The PassMark site has recently introduced version 10 of its PerformanceTest benchmark, with the result being that Intel processors now utterly dominate the single-thread performance chart. The Intel Core i9-9900KS tops the list with a result of 3,039, while the highest AMD chip is down in 35th position with the Ryzen 3 4300U on 2,673.
https://www.notebookcheck.net/AMD-obliterated-from-PassMark-s-single-thread-performance-chart-as-Intel-claims-top-34-places-thanks-to-version-10-update.457506.0.html
Real world tests are what matters, says intel
Quote
Can't stop but wonder how convenient this update is for Intel...but the explanation shows 0 partiality toward intel,...maybe the used features more optimized by intel (AVX512,..) but that stops here.
Tbh, If I were to build a desktop, it's a no-brainer I'd pick AMD, not that obsolete 14nm Intel is still trying to shamefully market after years of delay on an already obsolete 10nm node.
PS5 And Xbox Series X will be massively produced....in 7nm EUV for Both GPU/CPU...and here one of the most powerful chipset manufacturers can't even do 10nm correctly.
They rested so many years on their achievements, made close to no efforts because of monopoly...but know AMD is back, finally and for modern day usage (gaming, decoding/encoding, heat/consumption,...) they are ahead of Intel
AFAIK Intel still does have a noticeable single core advantage by most measures, so this adjustment should be in line with most other benchmarks.
"Passmark ... attempts to be TRANSPARENT about the new tests..." FAKE! How about show us how much you are paid way before this effort started and how much you are paid after the effort started and the next 24 months!
FCC should start investigating both companies for antitrust laws!
Are we serious? Comparing and 127W cpu with a 15W CPU?
Passmark also made to their GPU testing skewed at NVIDIA's advantage
Intel does lead in single threaded performance. If core counts are similar then intel leads multi threaded performance too. But when AMD puts more cores (12+) in the high end then AMD start to have an advantage in multi core performance while still lagging in single threaded performance. So I am not sure why anyone is surprised seeing Intel leading the single threaded benchmarks.
Uhh... Clock speed and AVX-512 that made AMD's single threaded performance score lower?
MORE SEVERE SECURITY ISSUES TOO!!! ZOMBIELOAD MELTDOWN ANY ONE
Totally believable, I trust Passmark unconditionally. Was swayed by these new results and gonna go buy Intel now.
A totally credible result, not. We continue with AMD.
Passmark 10.0 now sponsored by Intel
Except you are not buying single thread CPU, check this out - AMD Threadripper is unbeatable so far. The new top of the line
Intel Core i9-10980XE - 32,561 just a position after
AMD Ryzen 9 3900X - 32,829.
AMD Ryzen 9 3950X - 39,417.
Check PassMark high end CPU chart.
Of course intel leads single core, but when the 3700x exceeds a 7700 in single core and it is nowhere to be found.. That's a little odd. And the gap isn't as big as it used to be. Another problem is that AVX512 is limited to HEDT as an instruction set, so... I don't see the point in including a desktop processor.
Intel aee cheatinbg.
Anyone ever running one of many programs, called cheatmachine u can speed up the cpu speed,
So lets take a eksempel thay easy to understand, u are running a commodore 64 emulator. The emulator lets u run the game x10 faster
But hard to play 10 times faster then it suppose to go.
So u running a old demo but 10 times faster
Now u start benchmark program in cheat machine, 3dmark. Running cpu speed 1,3x enouge to beat amd. But it is cheating.
And cheat machine are detected by 3dmark anti cheat system, but northing prevent to write there own "cheay machine"
To the lost herd: AVX 512 has nothing to do with that result, as 9900ks does not support it. Although Intel does have the single thread performance lead, we are not talking bulldozer level gaps. Actually we do have even AMD leading ST performance for select workloads, but why have passmark selected the least favorable approach to AMD?
Uhh... this isn't very credible or their testing things that are clearly skewed towards intel via clock speed settings.
I was using a 7700k since launch and my new 3900x completely decimates the thing.. Even in single threaded work loads, CPU-Zs variation of single threaded performance testing is over 50 points higher than my 7700k.. which is reflective of reports that the 3000 series has slightly higher IPC over Skylake's multiple revisions.. it's a 4.3 vs 4.4 with "turbo" without "manually overclocking". My applications and REAL world use also reflect this result.
Just look at the clock speeds tested.. This isn't a very credible and they're purposely placing CPU's higher and lower than they should be. 7700k has a 4.2 base speed sure, but the thing literally runs at 4.5Ghz in a single core work load.. 4.4 multi core turbo without touching anything in the bios assuming you're using at least a half way decent $20-30 cooler and not throttling. The 9000 series chips run at anywhere from 4.6-5.0 in the same situation.
I just find it hard to believe that a 3.8ghz base clock 3900x is significantly behind a 4.0ghz base clock 9900k? Like what the hell are you even testing in your proprietary software?
Quote from: Shepard on March 15, 2020, 19:49:55
To the lost herd: AVX 512 has nothing to do with that result, as 9900ks does not support it. Although Intel does have the single thread performance lead, we are not talking bulldozer level gaps. Actually we do have even AMD leading ST performance for select workloads, but why have passmark selected the least favorable approach to AMD?
Intel does have the single threaded lead, but its solely limited to clock speed and Turbo values. (You can also include potential OC headroom as intel's chip do much better in terms of max frequency all around.)
Ryzen 3000 has higher IPC clock for clock over Skylake, but Skylake on 14++++(do i need more plus'?) clocks much higher and thus has the advantage typically.
The results are obviously skewed in a way that makes intel look better than they actual are. Like.. I'm not disagreeing that a CPU like the 9900KS shouldn't be in the number one spot, but the testing and overall results make no sense.
Intel: synthetic benchmarks are worthless.
Passmark: releases new version.
Intel: synthetic benchmarks are all that matter.
Total, honest, we are not tied.
This site blows.
No.
This site.
No. Not us.
news site.
This item blows.
Not news.
No, we are like Trump.
We can say anything, and well , generate cash like a banshee.
Intell customer service service announcement.
Spend around $450 more and we can give you a 6% increase.
Well worth the cash.
Intel fan boys , well suck it.
Here is a , supposed , reliable, news source.
AND Published.
on an archive.
For everyone, too look back on.
This is only saved in an archive.
Sad times my friends.
Intell:
We are the the better.
This site just showed HOW much they need advertising . Do Not Trust these Journalists/ News site anymore.
They EVEN published the findings.
They did this !.!.!
And keep changing the posts to suit their needs.
Hey,. !
How can I get these dollars so my daughter doesn't die from cancer?
Fucking liars..
Block these so called journals. Money, hungry, pussys. From your life.
And let them come back too reply.
You are my only lifeline to this world. Please help me?
are you people literally retarded? like literally dropped on your head when little stupid?
you are comparing a desktop processor against a mobile U (extra efficient processor) you incredible goshddarn moerons!
just because passmark didn't test the 4000 version desktop processor doesn't mean intel is better.
This pathetic excuse for an article should be taken down and thrown in the trash.
Lrn2test you Intel fanboys.
Nuff said...
Single thread performance isn't even important for workloads any longer. UNIX and UNIX-like operating systems and environments have been using multithread workloads for years, and only Windows has finally decided to do the same.
Applications are taking advantage of multithread workload features of COUs to spread the workload and boost processing capabilities for better performance.
Single thread is only useful for testing purposes. It had no valid functional usage in applications other than a staunch developer refusing to get with the times.
Passmark isn't even relevant anymore. Geekbench and other tools have far better results and can really show what your CPU is able to do.
The intend of the author is very suspicious. A managing editor will put the comparison of a low power mobile cpu to desktop chip in the subtitle?
To catch more views or simply showing off his IQ?
This is just a stupid article...
Intel has very long history of cheating the benchmark...
And this result in 15 years damaging in customer money... crating monopoly and stay the last 5 years in 14nm
If you supporting this... you supporting monopoly and anti innovation
Quote from: Curious on March 16, 2020, 03:54:34
The intend of the author is very suspicious. A managing editor will put the comparison of a low power mobile cpu to desktop chip in the subtitle?
To catch more views or simply showing off his IQ?
The author seems to bias more to Intel. Check out his other article, where he compared the upcoming AMD's desktop 4000 series APU (4 core?) with the mobile 4000 series APU (which has much lower TDP, even though it has more cores).