|
|
Post by stormchaser on Jul 9, 2020 11:10:06 GMT -5
Gauging interest in a CPUz mini comp. I was thinking we could have two leaders. One for dual socket systems and the other for single socket. Multi core test only Very simple, run the CPU "bench" and post that screenshot here. If we get enough interest I will start an excel spreadsheet for the leaderboards. I would also like to factor in multi thread ratio. So the two data sets will be A)Your multi thread result B)Your multi thread ratio How these results would tie into the actual placings on the leaderboard is yet to be determined, but I do have a couple ideas I'm mulling over at the moment... Any suggestions are welcome as well. Obviously not the most reliable benchmark in the world but it will give us a rough approximation of mutli core system performance. The format would go something like this... CPU...…….2x Intel Xeon E5 2673s @ 3.6GHz Memory...64GB Octal channel DDR3 @1866mhz Cooling...Stock (liquid) GPU...…..RX 5700 XT  |
|
|
|
Post by osmiumoc on Jul 9, 2020 12:24:24 GMT -5
I'm always in for quick fun.  Just take CPU-Z numbers with a grain of salt, it can bug out occasionally out of nowhere: valid.x86.fr/ipk1h5See the validation link, same single thread but much higher multi. Same system, no background tasks. EDIT: Just adding this validation so you see its up to date valid.x86.fr/9kn9fpCPU...…….1x Intel 10980XE Memory...32GB DDR4 Quad-Channel 4000MHz C15-15-15-34 1T Cooling... Water/ambient |
|
|
|
Post by Mr.Scott on Jul 9, 2020 12:35:43 GMT -5
SR-2/ x2 5675's  |
|
|
|
Post by stormchaser on Jul 10, 2020 1:13:26 GMT -5
Here is my other Z820 with the two E5 2696 12c/24t CPUs for a total of 24c/48t  Slight improvement in single core performance here... not counting it though. I thought perhaps it would be to cluttered to be chasing down three separate data sets.  |
|
|
|
Post by ShrimpBrime on Jul 10, 2020 20:47:27 GMT -5
Never mind the multi-core. This is set up for 3D right now. I can run all 12, but I don't want to make any changes atm.  |
|
|
|
Post by stormchaser on Jul 11, 2020 3:27:52 GMT -5
Thats a nice chip you got there shrimp is it a new processor to you? Cant recall seeing it before.
Very similar specs vs my 6 core 1151 rig as well.
Both are 6 core chips, based on similar architecture. only I don't have hyperthreading with the 9600KF
I exactly match your RAM, both are 4200mhz
This has interesting implications, if you are up for it, I'd like to go toe to toe with you.
And see how the rigs compare.
|
|
|
|
Post by ShrimpBrime on Jul 11, 2020 9:07:44 GMT -5
Sorta new. Had it a few months now. Main goal for this rig is 3D benching submissions.
I'd like to think the cpu is pretty average.
What kind of toe to toe bench are you thinking of??
|
|
|
|
Post by ShrimpBrime on Jul 11, 2020 10:50:43 GMT -5
Also, forgot to mention I can run that memory frequency at 16-16-16 timings.
|
|
|
|
Post by stormchaser on Jul 11, 2020 12:19:37 GMT -5
Also, forgot to mention I can run that memory frequency at 16-16-16 timings. I was thinking maybe start with a full aidia64 benchmark report? And we can pick some other benchmarks, perhaps a few CB runs or something. That will give us some decent baseline results. Mainly a CPU comparison. It's been a while since I checked the memory spec. I think its CL18 My other thread is actually about reviving my 9600KF rig. Currently waiting on watercooling parts before I can actually get the system back online. Expect about a week and I'll follow-up with you once I have all the kinks worked out and the system operational. I ordered the best 120 mm rad money can buy... its a a pure copper 80 mm thick rad, going to have two Noctua fans on it for push/pull however, I'll tell you right now I may be slightly Mhz limited by it but I do have air-conditioning here. I can get it down to about 62 or 63*F. Also NBs are close. You are at 4800 and Im at 4900. |
|
|
|
Post by ShrimpBrime on Jul 11, 2020 13:34:53 GMT -5
Also, forgot to mention I can run that memory frequency at 16-16-16 timings. I was thinking maybe start with a full aidia64 benchmark report? And we can pick some other benchmarks, perhaps a few CB runs or something. That will give us some decent baseline results. Mainly a CPU comparison. It's been a while since I checked the memory spec. I think its CL18 My other thread is actually about reviving my 9600KF rig. Currently waiting on watercooling parts before I can actually get the system back online. Expect about a week and I'll follow-up with you once I have all the kinks worked out and the system operational. I ordered the best 120 mm rad money can buy... its a a pure copper 80 mm thick rad, going to have two Noctua fans on it for push/pull however, I'll tell you right now I may be slightly Mhz limited by it but I do have air-conditioning here. I can get it down to about 62 or 63*F. Also NBs are close. You are at 4800 and Im at 4900. Ok. What is AIDa64 full benchmark? I usually only use that program for memory and cache benchmark. We can do restrictions if you like. 5ghz cpu 5ghz cache. Mine may do about 5.2ghz cache frequency. Memory at 4000mhz restriction . Or higher.... My water delta is mid 40f I think. Cooling isn't an issue. No rads installed. Not even a pump. Just some tubes and a water block. I'll play as long as we don't get super serious. I have the Team Cup to think about as well. |
|
|
|
Post by stormchaser on Jul 11, 2020 21:42:49 GMT -5
I was thinking maybe start with a full aidia64 benchmark report? And we can pick some other benchmarks, perhaps a few CB runs or something. That will give us some decent baseline results. Mainly a CPU comparison. It's been a while since I checked the memory spec. I think its CL18 My other thread is actually about reviving my 9600KF rig. Currently waiting on watercooling parts before I can actually get the system back online. Expect about a week and I'll follow-up with you once I have all the kinks worked out and the system operational. I ordered the best 120 mm rad money can buy... its a a pure copper 80 mm thick rad, going to have two Noctua fans on it for push/pull however, I'll tell you right now I may be slightly Mhz limited by it but I do have air-conditioning here. I can get it down to about 62 or 63*F. Also NBs are close. You are at 4800 and Im at 4900. Ok. What is AIDa64 full benchmark? I usually only use that program for memory and cache benchmark. We can do restrictions if you like. 5ghz cpu 5ghz cache. Mine may do about 5.2ghz cache frequency. Memory at 4000mhz restriction . Or higher.... My water delta is mid 40f I think. Cooling isn't an issue. No rads installed. Not even a pump. Just some tubes and a water block. I'll play as long as we don't get super serious. I have the Team Cup to think about as well. Good. I promise we don't have to go overboard, more of curiosity than anything else to see how the two CPUs measure up / and what impact hyper threading has from one to the other, etc. Here is how you run the full report in aida64: 1)at the top click report 2)then click report wizard 3)Under Report profiles select "Benchmark Pages" 4)It will take 5-8 minutes to run through the entire battery of benchmark tests 5)Save the Benchmark report when completed as an HTM or HTML file and then we can attach here |
|
|
|
Post by ShrimpBrime on Jul 11, 2020 23:01:27 GMT -5
Ok. What is AIDa64 full benchmark? I usually only use that program for memory and cache benchmark. We can do restrictions if you like. 5ghz cpu 5ghz cache. Mine may do about 5.2ghz cache frequency. Memory at 4000mhz restriction . Or higher.... My water delta is mid 40f I think. Cooling isn't an issue. No rads installed. Not even a pump. Just some tubes and a water block. I'll play as long as we don't get super serious. I have the Team Cup to think about as well. Good. I promise we don't have to go overboard, more of curiosity than anything else to see how the two CPUs measure up / and what impact hyper threading has from one to the other, etc. Here is how you run the full report in aida64: 1)at the top click report 2)then click report wizard 3)Under Report profiles select "Benchmark Pages" 4)It will take 5-8 minutes to run through the entire battery of benchmark tests 5)Save the Benchmark report when completed as an HTM or HTML file and then we can attach here Sounds easy enough. So for comparison, are we going to run the exact same clocks, or we gonna race? |
|
|
|
Post by ShrimpBrime on Jul 11, 2020 23:54:34 GMT -5
Here's 5.2ghz 6 cores no HT, memory at 4200mhz CL-19 and Cache at 4.9ghz. It's a little too much, temps are about 80c, FPU blue screen. Not looking to battle those out, so lowered the Cpu frequency. 100mhz won't be the end of the world in difference I don't think. Looks like some benchmarks are skewed. Is that from lack of delay between benches? If I run this individually, they seem to differ? AIDA64 Benchmark.htm (339.04 KB) |
|
|
|
Post by stormchaser on Jul 13, 2020 21:24:18 GMT -5
Here's 5.2ghz 6 cores no HT, memory at 4200mhz CL-19 and Cache at 4.9ghz. It's a little too much, temps are about 80c, FPU blue screen. Not looking to battle those out, so lowered the Cpu frequency. 100mhz won't be the end of the world in difference I don't think. Looks like some benchmarks are skewed. Is that from lack of delay between benches? If I run this individually, they seem to differ? Hmmm... you should get a relatively similar result regardless of whether the benchmark is done individually or as part of the complete benchmark report, at least from the testing I've done this seems to be the case. I get a variation of 1-2% between runs for the most part. Its possible you had some background tasks going for one of the individual runs? Perhaps that's why the numbers to match? |
|
|
|
Post by ShrimpBrime on Jul 13, 2020 23:45:49 GMT -5
Here's 5.2ghz 6 cores no HT, memory at 4200mhz CL-19 and Cache at 4.9ghz. It's a little too much, temps are about 80c, FPU blue screen. Not looking to battle those out, so lowered the Cpu frequency. 100mhz won't be the end of the world in difference I don't think. Looks like some benchmarks are skewed. Is that from lack of delay between benches? If I run this individually, they seem to differ? Hmmm... you should get a relatively similar result regardless of whether the benchmark is done individually or as part of the complete benchmark report, at least from the testing I've done this seems to be the case. I get a variation of 1-2% between runs for the most part. Its possible you had some background tasks going for one of the individual runs? Perhaps that's why the numbers to match? Very possible. I started the OC from scratch with very crude settings. Probably why 5.3ghz wasn't stable. Gotta pull the GPU so the rig is going to be down. I need to better arrange my benching set up. Also need to get a de lid done. So I hope the results are comparable. |
|
|
|
Post by stormchaser on Jul 13, 2020 23:48:28 GMT -5
Good. I promise we don't have to go overboard, more of curiosity than anything else to see how the two CPUs measure up / and what impact hyper threading has from one to the other, etc. Here is how you run the full report in aida64: 1)at the top click report 2)then click report wizard 3)Under Report profiles select "Benchmark Pages" 4)It will take 5-8 minutes to run through the entire battery of benchmark tests 5)Save the Benchmark report when completed as an HTM or HTML file and then we can attach here Sounds easy enough. So for comparison, are we going to run the exact same clocks, or we gonna race? Ill do my best to at least get 5.0Ghz stable. Im about 90 percent confident the temps will hold. |
|
|
|
Post by ShrimpBrime on Jul 14, 2020 19:54:26 GMT -5
Sounds easy enough. So for comparison, are we going to run the exact same clocks, or we gonna race? Ill do my best to at least get 5.0Ghz stable. Im about 90 percent confident the temps will hold. Just let me know. I'll re-run at the same speeds as yours for better comparisons. |
|
|
|
Post by stormchaser on Jul 19, 2020 1:21:08 GMT -5
There is scant information on the thread ratio result on the internet. It's obviously the result of dividing single core result into multi core result, but I'm thinking, is this some statistic we can use to determine placings? i.e. can it be compared globally to other rigs or is it core specific?
|
|
