News Intel Alder Lake i5-12400 Posts Impressive Benchmark

Status
Not open for further replies.

GaryM

Distinguished
Jun 17, 2008
16
1
18,525
What surprises me is how relatively unimpressive the Golden Cove performance is, compared to the Zen3 cores of my modest, stock-clocked AMD 5700G APU. I just ran the same CPU-Z benchmark, and my single-thread score was 624.4 (Alder Lake = 681.7), and my multi-thread score was 6288.4 (Alder Lake = 4983.8). Multiplying the 6-core Alder Lake score by 4/3 to level the playing field with my 8-core Zen3 APU, the Alder Lake extrapolated multi-thread score would be 6645.1. True enough, the new Intel scores are higher, but not by as much as I was expecting, given all the recent hype. And since the 5700G's integrated graphics performance is very unlikely to be matched by Alder Lake, I'm becoming even more comfortable with my decision not to wait for Alder Lake's release.
 
Well, thats a better CBR20 single core result than my R5 3600 at 485 pts. Only problem as with all this leaked benchmarks theres no info about the system (motherboard, power limits, RAM) and the cooling system.

Nevertheless it seems a decent start.

We will find out when the real product arrive to reviewers.
 

spongiemaster

Honorable
Dec 12, 2019
2,364
1,350
13,560
Oh come on, it's way more likely that it was simply Microsoft and Amd both being idiots than some conspiracy.
No, it was absolutely sabotage on MS's part. You know what's even worse? Microsoft refused to give anyone who worked for AMD access to beta versions of Windows 11 and made AMD purchase a retail copy from their local Egghead Software on release day. That's why the fix wasn't announced until after release day.
 

evdjj3j

Distinguished
Aug 4, 2017
371
396
19,060
What surprises me is how relatively unimpressive the Golden Cove performance is, compared to the Zen3 cores of my modest, stock-clocked AMD 5700G APU. I just ran the same CPU-Z benchmark, and my single-thread score was 624.4 (Alder Lake = 681.7), and my multi-thread score was 6288.4 (Alder Lake = 4983.8). Multiplying the 6-core Alder Lake score by 4/3 to level the playing field with my 8-core Zen3 APU, the Alder Lake extrapolated multi-thread score would be 6645.1. True enough, the new Intel scores are higher, but not by as much as I was expecting, given all the recent hype. And since the 5700G's integrated graphics performance is very unlikely to be matched by Alder Lake, I'm becoming even more comfortable with my decision not to wait for Alder Lake's release.

Keep in mind the Intel i5 x400 models are generally sub $200 processors.
 
What surprises me is how relatively unimpressive the Golden Cove performance is, compared to the Zen3 cores of my modest, stock-clocked AMD 5700G APU. I just ran the same CPU-Z benchmark, and my single-thread score was 624.4 (Alder Lake = 681.7), and my multi-thread score was 6288.4 (Alder Lake = 4983.8). Multiplying the 6-core Alder Lake score by 4/3 to level the playing field with my 8-core Zen3 APU, the Alder Lake extrapolated multi-thread score would be 6645.1. True enough, the new Intel scores are higher, but not by as much as I was expecting, given all the recent hype. And since the 5700G's integrated graphics performance is very unlikely to be matched by Alder Lake, I'm becoming even more comfortable with my decision not to wait for Alder Lake's release.
So what real-world use case is there for the CPU-Z benchmark?!
People using cinebench is already bad enough since pretty much nobody is using 3d rendering on desktop but cpu-z is even worse than that.
No, it was absolutely sabotage on MS's part. You know what's even worse? Microsoft refused to give anyone who worked for AMD access to beta versions of Windows 11 and made AMD purchase a retail copy from their local Egghead Software on release day. That's why the fix wasn't announced until after release day.
Lol, do you have a link for that? This sounds so ridiculous that I need to read about it.
We all had access to beta versions but AMD didn't?
 

wtfmang

Honorable
Nov 25, 2016
21
0
10,510
so this is like the 11400 with a tad better power usage.

if they jack up the price of this, then it would make sense to get the 11400 instead, since the 12400 just sounds like a repackaged 11400
 

PCWarrior

Distinguished
May 20, 2013
216
101
18,770
so this is like the 11400 with a tad better power usage.

if they jack up the price of this, then it would make sense to get the 11400 instead, since the 12400 just sounds like a repackaged 11400
How is this like a repackaged 11400? The 11400 has a single thread score in CPU-Z of 583. The 12400 has a score of 681.7 – that is 17% higher. And in multithreaded the 12400 scores 4983.8 versus 4341 of the 11400 – that is 15% higher. The single threaded score is enough to even place the 12400 at the top of the chart in guru3D (see here) – above even the 5950X. It also places it above the 5600X and 11600K in MT. It also is just 1% slower in MT than the octacore 2700X, and 11% slower than the 9900K. Not bad for an under $200 chip. And this is not even the final performance for the 12400 either. With final silicon, final microcode, better RAM, more updated Windows 10/11, etc you can easily have at least another 5% improvement.

I wonder if the tests were performed on Windows 11 where Intel and Microsoft worked together to have a "bug" giving AMD an automatic 5-15% reduction in performance.
Even if AMD gets a hit on Windows 11, it doesn’t get it on Windows 10. What is being compared here is the best score of AMD (i.e. in Windows 10) versus the score of Intel (either in Windows 10 or Windows 11). So no, AMD is not losing because of an OS-related regression in its performance but because Intel has made progress and leapfrogged them in performance fair and square. On top of that due to the incompetence of AMD to collaborate effectively with Microsoft they also see regressions in their performance in Windows 11 making the gap against Intel even larger than what is being reported here.
 
  • Like
Reactions: panathas and Why_Me

TJ Hooker

Titan
Ambassador
What surprises me is how relatively unimpressive the Golden Cove performance is, compared to the Zen3 cores of my modest, stock-clocked AMD 5700G APU. I just ran the same CPU-Z benchmark, and my single-thread score was 624.4 (Alder Lake = 681.7), and my multi-thread score was 6288.4 (Alder Lake = 4983.8).
Are you sure your CPU is at stock? Because your Cinebench R20 scores are 10% higher than what reviews of the 5700G are showing.

Edit: derp, I misread which benchmark was being discussed.
 
Last edited:

GaryM

Distinguished
Jun 17, 2008
16
1
18,525
Are you sure your CPU is at stock? Because your Cinebench R20 scores are ~10% higher than what reviews of the 5700G are showing.
Yes, it's the stock frequency (and standard on-demand boost). I was only quoting my CPU-Z benchmark scores as a direct comparison to a couple of the data points the article's author used. I don't have Cinebench installed. I have elevated my FCLK frequency to 1800 mhz for synchronized transfers with my 3600 mhz memory, but that shouldn't affect pure CPU benchmarks. The editor that wrote the headline was more impressed with Alder Lake's benchmarks than I am. They seem much more like normal generational progress.
 
  • Like
Reactions: TJ Hooker

NP

Distinguished
Jan 8, 2015
74
15
18,535
Keep in mind the Intel i5 x400 models are generally sub $200 processors.

I think this is exactly something that we should not keep in mind. The fact that i5 x(x)400 are generally sub-$200 processors means nothing in those instances where i5 x(x)400 processors are over $200. Like i5 12400 most certainly will be.
 
Like i5 12400 most certainly will be.
What makes you think that?!
The 12400 won't have any e-cores so why would the CPU be more expensive?

The last three generations of 1x400 where the exact same price down to the dollar.
Even with the 10th gen adding hyperthreading and the 11th gen adding a completely new iGPU. Even with all the shortages and added shipping costs in the last two years.
https://ark.intel.com/content/www/us/en/ark/compare.html?productIds=134898,199271,212270
 

NP

Distinguished
Jan 8, 2015
74
15
18,535
What makes you think that?!
The 12400 won't have any e-cores so why would the CPU be more expensive?

The last three generations of 1x400 where the exact same price down to the dollar.
Even with the 10th gen adding hyperthreading and the 11th gen adding a completely new iGPU. Even with all the shortages and added shipping costs in the last two years.
https://ark.intel.com/content/www/us/en/ark/compare.html?productIds=134898,199271,212270

It will be over 200 because it is 10nm. Last generations have been on a mature/more mature/super mature nodes, which makes all the difference.

And unlike you claim, there has not been three generations of 1x400 processors.
 
It will be over 200 because it is 10nm. Last generations have been on a mature/more mature/super mature nodes, which makes all the difference.

And unlike you claim, there has not been three generations of 1x400 processors.
Sure, so it's xx400 how does that change the prices? The last 3 gens are still the exact same price.
10nm is mature by now that's why they release it now to desktop, if it still were expensive we would wait for it for even longer.
 

NP

Distinguished
Jan 8, 2015
74
15
18,535
Sure, so it's xx400 how does that change the prices? The last 3 gens are still the exact same price.
10nm is mature by now that's why they release it now to desktop, if it still were expensive we would wait for it for even longer.
Soon you will see people buying 12400 for $200+. Let's discuss more then.
 
Status
Not open for further replies.