Well I do agree there should be no avx512 in geekbench, if that is in there it shouldnt be turned on. They need to wait until desktop apps start using it.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Intel Core i7-11700K beats Ryzen 9 5950X by 8% in Geekbench 5 single-core benchmark
- Thread starter Dave2150
- Start date
Well I do agree there should be no avx512 in geekbench, if that is in there it shouldnt be turned on. They need to wait until desktop apps start using it.
Caporegime
- Joined
- 8 Jan 2004
- Posts
- 32,579
- Location
- Rutland
And there's the problem. You got there!
Geekbench paints Intel (specifically the 11th gen) favourable in a way that doesn't reflect anyone's real world usage.
Soldato
- Joined
- 31 May 2009
- Posts
- 21,457
shillbench is worse as they've assigned cpu speed ratings to the processor groups that can't even be achieved by the processors in the group when you run the test.
It isn't even a different weighting now, it is complete and utter fabrication.
It isn't even a different weighting now, it is complete and utter fabrication.
Soldato
- Joined
- 31 May 2009
- Posts
- 21,457
Okay take a trip into the wonderful world of weird. The user benchmark site under CPU, listed by average bench rates all the new Intel chips above the old. And indeed well above all ryzen.
If you actually search user results however only 1 in 10 or less of chips tested say 11900k f.ex. achieve the result as stated in 'average bench.
It seems full on fabrication.
If you actually search user results however only 1 in 10 or less of chips tested say 11900k f.ex. achieve the result as stated in 'average bench.
It seems full on fabrication.
Right so they started by weighing 4 thread performance very heavily over anything more than that and Single threaded way above that, what was it about 98% to 4 threads or less?
Because Intel single core boost was that much higher it pushed 4 core Intel CPU's above 16 core AMD CPU's in the results charts.
Now that AMD have the Mutithreaded performance AND the single threaded performance they are weighing the results specifically to Intel's best results, those results that are in the top 10% and probably heavily overclocked.
Well, in their latest rant they complained that Mainstream Reviewers didn't use Core Enhancement or other such Motherboard specific overclocking features on Intel systems, they likened it to putting the handbrake on in a race.
You can smell the hate for AMD with these people and the raging anger that AMD are on top, they stink of it.
A third party LLC shell company that is wholly owned by a subsidiary of Intel, would be my guess.
Caporegime
- Joined
- 8 Jan 2004
- Posts
- 32,579
- Location
- Rutland
This is there 11900K review:
"The i9-11900K is the fastest CPU in Intel’s 11th Gen Rocket Lake-S lineup which brings higher IPC (early samples indicate +19%) and 50% stronger integrated graphics. There are also 500 series chipset improvements including: 20 PCIe4 CPU lanes (up from 16) and USB 3.2 Gen 2x2 (20 Gbps up from 10 Gbps). Rocket Lake’s IPC uplift translates to around a 10% faster Effective Speed than both Intel's 10th Gen and AMD’s 5000 series. Despite Intel’s performance lead, AMD continues to outsell Intel. Given the scale of Intel's operation, it’s inexplicable that their marketing remains so neglected. Little effort is made to counter widespread disinformation such as: “it uses too much electricity” or the classic: “it needs more cores”. Intel’s marketing samples are routinely distributed to reviewers that appear better incentivized to bury Intel's products rather than review them. Not enabling XMP or only testing with BIOS power limits enabled is akin to leaving the handbrake on during a race. Mind-numbing “scientific” and rendering benchmarks are presented as gospel. Different games, mostly unplayed by real users, are cherry picked for each “review”. When it's convenient, canned game benchmarks are chosen such as Ulletical’s CSGO which runs at nearly double the in-game fps. Credible benchmark data, which necessarily includes replicable video footage from popular games, is the exception rather than the rule. At every release, AMD’s marketers coordinate narratives to ensure another feast of blue blubber. Nonetheless, towards the end of 2021, Intel’s Golden Cove is due to offer an additional 20-30% performance increase. At that time, with a net 30-40% performance lead, Intel will probably regain significant market share despite AMD's class-leading marketing. In the meantime, most PC gamers need look no further than the 11400F."
Amazeballs.
Its hard to imagine even a fanboy would put that much effort and money into updating the site and paying Google for search rankings. I fully expect that it's either directly owned by Intel or is run by one of the top management at Intel
It's also worth noting that unlike Cinebench, no self respecting reviewers incorporate Geekbench into their review suites. If Geekbench was such a wonderful, realistic, important and accurate measurement of performance, reviewers would be falling over themselves to use it and yet...no one does
When amd launched the 3990x, which is 64 Zen 2 at 3.something ghz, they sold a bunch to RED, a company that develops and sells 8k recording equipment for Hollywood and they fell in love with it, said it speeded up their workflow a huge amount and were grateful for the time savings
So someone is claiming Intel achieved a 34% performance increase without a process improvement change?
That and it contradicting every performance review makes me skeptical. I was excited by earlier Ryzens but found the bulk of reviewers showed the 7920x had a single core lead so went with that, this time the 5000 series was clearly ahead.
Guessing Intel will be leaking marketshare until Kaby lake and likely beyond for a bit.