Question Geekbench 6 released and calibrated against Core i7-12700

Page 10 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,668
14,676
136
If you had written that 10 years ago you could substitute 80% get by with 1 core just fine, 19% 2 cores and 1% 4 cores. People are buying/using what the market is selling, but for the average person who hardly ever runs anything other than a browser - accessing their email in it, opening PDFs or word docs in it, etc. - 1 core would still be plenty if they still sold single core PCs with one of Intel or AMD's current high end cores with decent cache and clock speed. The needs of that average person haven't changed in 10 years, and the performance requirement to run Windows haven't either. Browsers are more heavyweight but that's mainly reflected in how much RAM they use rather than how much they load the CPU.

People are spoiled nowadays, and when they open up task manager think every active thread needs its own core or they will be bogged down lol
Or it could look like this:
 

Mopetar

Diamond Member
Jan 31, 2011
7,961
6,312
136
If they wanted to refocus their measures they should have split MT into two separate categories. Having something called MT that fails to scale past mid-range desktop CPUs in many cases when GB5 didn't have this issue feels like a regression. It's playing with semantics rather than giving users more useful information.

It's New Coke all over again.
 

hemedans

Senior member
Jan 31, 2015
201
100
116
Normal people do not need 16 cores. Period.

Only the enthusiasts, such people who you may found on a forum like this one, may find use for such a larger number of cores.

If you take all the people on the planet who own a computer - laptop or a pc:

  • 80% could get by with 4 cores just fine
  • 19% could require for what they do on the computer 8 cores, but they could survive with 4 cores as well, occasionally feeling restricted
  • 1% could require for what they do on the computer more than 8 cores, but they could survive with 8 cores as well, occasionally feeling restricted
Normal people benefit from increasing performance per core, and they need just a few of powerful cores.

All those chips with large number of cores which are now available on the consumer platforms will be on average massively underutilised.

Those few enthused consumers are lucky that they can purchase such powerful CPUs on the cheapest consumer platform.

Trying to apply a mindset of a miniscule sliver of all computer users to everybody is foolish.
Newer Games like Cyberpunk run terrible on 4 cores 8 thread Cpu, 6 cores 12 thread have Become minimum nowadays. Even in Site like Steams more than 50% of Gamers use 6 core or more.
 
Reactions: lightmanek

lopri

Elite Member
Jul 27, 2002
13,211
597
126
I have yet to decide what to think of their new methodology. lol. On one hand, what they say about multi-threading on a typical desktop is probably right. As the thread count increases their effectiveness decreases, especially after 8-cores or so. Other bottlenecks kick in, and programs themselves are not well coded to handle that many cores.

On the other hand, I do think a benchmark should show a full potential of hardware, and let users form their own preferences. Trying to influence it via opaque scoring system feels shady.
 

Doug S

Platinum Member
Feb 8, 2020
2,370
3,787
136
I can't wait for the latest issue of Car and Driver where they rank the world's fastest cars. I hear it's a 10 way tie, they all got up to 75 on the interstate and obviously no one needs to be breaking the speed limit or anything.


Flat out speed is essentially what Geekbench ST is measuring. MT is measuring something more akin to towing capacity, where some people might whine about wanting to see numbers for "what's the heaviest load I can tow over level ground". Even though in the real world people are towing stuff up (and down!) mountains, in rain or ice, in 110F or -40F, etc. and the numbers they would arrive it for max towing capacity are far lower than the idealized numbers that would be reported by Geekbench 5 MT equivalent for towing capacity.

The GB5 MT test for towing would probably be something like tying together trailers loaded with scrap iron and towing them down an airport runway. Look this F350 can tow 200,000 lbs, yay! And after people criticize it as unrealistic someone would claim that test approximates their real world use of towing a few trailers full of soybeans down a flat farm road to the grain elevator.
 

MrTeal

Diamond Member
Dec 7, 2003
3,580
1,725
136
Flat out speed is essentially what Geekbench ST is measuring. MT is measuring something more akin to towing capacity, where some people might whine about wanting to see numbers for "what's the heaviest load I can tow over level ground". Even though in the real world people are towing stuff up (and down!) mountains, in rain or ice, in 110F or -40F, etc. and the numbers they would arrive it for max towing capacity are far lower than the idealized numbers that would be reported by Geekbench 5 MT equivalent for towing capacity.

The GB5 MT test for towing would probably be something like tying together trailers loaded with scrap iron and towing them down an airport runway. Look this F350 can tow 200,000 lbs, yay! And after people criticize it as unrealistic someone would claim that test approximates their real world use of towing a few trailers full of soybeans down a flat farm road to the grain elevator.
Ehhh.. Honestly I think my joking backhanded analogy is more accurate than that, but there's not really a good car analogy that I can think of that really applies. If you really wanted to compare it to hauling, it'd be like saying one company has 8 of it's best trucks and 16 smaller trucks, while another company has 96 fast trucks all of the same design. If you just need a single load hauled, how fast can you get it from point A to point B?
Conversely if you have a big multi truck load, which company will get it there fastest?

Geekbench is like a Yelp review of trucking companies. Most people are moving a few pallets or their house. Some might have a lot more to move and need a few trucks to do it. The factory farm that has 10000 tons of soybeans to bring to the terminal can fill a couple runs of even the second company's fleet of 96 super B's, and doesn't give a *beep* about Yelp reviews.
 
Reactions: moinmoin

coercitiv

Diamond Member
Jan 24, 2014
6,280
12,297
136
The factory farm that has 10000 tons of soybeans to bring to the terminal can fill a couple runs of even the second company's fleet of 96 super B's, and doesn't give a *beep* about Yelp reviews.
So why is it that some people in this thread insist that Yelp should provide regular people with content for factory logistics? Somehow the "average Joe truck analogy" quickly derailed into a fleet management spreadsheet simulation.

I've seen people argue that GB should present a set of two MT scores, one of them being throughput oriented. If GeekBench truly wants to focus on consumer needs, keeping the prosumer benchmark score around can only hurt them, as the average person would not know how to interpret the results. There is a form of reassuring simplicity in presenting a simple set of scores, and most consumers need that as long as it's competent and honest.

We often complain that there is no perfect "all-round" benchmark, and we need specialized benchmarks to cover various use cases. Turns out Primate Labs may have actually embraced this reality and built a benchmark aimed at the common folk, something that will speak to them about what products scale with their needs. Personally I value the new benchmark more than I did the older ones, even with a narrower scope. We have better and more consistent "pro" MT benchmarks anyway.
 

lopri

Elite Member
Jul 27, 2002
13,211
597
126
If you had written that 10 years ago you could substitute 80% get by with 1 core just fine, 19% 2 cores and 1% 4 cores. People are buying/using what the market is selling, but for the average person who hardly ever runs anything other than a browser - accessing their email in it, opening PDFs or word docs in it, etc. - 1 core would still be plenty if they still sold single core PCs with one of Intel or AMD's current high end cores with decent cache and clock speed. The needs of that average person haven't changed in 10 years, and the performance requirement to run Windows haven't either. Browsers are more heavyweight but that's mainly reflected in how much RAM they use rather than how much they load the CPU.

People are spoiled nowadays, and when they open up task manager think every active thread needs its own core or they will be bogged down lol
I disagree. There are a lot going on even a commoner's desktop these days, be that Windows or Mac. A single-core CPU will get bogged down simply downloading a game installer. Athlon X2 and Pentium D came out 20 years ago. We are way past the single-core era.
 

Doug S

Platinum Member
Feb 8, 2020
2,370
3,787
136
I disagree. There are a lot going on even a commoner's desktop these days, be that Windows or Mac. A single-core CPU will get bogged down simply downloading a game installer. Athlon X2 and Pentium D came out 20 years ago. We are way past the single-core era.

Try disabling all but one core on your high end PC and you will find you can't tell any difference in performance for such common tasks as downloading an installer. It would have to be coded so poorly it would set some sort of a record for that to overload a modern core.
 
Reactions: Nothingness
Jul 27, 2020
17,155
11,022
106
We are talking about Windows after all. Just like running Windows on an HDD is sheer torture in terms of waiting for everything to load before you can use your PC without lag, it is possible that getting to the CPU idle state after boot with a single core may take way longer than anyone is prepared to wait. May check it out on my 12700K this weekend.
 

poke01

Senior member
Mar 8, 2022
997
1,091
106
We are talking about Windows after all. Just like running Windows on an HDD is sheer torture in terms of waiting for everything to load before you can use your PC without lag, it is possible that getting to the CPU idle state after boot with a single core may take way longer than anyone is prepared to wait. May check it out on my 12700K this weekend.
A Zen 4/Raptor P core/ M2 P cores are very powerful. Even a single one these is fine for installing games and web browsing.

Windows is good these days unlike the XP/Vista era.
 

hemedans

Senior member
Jan 31, 2015
201
100
116
A Zen 4/Raptor P core/ M2 P cores are very powerful. Even a single one these is fine for installing games and web browsing.

Windows is good these days unlike the XP/Vista era.
we had Pentium G3258 dual core unlocked Cpu during haswell era, some Games couldnt be installed because it had only 2 thread. Far Cry 4 is one of those games which need minimum 4 thread, unless you patch installer.
 

Just Benching

Banned
Sep 3, 2022
307
156
76
LOL at one modern core still being enough for most users. Windows update alone will crush your dream. Try it and report back, you will be disappoint.
Ι used a 2/4 CPU (G6900 or something? Don't remember) while I was waiting for RMA, I can't say it was dreadful, but it was a pretty mediocre experience. Your everyday user can definitely feel the difference with an actual 4c CPU even for normal everyday tasks like browsing excel word and the likes. Over 4c diminishing returns start kicking in for those simple tasks.
 
Reactions: lightmanek
Jul 27, 2020
17,155
11,022
106
Boot time results with my 12700K (from when "Restarting..." message disappears to WinLogin screen):

12C/20T: 24 secs

After login, CPU utilization was under 10%. AMD process (long name) took 5 to 7% CPU.

1C/2T: 22 secs

After login, CPU utilization was 100% for about a minute or two. AMD process (long name) took about 56% CPU.

1C/1T: Took 3 to 4 minutes to get to login screen. After entering password, several minutes to get to desktop. When desktop appeared, the taskbar is blank. No start menu to click. Hard reset. Tried to go into BIOS too get back to full cores. Screen went black. Hard reset didn't work multiple times. Hard shutdown. This time it booted faster and logged in and Start menu button appeared. Right clicking the Start button took some time for the menu to appear. Clicked Task Manager at least two times. It never loaded in the 5 minutes of my patience threshold.

So yeah, have fun running your 13900KS with 1C/1T.

Hope a 7950X/7950X3D user will report their own Win11 experience with 1C/1T.
 

Nothingness

Platinum Member
Jul 3, 2013
2,553
988
136
Boot is a very specific load: many independent tasks are being launched at the same time so more cores will help. But I guess there's a point of diminishing return, so testing would be needed with more or less cores (my old 4770K so 4C/8T started Windows 7 [yeah I know ] in basically no time and was immediately usable). I tried to find such a study, but failed miserably.
 

moinmoin

Diamond Member
Jun 1, 2017
4,988
7,758
136
It's possible Win7 can work fine with 1C/1T.
"Work fine". Even back in the day the change from 1 to 2 threads was significant. I'd say the change in basic responsiveness is best compared to the switch from HDD to SSD, and SATA to NVMe (edit: retracting the latter part, that's more comparable to adding even more threads).
 
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/    |