ZoSo

AIDA64 Benchmark Results Comparison

54 posts in this topic

1333 mhz RAM :unsure:

I'd say 1066MHz since DDR2 above that I recall being tricky compatibility wise and judging from your bandwidth scores. I'll change that to 1066MHz for now.

You can verify when your PC boots up as it lists the speed next to the amount installed - press Pause/Break to temporarily halt loading - or go into your BIOS and take a look. No good checking with AIDA64 as it doesn't show the overclocked speed only the stock speed, I did check :)

Edit: It's your FSB that is 1333MHz plus your score and is why I have your RAM down for DDR2.

0

Share this post


Link to post
Share on other sites

Ok my rig:

CPU @ 4.6 GHz

Memory Read: 20276 MB/s

Memory Write: 22923 MB/s

Memory Copy: 23690 MB/s

Memory Latency: 41.9 ns

CPU Queen 58128

CPU PhotoWorxx: 15405 MPikseli/s

CPU ZLib: 378.9 MB/s

CPU AES: 17946 MB/s

CPU Hash:3572 MB/s

FPU VP8: 7228

FPU Julia: 15295

FPU Mandel: 7820

FPU SinJulia: 6185

0

Share this post


Link to post
Share on other sites

WhyCry, I have updated the table above and soon ZoSo will update the initial post with the table.

In my general observations I do find it odd that your 3770K @ 4.6GHz is only scoring FPU Julia 15,295 against a non-HT 3570K @4.2GHz from ZoSo. Same below average score from FPU Mandel. It could simply be a bottleneck of your CPU caused by your RAM if you take a look at omegaa's scores and also the scaling of FPU Vs your laptop RAM @1333MHz. You could drop the clocks down to 4.2GHz and run AIDA64 again to see if there was much change across the tests to test that theory.

0

Share this post


Link to post
Share on other sites

Most of WhyCry's scores make no sense to me, how am I beating his scores with a lower clock speed and a 3570K vs his 3770K? Something isn't right there.

0

Share this post


Link to post
Share on other sites

Most of WhyCry's scores make no sense to me, how am I beating his scores with a lower clock speed and a 3570K vs his 3770K? Something isn't right there.

Yip, the 3770K is a very good chip WhyCry's should spank everyone in the table. I thought someone could pinpoint the oddity here but every time I think it's the RAM. If not I'd check the latest BIOS, my BIOS update fixed loads in the past including IB 1155 support, USB/SATA/PCIe performance improvements, and in December 2012 it sorted UEFI GTX 680 stability. Lastly, make sure you weren't running another activate window tabbed in the background or using the mouse/keyboard and run it again.

0

Share this post


Link to post
Share on other sites

are my bench good for my pc spec ?? honestly i cannot understand how to read bench

0

Share this post


Link to post
Share on other sites

Well I have no idea how this benchmark works. All I know is that I have not turned off the background apps, just went through the whole test without any preparation.

@ 4.2 Ghz

Memory Read: 19776 MB/s

Memory Write: 21276 MB/s

Memory Copy: 23214 MB/s

Memory Latency: 42.5 ns

CPU Queen 52954

CPU PhotoWorxx: 14045 MPixel/s

CPU ZLib: 342.9 MB/s

CPU AES: 16391 MB/s

CPU Hash: 3261 MB/s

FPU VP8: 6942

FPU Julia: 13995

FPU Mandel: 7143

FPU SinJulia: 5639

@ 4.6 Ghz

Memory Read: 20308 MB/s

Memory Write: 22939 MB/s

Memory Copy: 23759 MB/s

Memory Latency: 42.3 ns

CPU Queen: 57827

CPU PhotoWorxx: 12577 MPixel/s (second run 15393 MPixel/s)

CPU ZLib: 377.0 MB/s

CPU AES: 17948 MB/s

CPU Hash: 3570 MB/s

FPU VP8: 6747 (second run 7561)

FPU Julia: 15328

FPU Mandel: 7824

FPU SinJulia: 6178

I've noticed that during some tests my CPU is not utilized at full speed (PhotoWorxx for instance), it jumps from 1.6 to 4.6 GHz. Poor results, which were slower at 4.6 than 4.2, were much better after I run them seperately. Don't ask me why there's such a difference, because I have no idea :)

0

Share this post


Link to post
Share on other sites

Thanks for re-running that WhyCry. It suggest that your CPU bandwidth is being limited by your 1866MHz RAM as that has approached it's saturation point - can you borrow some faster RAM ie 2133MHz/2400MHz from a friend etc as I'm sure that will boost figures? Also when your being bottlenecked and overclock the CPU past what it's capable of with your components (think GPU overclocking as that does the same) in return you experience negative results and a degraded performance.

RE: PhotoWorxx: relying on CPU monitoring isn't showing a true real-time representation of what is going on as whole system can't work fast enough under heavy CPU loads to update it frequently. Normally you would see it tip (on my G19) from idle and a few seconds later it would hit 100% and stay static on that until it completes.

I know from reading into this that SpeedStep (jumping from idle of 1.6GHz) does affect the performance, normally knocks about 15% off so try it again with SpeedStep off as I did and post back. Have you updated your BIOS recently?

0

Share this post


Link to post
Share on other sites

are my bench good for my pc spec ?? honestly i cannot understand how to read bench

They are in line with your balanced setup and your memory bandwidth scores are excellent!

If your normally up for it: You could better my results if your willing to push 3770K a bit more to 4.6GHz - assuming you have excellent cooling or de-lidded the 3770K - should show +10% general improvement across the board as that would equal 5GHz on my 2600K based on the additional instructions per clock. I know I'm taking this literally, you could up it to 4.3GHz to equal my benchmarked 4.7GHz scores.

0

Share this post


Link to post
Share on other sites

Well I was using an Asrock Extreme Tuner app, so I don't think there is a better app for that :)

If my ram is crap here, then I might start thinking about upgrading this component in a next weeks. I won't be able to find a person with a better memory.

As for the clock, I tried going to 4.7 and 4.8 but there was a huge gap in cpu overvoltage with this additional 100 MHz. Can't remember precisely but it was something like 0.2V, it was a huge change just to reach 4.7 clock, so I went back to 4.6.

0

Share this post


Link to post
Share on other sites

Why don't you overclock in the BIOS, using 3rd party is nice but I havent heard good things about that Tuner app - any particular reason you need to use it? Get back to good old BIOS tweaking, and turn off SpeedStep cos you can't do that in utilities.

What is your vcore for 4.6Ghz, I'd guess 1.30v on an average chip? To reach 4.7/4.8GHz may take it up to 1.35v because I heard past 4.4GHz you need to start to whack the voltages up quite steeply to get any further. You can tweak BIOS voltage CPU PLL down 0.1v to 1.7v from default of 1.8v, will help drop temps a little too. Once booted, keep Coretemp open during testing to watch the vcore as that might jump about under load, simply decrease/increase again in BIOS to keep it inline with your target voltage. Also using LLC at 50-75% should help the vdroop, 50% for normal gaming use and 75% is what I use when I really push it (your motherboard may not have LLC but saying in case it does).

0

Share this post


Link to post
Share on other sites

You got me wrong, I was using the UEFI BIOS for overclocking, I just watched the clock speed tduring the benchmark using ASROCK Extreme turner. I used the Ivy Bridge ocing guide at overclock.net, which was so detailed that I'm sure I've somehow reached the limit of my CPU.

What's more important, I rarely use very intense overclocking configuration, because I need a stable system, this is a priority for me. I do not need additional 5% performance if I'm going to lose stability. If I was a gamer only, that wouldn't be a problem, but since I do a lot more work on my pc I cannot afford to lose data.

0

Share this post


Link to post
Share on other sites

i will try 4.3 when my RMA psu come back , i dont want to go over 1.2v . i hear somewhere , going over this voltage will reduce the life of the cpu.

0

Share this post


Link to post
Share on other sites

Totally understand WhyCry you require stability. As that is how you run your system everyday then that is how it shall be benched. I tell you what, it's a hell of a work PC ;) Now get a proper gaming rig :P:D

i will try 4.3 when my RMA psu come back , i dont want to go over 1.2v . i hear somewhere , going over this voltage will reduce the life of the cpu.

It's higher than that as I previously posted. This from Hardwarecanuks: "Ivy Bridge overclocks very well, generally better than mainstream Sandy Bridge at every voltage point up to 1.35V. Above that? We don't know. We don't know because we were told by an always reliable third-party that stress testing above 1.35V is a risk with IVB due to potential degradation, but we also don't know because we were scared that it would cause a house fire..."

source

0

Share this post


Link to post
Share on other sites

if you can help me out with my overclocking it will be great. once i have everything back and ready-to-go , i will let you know :)

0

Share this post


Link to post
Share on other sites

if you can help me out with my overclocking it will be great. once i have everything back and ready-to-go , i will let you know :)

It depends on your existing temperatures as that will be main factor for IB because the Intel TIM is like concrete and very messy. Beyond your 4.2GHz I'd leave as they are because then it gets tricky unless your up for it. Once de-lidded you can look at getting to 4.5GHz with 1.30vcore and other voltages at defaults otherwise without de-lidding the highest you'll get is 4.4Ghz because of the stupid voltage jump required and resulting heat. I'd wait until Haswell and forget about going for more because its not worth the time. On the 3570K I ordered for my Folding PC yesterday - only part so far - I'll be playing around with it and ONLY because of the temperatures.

0

Share this post


Link to post
Share on other sites

i can already post some screenshot that i have uploaded , or do i need to open new thread ?

0

Share this post


Link to post
Share on other sites

New thread cos it's off-topic, thanks omegaa :)

0

Share this post


Link to post
Share on other sites

As long as WhyCry is passing his stress test like Intel Burn Test, then he should not have any degradation in performance from an OC. That is more true for GPUs than it is for CPUs in my experience.

As for his CPU downclocking during the tests, I've seen it happen and normally it has to do with background apps and/or thermal throttling. WC go check your BIOS and see what you have your thermal limit set to, I think the default in our BIOS is 80C, which for Ivy is too low. It should be set to 90C for Ivy, even if you don't get near that temperature, or even 80C.

Thread on Downclocking During Stressing: http://www.overclock.net/t/1216036/i5-2500k-downclocks-during-stress-test

0

Share this post


Link to post
Share on other sites

There are more AIDA64 benches I happen to come across from Vortez where they are reviewing a system from Overclockers in the UK with 3770K @4.6GHz and 1600MHz RAM. BUT they cocked up the data as you can see below - makes no sense - maybe someone with an account there should let them know! I used version 2.70.2200.

CPU Queen - 58,571 / WhyCry's - 58,128 / ZoSo's - 38,109 [this is correct]

PhotoWorxx - 63,845 / WhyCry's - 15,405 / ZoSo's - 17,133 [incorrect by miles]

FPU Mandel - 12,863 / WhyCry's - 7,820 / ZoSo's - 11,288 [this is correct]

AES - 428,685 / WhyCry's - 17,946 / ZoSo's - 13,995 [incorrect by light years!]

link and screenshot link if they ever change it. The rest of the benches were fine ;)

0

Share this post


Link to post
Share on other sites

I overclocked my Phenom II 965 to 4.0GHz @stock voltage(1.4V) and its preety slow compared to your benches :(

I did 30k in CPU Queen. I could bump voltage because it is crashing @4.1GHz, but my cheap cooler can't keep up.

Well...there is no point comparing 100$ CPU to 300$ CPU

Ram was way slower... my 1600MHz cl9 dualchannel @1700Mhz cl9 vs your tripplechannel is like night and day

My system is running at stock 3.4GHz with lowered voltage to 1.25V (occt stable). I don't need more because I don't have GPU yet :P

I just upgraded from athlon 64 x2 5000+ 2,6GHz and DDR2 800MHz and it was huge jump!

0

Share this post


Link to post
Share on other sites

cpmPhKX.jpg

 

AIDA64 2.85.2400 is available for download

Changelog:

  • Optimized 64-bit benchmarks for AMD A-Series Richland APU
  • Improved support for AMD Kabini, Temash and Kaveri APUs
  • Improved support for Intel Haswell and Valleyview processors
  • Preliminary support for AMD Beema APU
  • Preliminary support for Intel Avoton SoC, Crystal Well CPU, Haswell-E CPU, Haswell-ULT SoC, Ivy Bridge-E CPU, Lynx Point-LP PCH, Wellsburg PCH
  • Preliminary support for DDR4 memory modules and GDDR5 system memory
  • GPU details for AMD Radeon HD 8000 Series and nVIDIA GeForce GTX Titan


DOWNLOADS:

 

ewtGmty.png  AIDA64 Extreme Edition Trial version, self-installing EXE package

       2.85.2400 Mar 27, 2013 13.8 MB

dJ7lvCU.png?1

 

 

hO8M1NE.png  AIDA64 Business Edition Trial version, self-installing EXE package

       2.85.2400 Mar 27, 2013 13.4 MB

dJ7lvCU.png?1

0

Share this post


Link to post
Share on other sites

Thanks for the update Skr13 :) Though could you correct the formatting as on the 5th bullet point after the word "Preliminary" it cuts to the next line, same after the word "Haswell-E CPU". Guess that was a copy-paste issue ;)

Whoever buys one these new CPUs/APUs, a quick post of benchmark results here would be nice.

0

Share this post


Link to post
Share on other sites

Yeah, it was a copy-paste issue indeed ;)

0

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now