• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Share your CPUZ Benchmarks!

As cpuz 1.87 is out, let's rejuvenate this thread with the new avx2 benchmark results, should be interesting

11142018-005554.jpg
 
Don't... buy that CPU. You had enough expensive fun with your LGA2066 suited quad-core.
Not expensive at all : ) - I was waiting for the skylake-x refresh though I'm tempted by the 7900x / 7980xe and I'm holding out to see overclocking results vs first gen. I couldn't care less about STIM or no STIM I've got my delid stuff still anyways.
 
New best scores on everything but mt ratio. Fairly sure if I tried a little harder I could get mt score at least a little closer to the nearly 20 point single thread jump I saw.

2l2YSR2.jpg
eerqnGn.jpg


Previous best score in 1.85.0. Followed by an update now that I noticed there is an AVX bench.

5FHh286.png
3OAzb2F.jpg
 
Last edited:
Lost 2 or 3 points on the single bench moving from 1803 to 1809 win 10. Anybody else?

Well within the margin of error. The benchmark is identical between versions 1.79 and 1.87.

---------

Soo ... I've been away, focusing on my degree for the past couple of months. I see there's a new version of CPU-Z, and a new benchmark accompanying it.

I've updated the graphs so that everyone who posted since my last visit has now been added. New separate graphs also exist for both CPU-Z 1.87 benchmarks. All graphs are available from the links within my signature.

In the meantime, I feel like I obtained a completely new system, even though I simply replaced a few parts. A new case, new power supply and a graphics card. The only thing left to get is an SSD. The old power supply was absolute garbage and I couldn't have gotten rid of it sooner. Suddenly, overclocks on this system aren't complete ass, although from what I've seen, the Richland APU overclocked better. It seems like 4.60 GHz is all I can get out of this, and there isn't much headroom with the BCLK now that I'm using a graphics card. The Richland APU played well with a 113 MHz BCLK, and a > 4,800 MHz core clock.

Here are my results for versions 1.86, 1.87 SSE2 and 1.87 AVX2.

CPU-Z 1.86 @ stock
186-stock.png

CPU-Z 1.86 @ 4.50 GHz
186-4500.png

CPU-Z 1.86 @ 4.60 GHz
186-4600.png

CPU-Z 1.87 SSE2 @ stock
187-sse2-stock.png

CPU-Z 1.87 SSE2 @ 4.50 GHz
187-sse2-4500.png

CPU-Z 1.87 SSE2 @ 4.60 GHz
187-sse2-4600.png

CPU-Z 1.87 AVX2 @ stock
187-avx2-stock.png

CPU-Z 1.87 AVX2 @ 4.50 GHz
187-avx2-4500.png

CPU-Z 1.87 AVX2 @ 4.60 GHz
187-avx2-4600.png
 
5930@4428
CPU-Z 1.81
CPU-Z Bench.PNG


4428 on 1.87
number2.PNG
 
Added. :)

I'm going to play around with some older versions, just to see how Godavari compares to Richland, given they are almost identical frequency-wise. Usually Richland comes out on top though, for some reason.
 
Got my hands in new Asus laptop with a i5-8250U cpu in it.

UmCHLUU.jpg
 
Wellp... I tried older versions of CPU-Z to compare against the other results.

No matter how hard I try, I can't get past 4.60 GHz with a much better power supply. The "trash" power supply handled the A8-6600K at 4.80 GHz, and even briefly at 5.00 GHz under light usage. I guess Kaveri/Godavari just doesn't overclock as well, which is a shame.

Granted, the results show that Steamroller doesn't need those clocks to perform better, but when clocked identically, it's actually quite hard for Steamroller to give better results than Piledriver in this specific SSE2 test. Interesting.

5,235 pts is the highest I can achieve* without tinkering with the BCLK, which puts the A10-7890K roughly 8% behind a Core i5-2500K at stock.

cpuz-178-4600-2000.png


*
CPU-Z 1.78
CPU @ 4.60 GHz (+ 0.048 V offset; can use stock voltage for 4.50 GHz)
NB @ 2,000 MHz (+ 0.024 V offset; can use stock voltage for 1,900 MHz)
My RAM is not overclockable (Kingston ValueRAM), hence the move to an RX 460.
 
Last edited:
cpuz8gb.jpg


perhaps bottlenecker is right at saying RAM bottlenecks cpus?
 
perhaps bottlenecker is right at saying RAM bottlenecks cpus?

The gains will drop off after 8 GiB of RAM.

It's just because 4 GiB is basically pushing it with 64-bit Windows and an integrated GPU.

---

Merry Christmas and Happy Holidays to you all. :)
 
gotta update ur specs... also what size rad are you using?:

This is just a bench rig. No need to update specs atm.

280mm with 2x140mm fans pushing through rad. Ambient was 24°C
 
Previous post in here with CPUZ 1.87 produced a single core number within easy tolerance of what I'm still seeing.

Multicore... 91 points better.

Which is to say I figured out how to maximize this bench score on my PC and could probably get that up to a round 100 with repeated attempts. o_O

Benchmark.JPG
CPU.JPG
Mainboard.JPG
SPD.JPG
 
Core i9 9900k @ 5,7 GHz 8C/8T , also runs stable @ 5,4GHz 8C/16T. Was hoping to breach 700 points single thread, but coulnd get her to run stable at 5,8Ghz


1548148484239.png
1548149638664.png
1548149686697.png
 
Last edited:
Back
Top