HP Z820 - upgrade CPUs or not?

[95Viper]

Stay on the topic, please.
Take your side arguments/disagreements to PMs; or, start a thread about your side topic.

Thank You and Have a Splendid Day.

 

[Zyll Goliat]

You guys are getting carried away with this. My thread has devolved.
View attachment 213277
Pls stop the arguing. Hamster is deeply concerned by it.

I made a typo in my first post, my CPUs are not E5 2620 V2s, they are just E5 2620s. Apparently it was ordered this way for some specification made by the project that was running it even though the Z820 has the V2 bootblock. The upgrade should make a substantial difference!

It also has DDR3 1333 DIMMs as it was specced with a slower CPU. According to the IT folk, it's not the only "weird" configuration in the pile, whatever that means. The Quadro cards I have are not what was shipped in the system either (nor were the SSDs or hard disks, apparently, but I do have enough drive carriers and 2.5" drive adapters to go full SSD if I choose to later, which is nice).

And having the dual Quadro M4000 cards setup with SLI is probably not needed by me, should I remove one to conserve a bit of power or just leave it?

Also took a peek in the chassis and - woooo - dust bunnies 'n stuff. A thorough cleaning is in order.

Looking forward to installing my 2680 V2s this weekend and I still don't know which free tools might be best to capture the before/after performance improvements to share with all y'all.

CPU-Z?
Cinebench?
...?

You can install any or all of those benchmarks:Cinebench15,20,23,Geekbench 4;5,Passmark 9,Corona Benchmark,Cpu-Z,RealBench,7-Zip....they are all good for measuring your CPU performance

 

[Toothless]

You listed two very good programs for benchmarking. Could always go by "feel" for quality of life things.

 

[lexluthermiester]

You guys are getting carried away with this. My thread has devolved.

Agreed. You can ask the mods to close the thread if you want when all is said and done. Just hit the "Report" button at the bottom of a comment and in the input field that comes up just ask for thread closure.

Pls stop the arguing. Hamster is deeply concerned by it.

BTW, not laughing AT you so much as just laughing because this was funny!

I made a typo in my first post, my CPUs are not E5 2620 V2s, they are just E5 2620s. Apparently it was ordered this way for some specification made by the project that was running it even though the Z820 has the V2 bootblock. The upgrade should make a substantial difference!

Ah fair enough and yes, those 2680v2 CPU's you ordered will be a very solid upgrade from the 2620v1's!

 

[storm-chaser]

Agreed. You can ask the mods to close the thread if you want when all is said and done. Just hit the "Report" button at the bottom of a comment and in the input field that comes up just ask for thread closure.

BTW, not laughing AT you so much as just laughing because this was funny!


Ah fair enough and yes, those 2680v2 CPU's you ordered will be a very solid upgrade from the 2620v1's!

Yeah I'm certain you are going to be happy with your new 2680 v2s, definitely one of the more ideal processors in the entire 2600 series family. Sorry to burden the thread earlier. This cpu strikes an excellent all around balance between core count and frequency. Kind of an aside, but If you really want to go all out or just future proof at some point down the road, you could always ditch the original 1333 mhz memory kit and upgrade to something like a 64gb kit containing a total of 16 memory modules running at 1866mhz, the maximum supported memory speed of the z820. Populating all 16 memory slots on the motherboard will unleash 8 channel effective memory bandwidth, since each CPU supports up to 4 individual memory channels on its own. Matter of fact, despite the age of the z820 and the obsolete DDR3 memory platform, this specific memory kit/combination will still net you almost twice the bandwidth of a brand new state of the art DDR4 system running at approximately 4000mhz.

 

[lexluthermiester]

ditch the original 1333 mhz memory kit and upgrade to something like a 64gb kit containing a total of 16 memory modules running at 1866mhz

This, yes! When I first got my 2667v2 I tried it with 1333 ECCreg DIMMs and performance was ok, but someone told me that 1866 was much better for this CPU. 1866ddr3 ECC Reg is cheap so I got a 32GB set(4x8GB) and the difference was a solid improvement. Nothing drastic, but totally worth the price of the set.

@ModemJunki
I don't want to make assumptions, but if you can afford a set of 1866 ECC Registered DIMMS, get them. In a Dual CPU config, 8x4GB(4x4GB for each CPU) will make a good difference in performance for you!

 

[storm-chaser]

Woo lots of good info, thanks everyone.

I went for a pair of 2680 V2s and will stick with the air cooling setup I have now as the liquid coolers aren't available cheap (IT dept doesn't have spares they are willing to give away as they still have to support a bunch of these older machines). I'm sure the 2680s will give this old workhorse a bit more pep with my single-threaded needs for a reasonable cost.

I did find out the reason my particular machine is specced with 2620s was for power budget reasons, apparently it was stuffed with mechanical drives, has 64GB of RAM, and used to have high-end Quadro cards back in it's day where raw CPU wasn't the target.

Question: what would be a good way to "benchmark" the existing vs. future upgrade? CPU-Z?

So forgive me, I had quite a bit of time to kill this evening, so just for fun, here are some reference benchmarks you can measure your new processors against when you get them installed (just running the basic CPUz benchmark first), this is my z820 with two 2696 v2 processors, which amounts to a total of 24 cores and 48 threads with an all core turbo of 3.1GHz and all out single core max turbo of 3.5GHz, which actually puts it on par with the retail 2697 v2 in terms of turbo specs, except that has a TDP of 130W, versus 120W for the 2696 v2 (at the same clock speeds, so the 2696 v2 is actually the better chip). And you can see the comparison to the 1950X Threadripper at the bottom as well....

Turbo Ratio Limits:
35x (1c), 34x (2c), 33x (3c), 32x (4c), 31x (5-12c)

There is the Cinebench R15 benchmark, for another comparison with your new processors.... (same z820 rig as above) (another good benchmark to run)

I posted this earlier in the thread, but for reference, this is the Cinebench R15 Extreme benchmark, another good test for your new CPUs. Im interested to see how your new CPUs stack up against two 2696 v2s!

Here is HWINFO64 to show you actual real time clock speeds. Highly recommend you download this, if you don't have it already. This is an excellent way to draw conclusions on air vs water cooling on a z820 because you can monitor clock speeds in REAL TIME. And make very accurate observations regarding turbo aggressiveness / and basic turbo operation when the system is under full load or at idle speed. With the liquid cooling system installed, the 2696v2 CPU turbo mode was far more aggressive and the system itself as a whole, was nearly silent, even when pushed hard, it was quiet, verses with the air coolers, the fans would spool up more, the system was noticeably louder and the turbo was less aggressive. Once again, more evidence to show you I am not kidding around here when in comes to understanding the difference between air and water cooling on a z820. These are legit results, period, the end. Water cooling is better, and it's easily observable if you pay close attention to your real time clock speeds and turbo modes when you run benchmarks and tax the system, it's very easy to see the advantage of water cooling. That's why even today I am still totally shell shocked that people here are recommending the weaker option for your system. Sorry, when I build z820 rigs, I like to build them the best way possible and that dictates the mandatory use of the liquid cooling system, ALWAYS.

Observe the real time clocks here, this is key, as you can see its boosting very well and aggressively engaging the turbo, and all cores are running near the maximum turbo "hard" limits set by Intel. I can say for certain, with a POS junk air cooler installed, you will NOT get the same result, and this is especially true with chips over 130 watts TDP. The bottom line is the air cooler just runs out of steam, and the choice or decision making process at that point becomes obvious.

If I were you, I wouldn't be happy until my z820 looked like this:

Here is my z820 measured against HWBOT, seems like I have a pretty rare "hardware configuration" and that's the way I like it. Yeah, you gotta bring some serious heat to the table if you want to match two 2696 v2 processors in a z820!

Here is my second z820 rig, also liquid cooled, of course, running two 2673 v2 OEM processors, in point of fact, this chip is definitely one of the most rare processors from the 2600 series family, never released, OEM only, which means it can be difficult to find and likely have to be purchased from a second hand supplier that parts out old retired servers. Probably the best 2600 series chip you can get with 8 cores and 16 threads due to the desirable 4.0GHz turbo. The 2687w has very similar turbo specs and clock speeds but comes it at like 20-30 watts more, who needs the extra heat, right?

The z820 motherboard is still something to marvel at.. it's sick... and still kicks some serious ass if you build it up the right way.

Oh yeah, one more thing, lets take a closer look at this so called better "liquid" cooling system, after all, storm-chaser has no idea what he is talking about, right? ... as you can clearly see here, this z820 liquid cooler is just a generic piece of junk. taken right off the factory shelf and basically just thrown at the z820 with no second thought and no care in the world. It's just a boring, generic cooler, you've seen the same thing a thousand times, who the hell cares, nothing to see here, it's junk, so just use the air cooler instead, that's so much smarter, of course you can do that, if you want more heat and more noise, and less aggressive turbo operation on your CPU, go for it!


(Just kidding, take a look at this masterpiece, it's pretty impressive if you ask me...) AND I SHOULDNT HAVE TO SELL IT TO ANYONE, ITS BETTER, PERIOD, THE END.

 

[newtekie1]

It's crazy how far we've come. My 10-Core 10850K outperforms the 16-cores of the dual 2673 v2 system in multi-threaded and just blows it away in single threaded. And actually comes respectably close to 24-core system.

(Just kidding, take a look at this masterpiece, it's pretty impressive if you ask me...) AND I SHOULDNT HAVE TO SELL IT TO ANYONE, ITS BETTER, PERIOD, THE END.

Yeah, that metal bracket that they designed to hold the 80mm slim rad of that bog standard Asetek cooler really just sends the performance through the roof.

 

[Zyll Goliat]

This is my z820 with two 2696 v2 processors, which amounts to a total of 24 cores and 48 threads with an all core turbo of 3.1GHz and all out single core max turbo of 3.5GHz, which actually puts it on par with the retail 2697 v2 in terms of turbo specs, except that has a TDP of 130W, versus 120W for the 2696 v2 (at the same clock speeds, so the 2696 v2 is actually the better chip).

I don't know how do you get this conclusion??I mean IF the logic is that 2696V2 is the better chip because you owned...then...sure...why not ....
I am pretty sure that Intel didn't put higher price for 2697V2 just to make their consumers stupid...also that 130W power consumption is there for a reason.....to cut the BS those chips are very similar 'tho I am pretty sure that 2697V2 are probably better binned chips and IF I could guess they could reach probably much higher clock when it comes to the BCLK OC which was the main reason why I got 2697V2 that was a bit more expensive compared to the 2696V2.......

P.S.Also don't see the real reason for water cooling...I mean sure why not if you prefer&love to play with that otherwise nah...those chips are pretty cold and personally in my case even when this CPU is pushed to work on 3,45Ghz on all cores and Boosting up to the 4,03Ghz it almost never reach 65c even in those torture tests 70c-73c is around the max temps I get ever...and all I got is the cheap/crappy air cooler

 

[storm-chaser]

It's crazy how far we've come. My 10-Core 10850K outperforms the 16-cores of the dual 2673 v2 system in multi-threaded and just blows it away in single threaded. And actually comes respectably close to 24-core system.

View attachment 213605


Yeah, that metal bracket that they designed to hold the 80mm slim rad of that bog standard Asetek cooler really just sends the performance through the roof.

It's crazy how far we've come. My 10-Core 10850K outperforms the 16-cores of the dual 2673 v2 system in multi-threaded and just blows it away in single threaded. And actually comes respectably close to 24-core system.

View attachment 213605


Yeah, that metal bracket that they designed to hold the 80mm slim rad of that bog standard Asetek cooler really just sends the performance through the roof.

Are u sure its even a metal bracket? Rumor has it Asetek (i just love when you throw that in like its a bad thing) was using plastic parts here from the junk pile for this liquid cooler. But hey, since you like more noise and lower performance, keep using your lovely air cooler. Since you apparently have zero comprehension of why a liquid cooler would be better for high end high TDP processors in the 2600 series family, you should probably read this HP documentation so you can finally put "the pieces together" for yourself.... of course it could all be a big lie, with the whole z820 cash cow thing, remember? According to you, HP is making millions off the z820 liquid cooling system, right?

Here, it's high time you finally figured this out for yourself!

http://h20331.www2.hp.com/Hpsub/downloads/Liquid_Cooling_HP_Z420_Z820_Workstations.pdf

A bit of a highlight from that link...

Acoustic noise in personal workstations There are multiple sources of noise in personal workstations, including fans, hard disk drives (HDDs), optical disk drives (ODDs), and other noises e.g., liquid cooling pumps, keyboards, etc. Fans and high speed (15K RPM) hard disk drives are typically the strongest sources of acoustic noise in a workstation. Workstations generally contain multiple fans—fans to exhaust air from the interior of the box, fans (blowers) on graphics cards, fans for memory cooling, and fans on the processor heat sink(s). There are two ways to reduce fan speeds (and noise). Reduce the amount of heat produced or use more efficient cooling mechanisms. Reducing the generated heat, while possible, is typically unacceptable; for example, reducing heat by lowering the processor frequency or number of cores results in lower performance. And with the advent of Intel® Turbo Boost Technology, the processor can now spend more time at even higher frequencies, increasing the processor cooling demands. Consequently, more efficient cooling mechanisms are typically the most effective way to get lower acoustics. This can include material changes, additional ducting, or heatsinks with larger surface area. In the case of liquid cooling, the small CPU contact area of a heatsink is replaced with a cold plate with many micro-channels that quickly remove heat from the CPU. And a smaller heatsink surface area (fins) for transferring heat to the environment is replaced by the larger surface area of a radiator, with liquid providing an effective means of transporting the CPU heat to the radiator. Figure 1 – HP Z420 Liquid Cooling HP recommends Windows® 7. 2 Acoustic benefits of liquid cooling The increased thermal efficiency of liquid cooling delivers lower CPU-induced acoustic noise at high frequency, high power CPU operating states. In fact, in a laboratory comparison of results between the liquid cooled and non-liquid cooled systems, the reduction in acoustic noise was measured at 2 dBA (perceived as about 15 percent quieter) while running mainstream CPUs1 at a heavy CPU load2 level. And with the most extreme CPUs1 , liquid cooling allows the workstation to maintain the same acoustic noise level as when using a mainstream CPU. The reduced noise level can make users more comfortable and productive, without sacrificing application performance. In addition, studies have also shown that certain kinds of acoustic noise are more distracting than others, especially higher frequencies and irregular modulations. In an air-cooled system at heavy workloads, the CPU fan(s) ramp quickly to keep the CPU cool. Under cyclic loading, the fan(s) ramp up and down with the workload. Because the LCS has higher thermal inertia, the associated fans do not ramp up and down as quickly.

Now, you have to remember, the low end air cooler is no bigger than the liquid cooling system, they both take up approximately the same space. If you prefer it, keep your air coolers. That way you will definitely know it's working overtime when the fans ramp up all the time!

I don't know how do you get this conclusion??I mean IF the logic is that 2696V2 is the better chip because you owned...then...sure...why not ....
I am pretty sure that Intel didn't put higher price for 2697V2 just to make their consumers stupid...also that 130W power consumption is there for a reason.....to cut the BS those chips are very similar 'tho I am pretty sure that 2697V2 are probably better binned chips and IF I could guess they could reach probably much higher clock when it comes to the BCLK OC which was the main reason why I got 2697V2 that was a bit more expensive compared to the 2696V2.......

LOL I'll give you one chance to just do this over again and delete this nonsensical comment before I show you the real turbo specs between the 2697 v2 and the 2696 v2... LOL @ the "binned better" comment you have no idea what you are talking about. After all, I am the one that is correct here, as the 2696 v2 has a 100MHz advantage in all core turbo over the 2697v2, you probably should have checked the processor specs before you blindly marched forward with bad and inaccurate information.

P.S.Also don't see the real reason for water cooling...I mean sure why not if you prefer&love to play with that otherwise nah...those chips are pretty cold and personally in my case even when this CPU is pushed to work on 3,45Ghz on all cores and Boosting up to the 4,03Ghz it almost never reach 65c even in those torture tests 70c-73c is around the max temps I get ever...and all I got is the cheap/crappy air cooler

You can keep your air cooler if you want, I've got nothing against that. More power to you! I just needed to clarify and correct the junk information in this thread regarding the z820 cooling systems.

It's crazy how far we've come. My 10-Core 10850K outperforms the 16-cores of the dual 2673 v2 system in multi-threaded and just blows it away in single threaded. And actually comes respectably close to 24-core system.

Okay, but then I can come at you with both systems, and your 10 core 10850K would get totally obliterated. By really old hardware, I might add! And no, it doesn't come anywhere near my 24 core / 48 thread system. You didn't even crack 7400! The 2696 v2 z820 rig is hitting 9500! That's a major difference, and plus I have no doubt I would totally obliterate your 10850K rig in memory bandwidth as well, so you probably don't even want to go there... You guys are making all the standard mistakes, sorry to capitalize on it like this! Don't underestimate (a well built) z820, it's far stronger than you know! Some people have to learn the hard way, lol!

 

[Zyll Goliat]

Are u sure its even a metal bracket? Rumor has it Asetek (i just love when you throw that in like its a bad thing) was using plastic parts here from the junk pile for this liquid cooler. But hey, since you like more noise and lower performance, keep using your lovely air cooler. Since you apparently have zero comprehension of why a liquid cooler would be better for high end high TDP processors in the 2600 series family, you should probably read this HP documentation so you can finally put "the pieces together" for yourself.... of course it could all be a big lie, with the whole z820 cash cow thing, remember? According to you, HP is making millions off the z820 liquid cooling system, right?

Here, it's high time you finally figured this out for yourself!

http://h20331.www2.hp.com/Hpsub/downloads/Liquid_Cooling_HP_Z420_Z820_Workstations.pdf

A bit of a highlight from that link...


Now, you have to remember, the low end air cooler is no bigger than the liquid cooling system, they both take up approximately the same space. If you prefer it, keep your air coolers. That way you will definitely know it's working overtime when the fans ramp up all the time!

View attachment 213614


LOL I'll give you one chance to just do this over again and delete this nonsensical comment before I show you the real turbo specs between the 2697 v2 and the 2696 v2... LOL @ the "binned better" comment you have no idea what you are talking about. After all, I am the one that is correct here, as the 2696 v2 has a 100MHz advantage in all core turbo over the 2697v2, you probably should have checked the processor specs before you blindly marched forward with bad and inaccurate information.


You can keep your air cooler if you want, I've got nothing against that. More power to you! I just needed to clarify and correct the junk information in this thread regarding the z820 cooling systems.

Right I forgot you are smarter then Intel engineers that decide to put higher price on actual worst product and named with the highest number in that series just for fun.....2696V2 have 25 multiplier and it's boosting up to the 3,3Ghz compared to the 2697V2 that have 27 multiplier and boosting up to the 3,5Ghz.....

P.S.I really don't mind water cooling and I never said that it's not better...I just said that I don't see the point for that when it comes to those CPU's because those are cold and there is not going to be any difference when it comes to the performance as they are also locked for OC and even IF you OC them via BCLK they are still pretty much cold even with the cheap air cooler and that was my point.....

 

[storm-chaser]

Right I forgot you are smarter then Intel engineers that decide to put higher price on actual worst product and named with the highest number in that series just for fun.....2696V2 have 25 multiplier and it's boosting up to the 3,3Ghz compared to the 2697V2 that have 27 multiplier and boosting up to the 3,5Ghz.....
View attachment 213618

LOL you are wrong again, (because you failed to do the legwork, but don't worry, this is a common mistake people make that don't know the real potential of the 2696 v2). I learned long ago, sonny boy, that even cpu world has it wrong sometimes. In other words, the 2696 v2 boosts to 3.5GHz, NOT 3.3GHz.

Turbo Ratio Limits:
35x (1c), 34x (2c), 33x (3c), 32x (4c), 31x (5-12c)

I will say it again, all core max turbo speed is 100MHz faster than 2697 v2. I'm sure you will have a hard time accepting this, but it is true. So just let it go, you were wrong, again.

 

[Zyll Goliat]

LOL you are wrong again, (because you failed to do the legwork, but don't worry, this is a common mistake people make that don't know the real potential of the 2696 v2). I learned long ago, sonny boy, that even cpu world has it wrong sometimes. In other words, the 2696 v2 boosts to 3.5GHz, NOT 3.3GHz.

Turbo Ratio Limits:
35x (1c), 34x (2c), 33x (3c), 32x (4c), 31x (5-12c)

I will say it again, all core max turbo speed is 100MHz faster than 2697 v2. I'm sure you will have a hard time accepting this, but it is true. So just let it go, you were wrong, again.

Right...right...everybody is wrong...those guy at Intel,me,@newtekie1 ,CPU World....everybody....all of us conspire just to get on your nerves ....

 

[storm-chaser]

Right...right...everybody is wrong...those guy at Intel,me,@newtekie1 ,CPU World....everybody....all of us conspire just to get on your nerves ....

No, no... Please, you have to understand, I'm being serious here, CPU world is actually wrong in this case, the 2696 v2 boosts 100Mhz higher than Intel's flagship 2697, and I've got screenshots to prove it, so yes, while there have been major inaccuracies coming from you, but those have been mostly corrected now, it's just the other guy who is apparently IN LOVE with the z820's air cooler, he will be in for a rude awakening if he ever upgrades to a 2687w, because the system will warn him on startup that you need the liquid cooling system is a pre-requisite when using this processor. Of course, said guy in love with the air coolers will write that off as HP trying to make millions on roughly 1% of z820 configured hardware. Yeah, that's brilliant logic! But it's all he's got left to say. The only other logical answer would force him to say the liquid cooling is better. And he cant do that now that he is totally invested in refusing to believe there are real world advantages to the liquid cooling system on a z820.

Any other questions, gentlemen? Seems like the truth is slowly hitting the brain stem now. That's a good thing!

 

[the54thvoid]

The OP asked earlier for the arguing to end. I'll make this a final warning.

 

[storm-chaser]

Okay, good deal, so I found a couple more benchmarks you can measure your new CPUs against as well, here you go, this is the same z820 with two 2696 v2 chips, as I mentioned earlier.... just for comparison purposes since you asked what other benchmarks you can run... I imagine your dual 2680s v2s will be pretty competitive here considering you have an all core turbo of 3.1GHz and even a slight advantage in single core turbo max @ 3.6GHz ...

This is Passmark Performance Test 9.0...

You can view detailed passmark results here:

PassMark Software - Display Baseline ID# 1360789

www.passmark.com

Here is PassMark 10 Performance Test Results:

CPU Placement of the 2696 v2 on the overall CPU Chart...

 

[Zyll Goliat]

Okay, good deal, so I found a couple more benchmarks you can measure your new CPUs against as well, here you go, this is the same z820 with two 2696 v2 chips, as I mentioned earlier.... just for comparison purposes since you asked what other benchmarks you can run... I imagine your dual 2680s v2s will be pretty competitive here considering you have an all core turbo of 3.1GHz and even a slight advantage in single core turbo max @ 3.6GHz ...

This is Passmark Performance Test 9.0...

View attachment 213636

You can view detailed passmark results here:

PassMark Software - Display Baseline ID# 1360789

www.passmark.com

Here is PassMark 10 Performance Test Results:
View attachment 213638

CPU Placement of the 2696 v2 on the overall CPU Chart...
View attachment 213639

View attachment 213640

BTW....That Passmark 9 CPU score seems really low for 2x2696V2,you maybe tested only with 1 CPU or something?I mean seriously it's impossible that I have better score with 12c/24t(even OC)then yours 24c/48t?Something is Off there for sure.....
P.S.You either tested with 12c/24t "only" or maybe your CPU's are throttling as they overheat on that famous liquid cooling ....

 

[storm-chaser]

BTW....That Passmark 9 CPU score seems really low for 2x2696V2,you maybe tested only with 1 CPU or something?

Come on buddy, now you are clearly looking for a reaction with a comment like that! Lets all play nice, I am here to enjoy and share in the success of all z820 systems, primarily because I think it was one of the sickest computers ever built, totally amazing tool less design and these rigs can still be very competitive against state of the art, 4500MHz DDR4 systems from 2021 (in fact they still obliterate these brand new rigs in memory bandwidth with the right memory setup) so when I see someone building one I have specific thoughts and ideas to share, and enjoy exploring the performance potential and seeing the results he will now hopefully share with the group once he gets his new processors installed.

FYI, to respond to your initial comment with sincere consideration, because I still like everyone here, we can all get along, look at the Passmark 9 CPU score, as I highlighted here, you will see im in the 99th percentile in CPU mark --------> In addition, I will also post an AIDA64 memory benchmark as well. It will be something for you to "work up to" (I have to crack a little smile here)... Seeing as you have the latest socket 1200 Intel rig, and a sick, state of the art Intel processor (and by the way, I mean that, it's a great chip and one that I have actually been eyeing for another build) and you probably some very fast memory, you should have no problems beating a 10 year old z820 in memory throughput, right? Well.. We will see about that, wont we!

 

[Zyll Goliat]

Come on buddy, now you are clearly looking for a reaction with a comment like that! Lets all play nice, I am here to enjoy and share in the success of all z820 systems, primarily because I think it was one of the sickest computers ever built, totally amazing tool less design and these rigs can still be very competitive against state of the art, 4500MHz DDR4 systems from 2021 (in fact they still obliterate these brand new rigs in memory bandwidth with the right memory setup) so when I see someone building one I have specific thoughts and ideas to share, and enjoy exploring the performance potential and seeing the results he will now hopefully share with the group once he gets his new processors installed.

FYI, to respond to your initial comment with sincere consideration, because I still like everyone here, we can all get along, look at the Passmark 9 CPU score, as I highlighted here, you will see im in the 99th percentile in CPU mark --------> In addition, I will also post an AIDA64 memory benchmark as well. It will be something for you to "work up to" (I have to crack a little smile here)... Seeing as you have the latest socket 1200 Intel rig, and a sick, state of the art Intel processor (and by the way, I mean that, it's a great chip and one that I have actually been eyeing for another build) and you probably some very fast memory, you should have no problems beating a 10 year old z820 in memory throughput, right? Well.. We will see about that, wont we!

View attachment 213641

No dude seriously that result is totally off....Your Passmark 10 score seems OKish I guess but Passmark 9 seems wrong for 24c/48t.....Look

I mean it's just impossible that I have better score with 12c/24t Vs 24c/48t...I mean sure my CPU is a bit OC but still this looks surreal......

 

[storm-chaser]

No dude seriously that result is totally off....Your Passmark 10 score seems OKish but Passmark 9 seems wrong for 24c/48t.....Look
View attachment 213643
I mean it's just impossible that I have better score with 12c/24t Vs 24c/48t...I mean sure my CPU is a bit OC but still this is just impossible...

No its right, everything is fine, there is no problem with the result, as I said, I am in the 99th percentile and if you look at the result, its accurate, it's basically just that you lack basic perspective to draw accurate conclusions here. No problem, let me explain, because other people here may have the same problem you have... it is actually fairly commonplace... let me explain... You have to remember, z820s are old. Think about it for a second, you have a brand new and fairly decently overclocked, state of the art, latest socket 1200 Intel processor from the year 2021 and I have two 9 year old processors from 2012-2013 that are not overclockable and we are basically still neck and neck. We are basically even, all things considered. To be honest, I would rather prefer use older hardware and fight to stay competitive with new generations, I see that as a more satisfying challenge than just going out and getting the latest hardware and then bragging about your brand new "state of the art" setup and how you set your multiplier to 55 and added a little voltage, and BOOM!! that's about it. (Although I do have a very sick liquid cooled i5 9600KF rig that goes to 5.3GHz stable and actually boots at 5.6Ghz as well) Perhaps you don't understand that aspect, it's a pure enthusiast aspect that some people just don't get/understand. And some never well, and that's there/your loss to be honest. All things considered, our rigs are just about even... Especially considering I've still got way more memory channels... so that's gonna maybe put a frown on your face!

 

[Zyll Goliat]

No its right, everything is fine, there is no problem with the result, as I said, I am in the 99th percentile and if you look at the result, its accurate, it's basically just that you lack basic perspective to draw accurate conclusions here. No problem, let me explain, because other people here may have the same problem you have... it is actually fairly commonplace... let me explain... You have to remember, z820s are old. Think about it for a second, you have a brand new and fairly decently overclocked, state of the art, latest socket 1200 Intel processor from the year 2021 and I have two 9 year old processors from 2012-2013 that are not overclockable and we are basically still neck and neck. We are basically even, all things considered. To be honest, I would rather prefer use older hardware and fight to stay competitive with new generations, I see that as a more satisfying challenge than just going out and getting the latest hardware and then bragging about your brand new "state of the art" setup and how you set your multiplier to 55 and added a little voltage, and BOOM!! that's about it. (Although I do have a very sick liquid cooled i5 9600KF rig that goes to 5.3GHz stable and actually boots at 5.6Ghz as well) Perhaps you don't understand that aspect, it's a pure enthusiast aspect that some people just don't get/understand. And some never well, and that's there/your loss to be honest. All things considered, our rigs are just about even... Especially considering I've still got way more memory channels... so that's gonna maybe put a frown on your face!

Ha ha ha...Dude WTF are you talking about ???I believe I told you already what I have...anyway if you missed that part I HAVE Xeon 2697 V2(see my system specs)!!! And that passmark 9 scores is also from my very "old" X79 Rig....and this is the reason why I am telling you that your score looks off but seems like you get offended or something.....

P.S.BTW TY for your honest mistake saying for my result and my Xeon 2697 V2 that is state of the art, believing that my CPU is the latest socket 1200 Intel processor from the year 2021 was very pleasant to hear especially after your previous claims that 2696 V2 is actually better CPU

 

[95Viper]

OP has made his purchase decision and his questions have been answered.
Thread is now closed as it has turned into a side topic disagreement/discussion.