disclosure:  I've been an Intel shareholder for 20 years.

[fonsui]

6 hours ago, Alan G said:

Intel's price risen by 30%?  Does the investing community know something that most do not?

Even before the Swann/Gelsinger transition was announced, Intel was nearing the mark with 10nm yield that would allow them to deploy it for desktop, the Xe DG1 had been shipped, and DG2 was well on the way. Then came the announcement, and since Gelsinger has taken over, Intel has shown consistent strengths by bringing back a sizeable amount of brainpower that had left the organization, and defining their IDM 2.0 strategy which includes 20 billion USD worth of fab investment supporting North American silicon independence from the East, as well as offering fab services to the market. Reports from inside Intel indicate a morale boost and re-energizing, based on the new leadership and the direction they are taking the company. The employees believe now, and that is one of the most important aspects of success.

 

If we compare the technology portfolio, even just for CPUs, it puts some things in perspective - Intel has been stamping out 28-core dies and is bumping that even higher, while AMD has not deployed a single CPU die, ever, that had more than 8 - this is not to say whether monolithic or chiplet is a better CPU design strategy, only making the point of the extreme capabilities of Intel engineering and fabrication. Intel EMIB packaging technology allows for high-performance interconnects between dies on a high-density package, giving an edge over the chiplet design used in Zen.

 

AMD has made some incredible strides catching and beating Intel on raw per-core and all-core total performance, beat them to the punch with chiplet design, beat them to the punch with PCIe 4.0 both on desktop and in the datacenter (which is a bigger deal than it sounds, since now all PCIe 4.0 interoperability validation is going to take place on AMD Epyc systems), is completely beating them up on their share in both of those markets, and we can most definitely thank them for providing such competition as to break Intel out of their years-long slumber during which almost no innovation took place. This fight is not over, it's only getting started, and the consumers are the ones who stand to gain. Good stuff all around.

 

So, to finally answer your question: yes, I think the investors know something that "we" the community haven't accepted, which is the entire reason that I began this thread. Years without innovation of any sort from Intel have worn on people, and they are allowing the emotions of disappointment and anger to define their perception of Intel as a company, which prevents them from "trusting" them or "giving them credit". That luxury is only available to home users who use the products - investors however, have Real Money on the line, and making emotional decisions with Real Money ends you up with less, or none.

[fonsui]

7 hours ago, James Evens said:

In a nutshell that's why the i9-11900k is is outright bad.

The BAD/GOOD false dichotomy is what I am addressing here, and the reason I very specifically chose the word "fine". The thing works, provides acceptable feature and performance parity to the competing products, and costs.. slightly more than we would like. It's not "good". It's also not "bad". It's just "fine".

 

The takeaway is that the deciding factors for a customer can be availability and preference, they do not have to feel bad for getting either product, both will serve the purpose well within a similar price range, whereas previously Intel was lagging notably in a number of areas. Catch-up, not a run at the crown.

[Bombastinator]

1 hour ago, fonsui said:

Even before the Swann/Gelsinger transition was announced, Intel was nearing the mark with 10nm yield that would allow them to deploy it for desktop, the Xe DG1 had been shipped, and DG2 was well on the way. Then came the announcement, and since Gelsinger has taken over, Intel has shown consistent strengths by bringing back a sizeable amount of brainpower that had left the organization, and defining their IDM 2.0 strategy which includes 20 billion USD worth of fab investment supporting North American silicon independence from the East, as well as offering fab services to the market. Reports from inside Intel indicate a morale boost and re-energizing, based on the new leadership and the direction they are taking the company. The employees believe now, and that is one of the most important aspects of success.

 

If we compare the technology portfolio, even just for CPUs, it puts some things in perspective - Intel has been stamping out 28-core dies and is bumping that even higher, while AMD has not deployed a single CPU die, ever, that had more than 8 - this is not to say whether monolithic or chiplet is a better CPU design strategy, only making the point of the extreme capabilities of Intel engineering and fabrication. Intel EMIB packaging technology allows for high-performance interconnects between dies on a high-density package, giving an edge over the chiplet design used in Zen.

 

AMD has made some incredible strides catching and beating Intel on raw per-core and all-core total performance, beat them to the punch with chiplet design, beat them to the punch with PCIe 4.0 both on desktop and in the datacenter (which is a bigger deal than it sounds, since now all PCIe 4.0 interoperability validation is going to take place on AMD Epyc systems), is completely beating them up on their share in both of those markets, and we can most definitely thank them for providing such competition as to break Intel out of their years-long slumber during which almost no innovation took place. This fight is not over, it's only getting started, and the consumers are the ones who stand to gain. Good stuff all around.

 

So, to finally answer your question: yes, I think the investors know something that "we" the community haven't accepted, which is the entire reason that I began this thread. Years without innovation of any sort from Intel have worn on people, and they are allowing the emotions of disappointment and anger to define their perception of Intel as a company, which prevents them from "trusting" them or "giving them credit". That luxury is only available to home users who use the products - investors however, have Real Money on the line, and making emotional decisions with Real Money ends you up with less, or none.

?! This “AMD has not deployed CPUs with more than 8 cores” thing is emphatically not true.  More than 8 cores is basically their deal.  The 3900 had 12, as does every “x900” chip.  The x950s have sixteen, and the threadripper stuff had more yet.  Not even going into the epic stuff which is what intel’s 28 core CPUs compete (unsuccessfully) with. There was a period where their single thread performance was so bad that they were not competitive with intel stuff with fewer cores.  That ended with ryzen2 though and they’re on ryzen3 atm. They may have not bothered to make individual chipsets with more than 8 cores but that is not the same thing.

[yenic]

8 hours ago, freeagent said:

Dunno what to tell you man.. my current AMD is just as, if not more stable than any of my Intel systems

 

Im not an AMD fanboy either. This will probably be my last one for a bit. I like to tinker and this has some but not enough tinkering built in.

I tried to head off comments like this with- "so if you love your Ryzen setup and it's stable, remember there's millions of hardware/software combinations for PCs.. "

 

I've had stable Ryzen rigs, "4" of them in fact.. 1800X, 1700, 2700X and for a brief glimmer in time, a 5900X. I'm not a bumbling fool, ComboAM4v2 really is a mess right now. They're supporting most of their chipsets and vendors are supporting an enormous amount of boards. It was a mess when PinnaclePi was launched too. I don't recall having issues off the top of my head with ComboAM4(v1). I think they're in over their head. CPUs they indeed have today. Chipset/platform, still a weakness. They're fortunate it's essentially an SoC..

1 hour ago, fonsui said:

Even before the Swann/Gelsinger transition was announced, Intel was nearing the mark with 10nm yield that would allow them to deploy it for desktop, the Xe DG1 had been shipped, and DG2 was well on the way. Then came the announcement, and since Gelsinger has taken over, Intel has shown consistent strengths by bringing back a sizeable amount of brainpower that had left the organization, and defining their IDM 2.0 strategy which includes 20 billion USD worth of fab investment supporting North American silicon independence from the East, as well as offering fab services to the market. Reports from inside Intel indicate a morale boost and re-energizing, based on the new leadership and the direction they are taking the company. The employees believe now, and that is one of the most important aspects of success.

 

If we compare the technology portfolio, even just for CPUs, it puts some things in perspective - Intel has been stamping out 28-core dies and is bumping that even higher, while AMD has not deployed a single CPU die, ever, that had more than 8 - this is not to say whether monolithic or chiplet is a better CPU design strategy, only making the point of the extreme capabilities of Intel engineering and fabrication. Intel EMIB packaging technology allows for high-performance interconnects between dies on a high-density package, giving an edge over the chiplet design used in Zen.

 

AMD has made some incredible strides catching and beating Intel on raw per-core and all-core total performance, beat them to the punch with chiplet design, beat them to the punch with PCIe 4.0 both on desktop and in the datacenter (which is a bigger deal than it sounds, since now all PCIe 4.0 interoperability validation is going to take place on AMD Epyc systems), is completely beating them up on their share in both of those markets, and we can most definitely thank them for providing such competition as to break Intel out of their years-long slumber during which almost no innovation took place. This fight is not over, it's only getting started, and the consumers are the ones who stand to gain. Good stuff all around.

 

So, to finally answer your question: yes, I think the investors know something that "we" the community haven't accepted, which is the entire reason that I began this thread. Years without innovation of any sort from Intel have worn on people, and they are allowing the emotions of disappointment and anger to define their perception of Intel as a company, which prevents them from "trusting" them or "giving them credit". That luxury is only available to home users who use the products - investors however, have Real Money on the line, and making emotional decisions with Real Money ends you up with less, or none.

TLDR; you're not going to stop Intel. They're still in the game today in every aspect, and they're coming back.

1 hour ago, fonsui said:

The BAD/GOOD false dichotomy is what I am addressing here, and the reason I very specifically chose the word "fine". The thing works, provides acceptable feature and performance parity to the competing products, and costs.. slightly more than we would like. It's not "good". It's also not "bad". It's just "fine".

 

The takeaway is that the deciding factors for a customer can be availability and preference, they do not have to feel bad for getting either product, both will serve the purpose well within a similar price range, whereas previously Intel was lagging notably in a number of areas. Catch-up, not a run at the crown.

 

They're still in the performance game for the majority of buyers. 6-8 cores are most of the rigs I see around in sigs in places like this forum. That's what the kids, gamers and hobbyists are buying. The fact Intel isn't currently able to market to a buyer like me that wants 10-24 cores in a mITX system for a great price, really is out of the realm of the mass market. I'm paying for it in other ways though. My cheap, high core count AMD 5900X does not include the Intel or Apple markup, where you pay more and receive more in higher quality longterm support. Rocket Lake is indeed fine. Single core performance is within striking distance, and bang for buck on the 6-core parts is pretty compelling. I'd be typing this post on one right now if they had a 10-core 11th gen available.