Jump to content
HWBOT Community Forums

CPU benchmarks applicable for global points - 2019


Which CPU benchmarks should get global points?  

129 members have voted

  1. 1. Which CPU benchmarks should still get global points in 2019?


This poll is closed to new votes


Recommended Posts

10 hours ago, havli said:

So... when people want to bench legacy HW, say Radeon 9800 Pro, they are supposed to run timespy on it or what? 9_9 Also rest assured - everything can be be hacked with some effort, even the hwbot API equipped benchmarks.

 

As to the x265 killing CPUs - you know what they say... no pain, no gain. :P If you are afraid, probably running something easier would be the solution, for example superPI 1M. B| On more serious note - if your CPUs are dying, probably you are pushing it too hard. Every SW using AVX is going to be more heavy on CPU that legacy (non-AVX) benchmarks. But X265 really isnt that heavy, there are many applications that are even worse... much worse in terms of power consumption and heat output. Also when CPU is X265 stable, it is still very far from general stability for every day use... my daily rig 8700k could do 1 hour long 4k 4x overkill easily but crashed during simple handbrake video conversion or gaming.

I was being Polite... Avx is harder than legacy benches thanks so much for teaching me /s

I'm curious which bench that people run is higher Power consumption and heat output? Do you have some data on this?  

Actually I think x265 is the highest load, the next highest load is when you open then x265.exe wrapper and you see 800w being used on an 18 core. 

Pushing too hard is not even the issue, the chips will even finish the session for hours until then next day wont boot dead cord or dead imc. 

There is one bench that is a massive killer of hw and it's yours. 

 

6 hours ago, Mr.Scott said:

x265 is the ultimate benchmark. What's a gold or WR worth to you? You're in total control of your own destiny.

Easy for you to say you don't bench anything new :P

 

  • Thanks 1
Link to comment
Share on other sites

Sure - Aida 64 Ray Tracing benchmark, Y-Cruncher perhaps (that one makes very heavy use of AVX2 and even AVX-512 on Skylake-X). And then of course various stability testing applications like OCCT and such... while not a benchmark, this is still SW that performs some sort of calculation and CPU should be able so survive it.

On 8700k x265 power consumption wasn't really that much higher than R15, the difference is it takes much longer to complete. For example when R15 or X265 were consuming 180 W (CPU only), Aida 64 RT easily yould hit 200+.
You mention 18-Core... We all know SKL-X is extremely power hungry and considering it takes some time to finish x265 4k - maybe even LN2 can't keep it "cool enough" the whole run and after some time some irreversible changes or damage happens on the silicon level or perhaps at the connection of the die to package. Most likely due to extreme current and heat output / local overheating. Just guessing here - but something like this is possible.


Every CPU is designed to run at specified power level... and obviously when you run it several times as high, then some damage might occur. Some CPUs just degrade and as it turns out some can die completely if you overload them too much. It is not the SW fault, it is simply running CPU far beyond safe levels... and something can fail.

For instance have you ever seen someone complaining 3DMark killed their voltmodded GPU? No, it simply can hapen when pushing HW to the limits.

Oh - and the loading screen, do you know what kind of load is used there? It is very funny in fact - I just spawn number of threads that matches the CPU... and run empty loops in them. So it is in fact doing nothing... and the only purpose of this is to determine more or less correct multi core turbo clock.


If you don't like running x265, noone is forcing you to use it. Perhaps it is HW killer, but as long it is not killing CPUs on default voltage / clock... I don't see it as a problem. And even so, there is warning written in the manual and on my website too - which states clearly "use at your own risk". Btw - only the java based wrapper is my work, the encoder itself is actually an opensource project and is used in many popular applications. So if you are afraid of X265 Bench, I suggest not to do HEVC video encoding in other SW either, the risk of HW dying is the same.

  • Thanks 1
Link to comment
Share on other sites

  • Members

The issue is the way it kills hw, I've never heard of benches that run no problem, you can save scores etc, then shut down after the session, but when you try to use the CPU again it's dead. That seems off to me, but maybe it happens with other benches as well? Killing a CPU whilst benching is fair enough, them dying after benching whilst off seems wrong. 

Edited by GeorgeStorm
Link to comment
Share on other sites

As we can see from pool, x265 will probably keep it's gp. I like this bench, never killed any hardware on it, even when pushing my 7700k at 1.8v i killed it, but i'm pretty sure i was using less vcc, pll etc... voltages than top overclockers.

But i know x265 can really kill much more cpus than other benchs, so, as Scotty said, the limit will be our goal, and as havli said, it's our own risk.

Edited by Casanova
Link to comment
Share on other sites

6 hours ago, havli said:

Sure - Aida 64 Ray Tracing benchmark, Y-Cruncher perhaps (that one makes very heavy use of AVX2 and even AVX-512 on Skylake-X). And then of course various stability testing applications like OCCT and such... while not a benchmark, this is still SW that performs some sort of calculation and CPU should be able so survive it.

On 8700k x265 power consumption wasn't really that much higher than R15, the difference is it takes much longer to complete. For example when R15 or X265 were consuming 180 W (CPU only), Aida 64 RT easily yould hit 200+.
You mention 18-Core... We all know SKL-X is extremely power hungry and considering it takes some time to finish x265 4k - maybe even LN2 can't keep it "cool enough" the whole run and after some time some irreversible changes or damage happens on the silicon level or perhaps at the connection of the die to package. Most likely due to extreme current and heat output / local overheating. Just guessing here - but something like this is possible.


Every CPU is designed to run at specified power level... and obviously when you run it several times as high, then some damage might occur. Some CPUs just degrade and as it turns out some can die completely if you overload them too much. It is not the SW fault, it is simply running CPU far beyond safe levels... and something can fail.

For instance have you ever seen someone complaining 3DMark killed their voltmodded GPU? No, it simply can hapen when pushing HW to the limits.

Oh - and the loading screen, do you know what kind of load is used there? It is very funny in fact - I just spawn number of threads that matches the CPU... and run empty loops in them. So it is in fact doing nothing... and the only purpose of this is to determine more or less correct multi core turbo clock.


If you don't like running x265, noone is forcing you to use it. Perhaps it is HW killer, but as long it is not killing CPUs on default voltage / clock... I don't see it as a problem. And even so, there is warning written in the manual and on my website too - which states clearly "use at your own risk". Btw - only the java based wrapper is my work, the encoder itself is actually an opensource project and is used in many popular applications. So if you are afraid of X265 Bench, I suggest not to do HEVC video encoding in other SW either, the risk of HW dying is the same.

I said benchmark that people actually run :P not sure of the last time someone ran OCCT for fun on ln2

And yes x265 is no harder than r15 ffs have a look man 

x265.PNG.d4f7a55209694398701615994404070b.PNG

 


The defense that its safe to use at stock.....you are on the most popular overclocking ranking website? 

Im done discussing it. Just bothers me you seem proud that it kills chips rubs me the wrong way completely.

Edited by Splave
  • Like 2
Link to comment
Share on other sites

Like I said - my 8700k draw 180W in R15, much like in X265. Obviously There is a huge difference in time needed to complete the benchmark which is the key.
R15 on these multicore monsters is just a few seconds long suicide run. Not much stability is needed for this and not much cooling power either (the total amount of heat generated during short R15 run is rather low). On the other hand x265 is much longer and therefore it is possible insufficient cooling is the problem here and stability on high clocks too. Even LN2 may not be strong enough for this much power hungry CPU.

CPUs are designed to run at specified condition... and when those conditions are met, it should any SW for unlimited amount of time. If x265 were causing issues at default, it would be a problem. But any OC is not guaranteed by anyone and how overclocked CPU runs this or that benchmark can be completely unprediclable. And as it turns out, it is.


I'm not proud about your dead CPU... in fact (sorry to say it hard like this) I don't care at all about other people's HW. I only care about my own, because it was my money spend on it. If your CPU died, then it is unfortunate of course - but it was you who overclocked it and you pressed the start button knowing this could happen, so it is entirely your responsibility. Otherwise you could blame pretty much anyone for this accident. Will you blame Microsoft for selling you Windows... because Windows allowed you to run the benchmark? Will you blame your ISP for letting you join HWBOT and download x265...? Because without Internet connection you'd have never run it. So why blame x265? It is just a SW that do nothing on its own.

You are not the only one having HW killed during benchmarking, it is part of this hobby. For example once my MB VRM blew up when running wprime... so I just said blah, tough luck, next time I'll be more carefull, time to move on. No forum post about how evil wprime is and how points should be removed, etc.

Not much more to say about this matter. Everyone should evaluate how risky is to run their HW at the limit and then decide whether the score / fun during benching / points gained is worth the price of potentially dead HW or not. If it is, then by all means push for the WR... if it isn't then you can always play it safe using more conservative voltage/clocks and minimize the risk of killing stuff to almost zero.

 


 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

  • Administrators

@havliApart from the topic to run x265 or not, you want to tell me that a non avx benchmark like cb15 draws same wattage like an avx benchmark like x265? And the heat generated even at running it 25 seconds maybe like cb15, which is much higher, falls from the sky? This is like telling people prime with avx is same like prime without avx enabled, and I wonder why I need a lot more vcore for same frequency... Just tested it again with 7600k, cinebench 15 takes at least same time like 1080p on the system but power drawn is clearly higher on x265.

  • Thanks 1
Link to comment
Share on other sites

49 minutes ago, Splave said:

Will continue using all but your benchmark in confidence. Have fun

Good, for everyone I think.

 

42 minutes ago, GeorgeStorm said:

@havli can you explain why it kills them in the way it does? Killing something whilst running a bench is something everyone is aware of, chips dying after being shut down without issues having run x265 make little sense to anyone, and that is the issue.

I can only speculate. But most likely it is related to the extreme current and heat stress In the CPU silicon itself (I'd say less likely). Or (IMO more likely) the connection between silicon and package. The silicon itself is attached to the package (which is similar to regular PCB) using thousands of little metalic balls.. looks similar to any BGA chip. Of course this is much smaller, more dense and the metal for sure isn't just ordinary solder. Still, I can imagine, those balls can be damaged either by extreme temperature change of the silicon above (from -190C to +100C in few seconds). Or some of them may partially melt due to high current (the power pins). If this theory is true, then cracks in the connections may develop while running stress test and while it still contacts when everything is hot and powered on... after you turn the system off it can break because of thermal shrinking, warping, etc. Also the fact SKL-X is using something like two layer interposer - not directly attached die to the base package (like LGA 1151) might do this even worse.

The other possibility is something "burns" or degrades in the silicon itself and while the CPU continues to run, on another cold start it perkaps fails some kind of self test and doesn't post.

33 minutes ago, websmile said:

@havliApart from the topic to run x265 or not, you want to tell me that a non avx benchmark like cb15 draws same wattage like an avx benchmark like x265? And the heat generated even at running it 25 seconds maybe like cb15, which is much higher, falls from the sky? This is like telling people prime with avx is same like prime without avx enabled, and I wonder why I need a lot more vcore for same frequency... Just tested it again with 7600k, cinebench 15 takes at least same time like 1080p on the system but power drawn is clearly higher on x265.

That is what I measured directly on the CPU 8pin (12V). Btw - prime AVX vs non-AVX is different from R15 and X265. Prime AVX is designed to make heavy use of it, while x265 is not using AVX that much - that is the reason btw why Ryzen is not that much slower than CL evethough it has much weaker AVX implementation..

Edited by havli
  • Thanks 1
Link to comment
Share on other sites

Well, if someone is suspecting that there is a killer bug in x265 benchmark, than i think the better way to test this is to pick the same videos that are in x265 and encode then using some software like handbrake during a extreme overclock session, pushing it the same way x265.

If after some extreme oc sessions no hardware is killed then there is a x265 killer bug problem. But if hardware is killed while encoding using handbrake we will know that it is not a problem related to the benchmark, but to the encoding process itself. Anyway, it looks a crazy suggestion, but would be a non questionable proof.

Edited by Casanova
Link to comment
Share on other sites

  • Administrators

a 2700x ryzen az 5,6ghz can beat a 7820x and 9900k at same frequency on Cinebench15, often ryzen is even more efficient on this - a 4,8ghz 9900k or 7820k beats the same 5,6ghz ryzen 2700x easily at x265 1080p.... I do not criticze the benchmark itself, but your argumentation about avx not being that important and load being same as cinebench. Every test I made was different, everyone I talk to about this tells this is different and x265 draws a lot more power on Intel new systems than cb15, needs more volts for same frequency and put higher load. On the 2700x vs 7820x or 9900k comparison, you can look up results at database here. I also have to say this problem is lot worse on 4K than 1080p, but this doesn´t change things

  • Like 1
Link to comment
Share on other sites

Considering Ryzen theoretical AVX performance is more or less half compared to 9900K, I really think it performs well enough in x265.

Anyway x265 is what it is and people should either accept it or bench something else if they dont like it. I think it is not such a bad idea to actually have something really CPU demanding that can work on any CPU you can think of (both old an new) and can take advantage of modern CPU instructions too. Recent x265 encoder builds even use AVX-512 to some extent... but this is not the case with HWBOT version and I have no plans to change it. Also interesting fact - at some point there was plan to add 8k preset :ph34r: But ultimately that plan was scrapped and when reading this thread I'm glad it ended like that.

I'm really curious whether in the future we will also have similar "killer CPU" complains about new GPUPI (which wil make use of AVX too, IIRC) or Y-Cruncher if it becomes more popular and get attention of extreme overclockers.

Link to comment
Share on other sites

I believe x265 killed my golden caseking 6700k (#19: 99X1135 HWBOT Prime @ 1.90 V: 6627 MHz* 6th best out of 700 Price: 1000 €)
ok part of benching. Then it killed a decent 7700k (L639Fxxx), frustrating but it happens.
Most recently it killed my very first golden cpu (4770k L314B105) while benching for Country Cup.
No more x265 4k for me on any cpu I care about

Edited by GtiJason
  • Thanks 1
Link to comment
Share on other sites

4 hours ago, GtiJason said:

I believe x265 killed my golden caseking 6700k (#19: 99X1135 HWBOT Prime @ 1.90 V: 6627 MHz* 6th best out of 700 Price: 1000 €)
ok part of benching. Then it killed a golden 7700k (L639Fxxx), frustrating but it happens.
Most recently it killed my very first golden cpu (4770k L314B105) while benching for Country Cup.
No more x265 4k for me on any cpu I care about

I still think it's not a benchmark problem, encoding to h265 using avx is very hardware stressful.

Still hoping to see someone here testing it the way i mentioned above.

Link to comment
Share on other sites

I agree, encoding h265 with avx is very stressful and may simply be the issue here. I can't say if there is a problem or not, not an expert on the subject nor will I pretend to be. Just wanted to include my experience with the subject. The 6700k finished benching fine and was shut down/ warmed up by me, did not realize it was dead until a couple weeks later. The 4770k died the day after heavy 4k testing while checking my temp delta between cpu and pot. Booted into the OS at pot temp of around zeroC / 1.35v core, cpu was dead 30 seconds later

Edited by GtiJason
Link to comment
Share on other sites

Well...

1) I totally agree with what @websmile ( Michael ) wrote and I have made the same observations myself.

2) My very good I7-7740X running 6ghz on SS died EXACTLY as @Splave ( Allen ) described, running one single session of H265 - 4K - 2X overkill mode ( 2 runs ).

    Shut down the system normally, next day I got that wonderful 00 - cpu dead. And it is definitely not 18 or 16 cores where I suppose absolute hell will break loose.

So, my 16 and 18 core cpu's will never see H265 4K or even H265 at all, however many points it gives.

And since @havli prefers to put things the hard way ( which I like ), I am not prepared nor will I ever be again to lose a cpu ( price irrelevant ) on a benchmark.

Unless a sponsor suddenly falls in love with me and starts shipping boxes of FREE cpu's to me.

No XTU and no X265 ( for me at least )

GPUPI - 1B is very interesting though. I think it should get globals.

 

Edited by Fasttrack
  • Thanks 1
Link to comment
Share on other sites

Hey, x265 encoding is just like your daily converting video, it just works like that, if your cpu died on LN2 most probably because too much current flow on your cpu, and yes cpu doesnt like too much current for long time and probably for short time if you cpu cant handle it anymore. Using it on Ryzen 7 1700X LN2 like 50x run for competition and it mostly survived. It just risk like when you do CPU-Z suicide clock & memory clock with high vccsa & vccio. Just that. And you can disable AVX from OS if you want to...

Edited by speed.fastest
  • Thanks 1
Link to comment
Share on other sites

Drop the buggy XTU please , drop x265 4k and the others can stay :) . Agree with my fellow OCers x265 4k killed one of my cpus recently :( , meaning no more xtu265 4k runs for me on any hardware. It died exact same way as Allen and George's cpu , bench hard on subsero one day , after running x265 4k a few too many times and i got the score i was looking for happy me  , next day cpu won't post ?  , nada , nothing , dead as fish out of water ... 

Edited by chispy
  • Thanks 2
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...