Are we going to see good Intel processors anymore?

I wanted to go AMD but the lack of stock made it impossible. Been waiting a couple years for the right time. Wanted a 5900x but settled on a 10850k and Z490 that I'll be installing tomorrow. I'm already going from 4 cores to 10 and the benchmarks between the two are very close. I think Intel likely to leap frog AMD on single core performance next gen at the expense of power, heat, and cores. Intel is doing what AMD was going with 200+ watt CPUs to make up for IPC disadvantage. Not ideal but at least I can get my hands on an Intel setup and it was $160 cheaper and my H110 AIO will work.
 
  • Like
Reactions: noko
like this
The current situation basically mirrors where AMD was in prior years. Not winning, but still perfectly fine. With supplies being limited for AMD's stuff and gaming performance still neck and neck, there's till a place for Intel.

There's one huge difference between Intel's current problems and AMD's problems pre-zen.

AMD had access to competitive manufacturing, but was crippled by a bad core design. That was something that can be fixed in one shot with a new design with only the time needed to do so a limiting factor.

Intel's problem is that they apparently tried to do too many changes at once going from 14 to 10nm. When something didn't work as expected were unable to figure out how to fix it for years; and apparently are still having enough problems that their yields are really low. Every year of delay here is also a year of delay on their future 7nm, 5nm, etc processes because each one builds on the one before so you can't just skip one entirely to catch up without having an even more extreme version of trying to do too much at once than what caused the 10nm disaster. In the best case Intel manufacturing might be able to slowly claw back their lost process lead and catch back up to Samsung and TSMC; but doing so is going to take a number of years. And the rumor mill says tehy're having problems with their 7nm process and its availability date is slipping too.

Giving up on in house manufacturing and switching to Samsung/TSMC for a majority of their processors isn't an option either; unlike when AMD dropped Global Foundaries, Intel's manufacturing needs are too large to be absorbed by either of the major 3rd party foundries. Intel's ~800k wafers/month is roughly 1/3rd of TSMC's total capacity and 1/4th of Samsung's total capacity and a much larger share of their relevant capacity (since a large but unspecified share of their wafer starts are from DRAM/NAND foundries that can't make CPUs). Neither company is willing to spend 10s of billions of dollars on new capacity that they might only have a customer for for a few years. It'd take Intel giving up on developing new higher density processes to make that investment non-suicidal for TSMC/Samsung; but the several year lead time it'd take either of those companies to build new fabs for making Intels chips would leave Intel stuck with their current obsolete production for years of being increasing noncompetitive with AMD/ARM.


https://www.extremetech.com/computi...-capacity-for-intel-views-orders-as-temporary
 
There's one huge difference between Intel's current problems and AMD's problems pre-zen.

AMD had access to competitive manufacturing, but was crippled by a bad core design. That was something that can be fixed in one shot with a new design with only the time needed to do so a limiting factor.

Intel's problem is that they apparently tried to do too many changes at once going from 14 to 10nm. When something didn't work as expected were unable to figure out how to fix it for years; and apparently are still having enough problems that their yields are really low. Every year of delay here is also a year of delay on their future 7nm, 5nm, etc processes because each one builds on the one before so you can't just skip one entirely to catch up without having an even more extreme version of trying to do too much at once than what caused the 10nm disaster. In the best case Intel manufacturing might be able to slowly claw back their lost process lead and catch back up to Samsung and TSMC; but doing so is going to take a number of years. And the rumor mill says tehy're having problems with their 7nm process and its availability date is slipping too.

Giving up on in house manufacturing and switching to Samsung/TSMC for a majority of their processors isn't an option either; unlike when AMD dropped Global Foundaries, Intel's manufacturing needs are too large to be absorbed by either of the major 3rd party foundries. Intel's ~800k wafers/month is roughly 1/3rd of TSMC's total capacity and 1/4th of Samsung's total capacity and a much larger share of their relevant capacity (since a large but unspecified share of their wafer starts are from DRAM/NAND foundries that can't make CPUs). Neither company is willing to spend 10s of billions of dollars on new capacity that they might only have a customer for for a few years. It'd take Intel giving up on developing new higher density processes to make that investment non-suicidal for TSMC/Samsung; but the several year lead time it'd take either of those companies to build new fabs for making Intels chips would leave Intel stuck with their current obsolete production for years of being increasing noncompetitive with AMD/ARM.


https://www.extremetech.com/computi...-capacity-for-intel-views-orders-as-temporary
I guess I'll ask the question about what "noncompetitive" means. For 99% of users, is there a noticeable difference between using mid-tier processor and the latest and greatest? For normal desktop usage (MS office, web browsing, watching videos, cloud apps like Salesforce, light gaming, etc.) I think you'd find there's little no difference at all. If Intel can put out a product that's cheaper and on the shelf they can remain competitive for most users for quite a while. Especially with the right OEM contracts. Those are far more important than pure horsepower that most applications don't use. I'd look to Intel's business strategy moreso than their manufacturing capabilities.
 
Intel will have to do something if they want to stay in business. AMD and Apple make processors that pretty much destroy Intel. So I would expect Intel to do something soon.
I mean... not really. All that will happen is they lose revenue. Well, for now. Nobody can produce as many desktop chips as Intel and I think even Apples chips are getting larger even on 5nm so that's a lot of wafers if they want to supply that market.

There will be better intel chips once they get their process under control and there is a chance they will have the process lead on AMD if apple gobbles up all the capacity on the latest nodes from TSMC.
 
I guess I'll ask the question about what "noncompetitive" means. For 99% of users, is there a noticeable difference between using mid-tier processor and the latest and greatest? For normal desktop usage (MS office, web browsing, watching videos, cloud apps like Salesforce, light gaming, etc.) I think you'd find there's little no difference at all. If Intel can put out a product that's cheaper and on the shelf they can remain competitive for most users for quite a while. Especially with the right OEM contracts. Those are far more important than pure horsepower that most applications don't use. I'd look to Intel's business strategy moreso than their manufacturing capabilities.

at lower levels the performance difference isn't that noticeable (especially for people who've never bought fast systems), but margins are much higher on top of the line Xeons and I7/9 chips than i3s and celerons; losing a big chunk of the former is going to hammer their bottom line hard and make trying to fund a recovery even harder. having to sell high end chips at low end prices is even worse than just the sticker amounts imply because the big dies are a lot more expensive to make. And while bargain hunting consumers might not immediately notice, at some point they're going to realize that the Intel laptop with a 45W i9 rebadged to i3 only lasts 4-7 hours on battery while the equally fast AMD R3/arm laptop gets 10-20 hours and as much less or no fan noise. The OEMs will know that is coming as well, and will start shifting more of their product lines over to avoid it.

It will be a top down implosion though. They're already losing share in the enthusiast market rapidly; and while Data Center customers are slower to change that's happening too and will probably overtake the enthusiast market in retreat at some point since they almost all care about perf/watt while a lot of enthusiasts are indiferant to how much power their box draws. If they can't get their manufacturing sorted in the next few years Intel's probably screwed.
 
Yeah with the cash they have I am sure we will see them around with something good in a few years time or more.
 
Yeah with the cash they have I am sure we will see them around with something good in a few years time or more.
Yeah, sure. Everyone has been saying that for years. Money doesn't equate to technology unless -

1) You have the talent in-house, which clearly Intel is struggling with at the Management and Engineering Level
2) Outright buy it. Buying a high level fab is out of the question here. Intel has failed also to bring in adequate outside talent to fix (1)
 
If you want a rig that "just works", and not have to worry about making it work, I think Intel is still better. Especially since AMD jacked up their MSRP. Aslo now that AMD has gained somewhat equal marketshare in the home PC/ console segment, look for more side channel attacks and other exploits to surface, that work on AMD platforms.
 
I don't think I'm ready to say the whole world has gone AMD yet. I think once I see the standard lines of HP, Lenovo, Dell, etc.. all switch to AMD (where you have hunt to find Intel, that is, the positions have swapped), then I'll be ready to say that Intel is in real trouble.

In short, define good? Right now, looks like "good" means Intel is still pretty dominant. With that said, AMD getting 20+% of the CPU market (if that's accurate), does mean that Intel is probably going nuts right now.

AMD CPU prices have gone up while Intel's have plummeted. So... while AMD might be cheaper, it's not that much cheaper nowadays. Competition is good. And Intel still reigns by some pretty hefty margins market share wise. Now... if AMD can continue to move up and get to 40% and get the major manufacturers to take them seriously. Again, that's going to spell trouble for Intel.

Right now, desktop/laptop wise, AMD still has "toy" status. Would love to see this change. Server wise, I think AMD is doing ok, but still not nearly as available (perceptively) as Intel servers.

When you throw in a "change point" to a company, it puts all things on the table. Thus, in the server world, switching from x86-64 made by Intel can mean switching away from x86-64 altogether.

If you want me to obtain a reasonably fast x86-64 host on the cheap with at least 32GB of memory... it's going to be Intel, and usually by a price factor of being less than an AMD system by 2-3x. I was hoping to see this change as 1st gen Ryzens are pretty old now. But it hasn't. Why? Volume.
 
Intel will have to do something if they want to stay in business. AMD and Apple make processors that pretty much destroy Intel. So I would expect Intel to do something soon.
NIZMOZ Destroy, strong word, depends on what you are doing..4k gaming? almost identical, multi-tasking, depends on the task..

Do not forget, Intels market is servers and business and OEMs (Dell / HP et cetera) that is where they make all their money.
 
Hard to recommend the 10900k over the 10850k being that they offer effectively identical performance. 10850k @ $400 IMO is the processor to get if you want more than 6 cores and you can't find a 5900x.
sabrewolf732 except your playing the lottery with a 10850k which can use more power than the 10900k, upwards of 20W or more for some chips tested and some shown to hit 102C. Where as the 10900k has very stringent requirements to be sold as the 10900k.
 
NIZMOZ Destroy, strong word, depends on what you are doing..4k gaming? almost identical, multi-tasking, depends on the task..

Do not forget, Intels market is servers and business and OEMs (Dell / HP et cetera) that is where they make all their money.
Well they're losing market share in that space as well.
 
If you want a rig that "just works", and not have to worry about making it work, I think Intel is still better. Especially since AMD jacked up their MSRP. Aslo now that AMD has gained somewhat equal marketshare in the home PC/ console segment, look for more side channel attacks and other exploits to surface, that work on AMD platforms.
vick1000 which almost no home user has to worry about side channel attacks and Scepter anyways since they are not running shared work loads on their computers and for someone to expoit would already need to be in their network / on their system, which you have bigger problems.

I do agree though, when you read over the AMD threads of so many people having problems using memory profiles and not being able to use the rated speeds cause it isnt stable and such...makes me wonder if I should go AMD or not..
 
vick1000 which almost no home user has to worry about side channel attacks and Scepter anyways since they are not running shared work loads on their computers and for someone to expoit would already need to be in their network / on their system, which you have bigger problems.

I do agree though, when you read over the AMD threads of so many people having problems using memory profiles and not being able to use the rated speeds cause it isnt stable and such...makes me wonder if I should go AMD or not..
To be fair a lot of issues are from users who do no research and just assume things. Though I concur memory support could improve all around.
 
To be fair a lot of issues are from users who do no research and just assume things. Though I concur memory support could improve all around.
Memory issues can exist, but not to go with ryzen because of it is foolish. Like you said, research and a little due dilligence goes a long way in preventing such issues.
 
Now... if AMD can continue to move up and get to 40% and get the major manufacturers to take them seriously. Again, that's going to spell trouble for Intel.
For starters one would need to be able to buy said AMD processors which atm is pretty hard to do.
 
For starters one would need to be able to buy said AMD processors which atm is pretty hard to do.
True. And the Ryzen 5000 are more much expensive than Intel at the same core count. Even the Ryzen 3000 line is as or more expensive than Intel. And Intel is very easy to find and buy everywhere. Not AMD.
 
This is the boat I am in right now, I wanted an AMD 5000 series, my i5-8600 is holding back my new 6800 in some games. I can get an intel 10850 or 10900k right now and build a rig off that and see what comes later for socket 1200, i dont need PCIe 4 or DDR5 either....I would love to get a 5800X right now and be done with it....but then i do not want to have to wait until who knows when. Intel annoucing things in CES 2021 (i think and soon) the waiting game then starts, or I just say screw it and get an intel platform again...
 
Mega6 Here in CAD they are even more rare it seems. I am in a discord channel that has stock updates for all Canadian sites and the only alerts we get are 5800 in Japan...
 
Mega6 Here in CAD they are even more rare it seems. I am in a discord channel that has stock updates for all Canadian sites and the only alerts we get are 5800 in Japan...
Seems like you've answered your own question then if you need something now. Life is full of compromises, especially this past year.
 
5800x

from china but it says international ship 6-11 days. Approximately C $368.14

7 left
 
That seems too good to be true at that price for CAD considering retailers are all selling for $600 CAD +
 
since AMD now has 50% desktop cpu market, i hope intel undercuts in price

that's the kind of competition we need

i5-11400 for $150 would be nice
 
There certainly is a lot of hyperbole and nonsense concerning Intel "not being any good anymore" or being just bad in comparison. While it is true they are no longer the outright performance leader currently, that by no means translates to them being terrible. People blow things WAAAY out of proportion and it's really just the fanboys of either company in a pissing match that perpetuates it. If I were to sit you down at either of my gaming rigs and put my RTX 3090 in either of them, I 100% guarantee no one, not even the most hardened fanboy of AMD or Intel would be able to tell the difference with any measure of certainty. They are virtually identical in game performance with one a couple percentage points ahead in one game and then trailing by percentage points in the next. You will not be able to differentiate them in a blind test. This is a good thing, that means there is choice again in the CPU arena. Go with whatever you want or with what is cheaper that suits your needs, neither option is a "bad" choice.
 
There certainly is a lot of hyperbole and nonsense concerning Intel "not being any good anymore" or being just bad in comparison. While it is true they are no longer the outright performance leader currently, that by no means translates to them being terrible. People blow things WAAAY out of proportion and it's really just the fanboys of either company in a pissing match that perpetuates it. If I were to sit you down at either of my gaming rigs and put my RTX 3090 in either of them, I 100% guarantee no one, not even the most hardened fanboy of AMD or Intel would be able to tell the difference with any measure of certainty. They are virtually identical in game performance with one a couple percentage points ahead in one game and then trailing by percentage points in the next. You will not be able to differentiate them in a blind test. This is a good thing, that means there is choice again in the CPU arena. Go with whatever you want or with what is cheaper that suits your needs, neither option is a "bad" choice.
Sure we could, play long enough and the AMD rig would be the one that crashes. :) [/fanboitrigger]
 
since AMD now has 50% desktop cpu market, i hope intel undercuts in price

that's the kind of competition we need

i5-11400 for $150 would be nice
If you are talking about this:
https://www.techradar.com/news/amd-...u-market-share-for-the-first-time-in-15-years

it look to me they are not talking about market share, they are talking about the percentage of people that run a particular benchmark (passmark), a bit like looking at a steam hardware survey but a much smaller sample size.

It could mean that AMD has 50% of the enthusiast run benchmark on my machine market share more than a 50% overall market share, with just extremely misleading headline.

FZqmVVPkgwK3Nd96cZnxJP-970-80.jpg.webp
 
Yeah, I missed that comment. AMD has no where near 50% of the desktop CPU market. It still isn't even close, just a cursory glance at the latest Steam survey shows Intel still has a commanding 75% share. While sure, that's not an all encompassing measure it is wildly popular and common for a PC to have Steam installed even on the "lightest" of PCs in terms of gaming hardware.
https://store.steampowered.com/hwsurvey/processormfg/
 
Steam Survey is bullshit. We all have our sources.

View attachment 317782
LOL ok I guess you just ignore the fact that Steam is the most widely used and popular digital storefront? Calling Steam "bullshit" is incredibly ignorant on your part. While true it's not all encompassing as I said, it is a very good indicator of what the majority of people are using.

Again that's from "passmark" benchmark which is a niche piece of software especially in comparison to Steam.
 
LOL ok I guess you just ignore the fact that Steam is the most widely used and popular digital storefront? Calling Steam "bullshit" is incredibly ignorant on your part. While true it's not all encompassing as I said, it is a very good indicator of what the majority of people are using.

Again that's from "passmark" benchmark which is a niche piece of software especially in comparison to Steam.
Surveys are niche by the very fact they are "Surveys". The point i am making is that there is no reputable data source. They all over the place. Cherry picking data is pointless. I'm "ignorant" because you like the data your source has vs something else? That's ignorant all right.
 
Surveys are niche by the very fact they are "Surveys". The point i am making is that there is no reputable data source. They all over the place. Cherry picking data is pointless. I'm "ignorant" because you like the data your source has vs something else? That's ignorant all right.
You clearly are missing the point and trying to tailor the argument to suit your naive narrative. Steam is far and away the more popular software, and it is installed on almost every "consumer" level gaming PC. You can't say the same for Passmark. It's people like you that perpetuate the nonsense I was speaking about earlier. You read those clickbait articles and then want to try and apply it to the whole market. Your argument that AMD has 50% of the CPU market is ONLY valid if you are talking about in the Passmark benchmark alone NOT the industry as a whole. There is a difference, I highly suggest you learn it.
 
It depends on what is meant by "desktop market". If you look at complete historical ownership, yeah there are more old intel boxes laying around. If it means recent desktop sales, AMD is over 50%.
 
You clearly are missing the point and trying to tailor the argument to suit your naive narrative. Steam is far and away the more popular software, and it is installed on almost every "consumer" level gaming PC. You can't say the same for Passmark. It's people like you that perpetuate the nonsense I was speaking about earlier. You read those clickbait articles and then want to try and apply it to the whole market. Your argument that AMD has 50% of the CPU market is ONLY valid if you are talking about in the Passmark benchmark alone NOT the industry as a whole. There is a difference, I highly suggest you learn it
So we are talking about a sliver of desktop market that:
1) have steam installed
2) actually take the survey
3) are gamers

Hardly representative of the desktop market. Stop comparing passmark, steam or whatever the fuk else and look at the big picture. NO software or survey on a specific platform is going to give an accurate picture of PC ownership.
 
Back
Top