Question Rename E-cores to B-cores

fastandfurious6

Senior member
Jun 1, 2024
456
605
96
Bench-cores

Is it true that E-cores serve the purpose of inflating MT Benchmark numbers?

Since intel's 13th gen where there's tons of E cores, there is a consistent pattern:

benchmarks produce higher scores than actual time-based results

i.e. they may have higher Geekbench or Cinebench numbers than competitors, but they're slower when it comes to majority of real tasks i.e. blender, 7zip, etc etc, also games FPS
 
Last edited:

Meteor Late

Senior member
Dec 15, 2023
289
313
96
Main purpose is to increase MT peformance on the cheap, that is, you get more MT performance in the same area. Not necessarily just in benchmarks, it's just that most people don't actually use apps that leverage all that many cores where that extra MT performance is useful.
Exception would be in Laptop and mobile space, where obviously, if you only have two or 4 P cores, then the E cores are more useful obviously, assuming similar cost.
 
Jul 27, 2020
23,643
16,599
146
Intel always does stupid things in my opinion.

Where is the 10 core Alder Lake with only Golden Cove cores with AVX-512 enabled?
Where is the 10 core Arrow Lake with only Lion Cove cores with AVX-512 enabled?
Where is the 32 or 48 core Bench Monster comprised of only Gracemont cores and now Skymont cores?
Where is the Skymont based Core i3-N405?

Intel is mostly always late. It's like they are paying peanuts to the engineers who can make things happen while it seems AMD respects their talent a lot more which is why they keep on executing well. I hope LBT pops each and every cancerous tumor in the Intel culture and blacklists such people from ever being hired at Intel again.
 

alcoholbob

Diamond Member
May 24, 2005
6,351
423
126
It's too many threads for most apps is the main problem. If they went with 16 P-cores they would need more die surface area for maybe slightly better real-world MT performance in apps. They decided margins were more important than slightly better practical performance.
 

Thunder 57

Diamond Member
Aug 19, 2007
3,438
5,675
136
Intel always does stupid things in my opinion.

Where is the 10 core Alder Lake with only Golden Cove cores with AVX-512 enabled?
Where is the 10 core Arrow Lake with only Lion Cove cores with AVX-512 enabled?
Where is the 32 or 48 core Bench Monster comprised of only Gracemont cores and now Skymont cores?
Where is the Skymont based Core i3-N405?

Intel is mostly always late. It's like they are paying peanuts to the engineers who can make things happen while it seems AMD respects their talent a lot more which is why they keep on executing well. I hope LBT pops each and every cancerous tumor in the Intel culture and blacklists such people from ever being hired at Intel again.

I don't know if it's how they do things in that part of the world or if it's just you but your "hiring" statements are always bonkers. Aren't you the one who said something like "bring back Jim Keller and fire everyone who disagrees with him"?
 
Jul 27, 2020
23,643
16,599
146
I don't know if it's how they do things in that part of the world or if it's just you but your "hiring" statements are always bonkers. Aren't you the one who said something like "bring back Jim Keller and fire everyone who disagrees with him"?
It's just my opinion that people dragging down an effort should be let go immediately. All that debt of keeping people who are better off doing something else stifles innovation and breeds a whole lot of resentment as the productive employees start wondering, hey, this guy is so happy all the time yet he hardly contributes whereas I'm always worried about my responsibilities and pushing myself harder to accomplish more. You can be happy or you can be responsible. They are almost mutually exclusive. The weight of a responsibility makes you serious. You hardly have time to smile.

I see it in my workplace all the time. The lamest employees who dress the best and smile the most are also some of the least productive and since I am the one processing everyone's payroll, I know that the management gets fooled by these dishonest people's presentation skills into giving them money they don't really deserve. It happens EVERYWHERE.
 
Reactions: Thunder 57

johnsonwax

Member
Jun 27, 2024
140
227
76
Bench-cores

Is it true that E-cores serve the purpose of inflating MT Benchmark numbers?
The purpose of E-cores is to relieve the CPU from having to do weak compute/high latency tasks, usually system tasks, so that the performance cores can concentrate on the high compute tasks.

Stop acting like the point of a CPU is to run benchmarks. That's not the point. And if it is becoming the point (Goodhart's law), the plot has been lost and those who engage in it will eventually fail. Don't invite that to happen.
 
Reactions: Nothingness

johnsonwax

Member
Jun 27, 2024
140
227
76
Intel always does stupid things in my opinion.

Where is the 10 core Alder Lake with only Golden Cove cores with AVX-512 enabled?
Where is the 10 core Arrow Lake with only Lion Cove cores with AVX-512 enabled?
Where is the 32 or 48 core Bench Monster comprised of only Gracemont cores and now Skymont cores?
Where is the Skymont based Core i3-N405?

Intel is mostly always late. It's like they are paying peanuts to the engineers who can make things happen while it seems AMD respects their talent a lot more which is why they keep on executing well. I hope LBT pops each and every cancerous tumor in the Intel culture and blacklists such people from ever being hired at Intel again.
This is why Apple broke from the CPUs makers. Intel isn't doing stupid things, they are doing the only thing they can do which is to design to SKU. They need to nerf half their product line in one way or another to justify you moving upmarket to more expensive CPUs. It's why Thunderbolt wasn't available on low-end SKUs for people who needed high throughput but low compute, because it forced them to move upmarket and buy things they didn't need, because Intel needed the revenue. AMD does the same thing. So does Broadcom, and Qualcomm, etc. Every component maker does the same thing. Hell, Qualcomm forced you to buy compute even when you only wanted radio. They cranked it so hard that they forced you to buy a whole category you didn't want.

Apple meanwhile wanted to have Thunderbolt across their product line, and not just on the high end stuff because it simplified their marketing and they didn't see connectivity as a thing to nerf for differentiation, and Intel wouldn't make that product because it would collapse some of their higher end sales.

Intel is doing what always happens to component makers. Apple isn't a component maker but does the same thing with their end products - the low end iPhones get nerfed in critical ways to upsell you to the better ones - but Apple wanted to control how that manifested rather than having Intel control it. Welcome to how markets work.

The answers to the questions above likely have more to do with the marketing team than the engineers. So long as Intel has 70% of the market to AMDs 30%, the marketing team is succeeding and nothing needs to change.
 

LightningDust

Member
Sep 3, 2024
43
76
51
Intel always does stupid things in my opinion.

Where is the 10 core Alder Lake with only Golden Cove cores with AVX-512 enabled?
Where is the 10 core Arrow Lake with only Lion Cove cores with AVX-512 enabled?
Where is the 32 or 48 core Bench Monster comprised of only Gracemont cores and now Skymont cores?
Where is the Skymont based Core i3-N405?

Intel is mostly always late. It's like they are paying peanuts to the engineers who can make things happen while it seems AMD respects their talent a lot more which is why they keep on executing well. I hope LBT pops each and every cancerous tumor in the Intel culture and blacklists such people from ever being hired at Intel again.

Work in a large semiconductor company for a while. Please.
 
Reactions: Nothingness
Jul 27, 2020
23,643
16,599
146
Work in a large semiconductor company for a while. Please.
Please elaborate on what's wrong with my post instead of suggesting something improbable. Besides, at my age, I wouldn't want to work anywhere now unless I have the authority to fire anyone I want

But don't worry. I look at all sides when making a decision so I don't think I would be a mean dictator.
 

LightningDust

Member
Sep 3, 2024
43
76
51
Please elaborate on what's wrong with my post instead of suggesting something improbable. Besides, at my age, I wouldn't want to work anywhere now unless I have the authority to fire anyone I want

But don't worry. I look at all sides when making a decision so I don't think I would be a mean dictator.

You seem to be under the impression that Intel's problems are as simple as "hire the right engineers, fire the wrong ones." They aren't. The presence or absence of a design engineer, or ten design engineers, would have made no real difference in Intel's trajectory over the last decade.

Intel's problems now are the same ones they've been for a long time - that Intel is a horrifying chimera of dozens of different silos, many of which are externally acquired, and making a monster that big dance is a difficult assignment. The amount of expertise in Intel is insane, just like it is (or was until recently) in IBM, which has many of the same problems. But large companies do not change direction easily or quickly, and cannot be made to do so by hiring a few of the right people or issuing memos with bold statements. These issues, historically, have been compounded by the dependence on Intel's internal lithography capabilities and a tendency to use a lot of custom physical design in a given product rather than a high-reuse SoC flow like everyone else in the industry. Intel has been taking steps on both of these, but semiconductor products take a long time to develop, and the ripples are going to be affecting their roadmap for a long time - and there's very little that can be done about that, because every generation of complex IP is building on the generations prior.

To be honest, I think Intel probably should have split up into several pieces a decade or more ago.
 
Mar 11, 2004
23,444
5,845
146
You seem to be under the impression that Intel's problems are as simple as "hire the right engineers, fire the wrong ones." They aren't. The presence or absence of a design engineer, or ten design engineers, would have made no real difference in Intel's trajectory over the last decade.

Intel's problems now are the same ones they've been for a long time - that Intel is a horrifying chimera of dozens of different silos, many of which are externally acquired, and making a monster that big dance is a difficult assignment. The amount of expertise in Intel is insane, just like it is (or was until recently) in IBM, which has many of the same problems. But large companies do not change direction easily or quickly, and cannot be made to do so by hiring a few of the right people or issuing memos with bold statements. These issues, historically, have been compounded by the dependence on Intel's internal lithography capabilities and a tendency to use a lot of custom physical design in a given product rather than a high-reuse SoC flow like everyone else in the industry. Intel has been taking steps on both of these, but semiconductor products take a long time to develop, and the ripples are going to be affecting their roadmap for a long time - and there's very little that can be done about that, because every generation of complex IP is building on the generations prior.

To be honest, I think Intel probably should have split up into several pieces a decade or more ago.

I agree that Intel's issues aren't really the engineers (although it might actually be as they've let a lot of people go). Its always been management, with their engineers occasionally giving them such advantage that their management can then turn it into an unfairly dominant market position. When things were good, Intel management wasted it on bizarre stuff (contra-payments to try and shove Atom down our throats in mobile, their weird fashion era where they were paying clothing designers to slap LEDs into stuff, think they spent like a billion on that). When things aren't good they don't know how to operate so they make rash decisions (spin off NAND, ditch 3D XPoint which was one of the few things they took a chance on that had potential; repeatedly trying GPU designs with what should have been glaring issuses from the get go).

They did that (ditched many areas including ones that were profitable even in order to focus back on x86) and it didn't really help them. Their processes stopped keeping them competitive and they've been in perpetual management shakeup. Splitting foundries off likely would've just been GF 2.0, if not worse, which would've exacerbated the entire situation. If Intel hadn't been able to flood the market with chips when pretty much no one else could, they'd have been out of business long before now.
 
Reactions: igor_kavinski
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |