Intel Xeon 2696 v2 vs Intel 2697 v2 CPU performance comparison

[storm-chaser]

Any suggestions on which is the better CPU here? From a quick ebay search it looks like the 2696 v2 runs about $100 ish and the 2697 v2 runs about $15-20 more. For very general background, these are both 12 core / 24 thread processors at the top end of the 2600 series family, with the 2697 v2 generally regarded as the "flagship" processor of this series. Feel free to offer up your own opinions and advice, I want a productive conversation and focus on real, attainable facts and specifications (and please double check your specifications, because not everything you read online is accurate), if that is possible here. As I really do want this to be a productive thread, because this has become a hot topic as of late and I've yet to really see exacting information on which CPU is better. I will reserve my opinion for later, since I'm completely certain it will be very controversial to some members here. I want to hear from the forum first, so lets get expert advice, I'm sure there are plenty of Xeon wizards out there who can answer this question for us and settle this once and for all. Which chip is the better performer? Lets hear your thoughts on the matter!

 

[dark2099]

How will you be using the chip? Need to look at specs, but that information will help us offer the cleanest opinion to give back to you.

 

[storm-chaser]

How will you be using the chip? Need to look at specs, but that information will help us offer the cleanest opinion to give back to you.

Okay, so in this specification we will be working with an HP z820 workstation, basically, folding and only folding is what I am going to be doing with this machine, with a focus on heavy, sustained work loads. And occasional gaming with titles like COD and a couple other first person shooters as well, from time to time. This particular z820 is going to get two CPUs, so I dont know if that will have any impact here on actual performance comparisons here, but yes, thats about it. 64GB Ram, 8 channel, 1866MHz RAM setup.

 

[bogmali]

I see where this is going, you want to continue where you left off from this thread. I will allow it but the minute that I see a back and forth arguing about "who is right", consider it shut OP and I might issue some points to boot.

 

[eidairaman1]

Okay, so in this specification we will be working with an HP z820 workstation, basically, folding and only folding is what I am going to be doing with this machine, with a focus on heavy, sustained work loads. And occasional gaming with titles like COD and a couple other first person shooters as well, from time to time. This particular z820 is going to get two CPUs, so I dont know if that will have any impact here on actual performance comparisons here, but yes, thats about it. 64GB Ram, 8 channel, 1866MHz RAM setup.

Hey man, heres 2 spec links.

Does that motherboard you are trying support these 2 way xeons?

https://www.cpu-upgrade.com/CPUs/Intel/Xeon/E5-2696_v2.html

 

[storm-chaser]

Yes, but I want you to know, I will do my best to now leave this rather interesting question in the hands of our resident xeon experts, and will remain in the peripheral, so as to get accurate and exacting information on real performance results between the two processors. Because no doubt, I want the truth here as well, so I will wait for some other expert forum members to offer perspective here. That's all I'm looking for. I will be glad to share my perspective on the matter when the time is right.

Hey man, heres 2 spec links.

Does that motherboard you are trying support these 2 way xeons?

https://www.cpu-upgrade.com/CPUs/Intel/Xeon/E5-2696_v2.html

Yes, the HP z820 supports up to two LGA 2011 processors.

 

[newtekie1]

If you are in the rare situation where the CPUs will be under full load most of the time, the 2696 v2 is better due to the minor 100MHz higher all core boost. If you use your computer as a normal computer most of the time and only occasionally fully load the CPUs, then the 2697 v2 is the better CPU thanks to the higher maximum boost and higher base frequency.

 

[eidairaman1]

Heres the other part.

https://www.cpu-upgrade.com/CPUs/Intel/Xeon/E5-2697_v2.html

 

[storm-chaser]

If you are in the rare situation where the CPUs will be under full load most of the time, the 2696 v2 is better due to the minor 100MHz higher all core boost. If you use your computer as a normal computer most of the time and only occasionally fully load the CPUs, then the 2697 v2 is the better CPU thanks to the higher maximum boost and higher base frequency.

Great. Excellent information. And you say the 2697 v2 has a higher maximum boost, are you referring to something also know as like single core max turbo speed here?

Because there is no doubt these processors have different base frequencies, but we all know going into this one, that given proper CPU cooling both of these processors should really run no less than all core boost speeds at all times. Is that right?

 

[eidairaman1]

Great. Excellent information. And you say the 2697 v2 has a higher maximum boost, are you referring to something also know as like single core max turbo speed here?

Because there is no doubt these processors have different base frequencies, but we all know going into this one, that given proper CPU cooling both of these processors should really run no less than all core boost speeds at all times. Is that right?

Yes the 2697 has a slightly higher base clock and boost clock, only the 8 core has way more...

 

[storm-chaser]

Yes the 2697 has a slightly higher base clock and boost clock, only the 8 core has way more... You

Total side conversation here, but you bring up an interesting point, I am well aware of the 8 core chips and their boost clock advantage, that you speak of here, and my response to that is my second z820 has two 2673 v2 processors, which from my research is the best all around 8 core chip in the 2600 series family due to highly desireable 4.0GHz single core turbo, and 3.3Ghz base turbo, and a very low TDP of 110W. Now you could say that the 2687w is better, and I would be forced to agree, considering it has a 100mhz advantage in base turbo, at 3.4 GHz, but that is a 150w TDP chip, so your gains quite literally go up in smoke, if you get what I'm putting down! Plus that requires the mandatory liquid cooling system which on the z820 is probably just a bunch of hocus pocus anyway! Now the common middle ground and most popular choice for the uninformed would definitely be the 2667 v2, sporting identical turbo specs but again that chip still has a substantial TDP disadvantage when compared with the 2673v2. So those are my thoughts on the 8 core chips from the 2600 series family, just for reference purposes I guess. my logic is I have to have the best of both worlds from intels 2600 series chips, even if that involves two z820s, four processors and four liquid coolers to keep everything nice and cool (I was kidding earlier, z820 liquid coolers are the obvious choice here) so now I can sleep soundly at night knowing I have taken full advantage of intels 2600 series CPU both in core count and boost frequency. I guess the only question remaining will be hearing more about single core boost clock advantages, and whether those advantages lie with the 2696 v2 or the 2697 v2. Because thats the last question we need answered here, since someone already established the fact that the 2696 v2 has a 100mhz all core boost clock advantage over the flagship 2697 v2. Which is slightly odd and paradoxical
....
when you consider it's coming into the game with a lower TDP of 120 watts. It's gotta make you wonder if perhaps even in the retail space we should be buying oem CPUs? Well yeah, I think that is totally clear cut, at least within this cpu family, as it's clear nearly all the stand out cpus are oem only! And yes the 2696 and 2673 are both oem for the record. There has to be some corellation here, does anyone else see it?

 

[newtekie1]

Great. Excellent information. And you say the 2697 v2 has a higher maximum boost, are you referring to something also know as like single core max turbo speed here?

Because there is no doubt these processors have different base frequencies, but we all know going into this one, that given proper CPU cooling both of these processors should really run no less than all core boost speeds at all times. Is that right?

Yes, the all core turbo is just that, the max turbo when all the cores are under load. But like I said, that is a rather rare in most use cases. It isn't just single core max turbo, but the 2697 v2 maintains a higher turbo speed any time 4 cores or less are under load. Which will be most times on a computer under normal use or playing games. Which is why the 2697 v2 would be faster in most use cases. The only time the 2696 v2 would be faster would be in uses where you are using 6 or more cores under load. But with this level of machine, it is admittedly more likely that the person using them would be putting the processor under significant load rather often. Otherwise it'd really a waste of a system like this.

 

[bogmali]

I will reserve my opinion for later, since I'm completely certain it will be very controversial to some members here. I want to hear from the forum first, so lets get expert advice, I'm sure there are plenty of Xeon wizards out there who can answer this question for us and settle this once and for all. Which chip is the better performer? Lets hear your thoughts on the matter!

Yes, but I want you to know, I will do my best to now leave this rather interesting question in the hands of our resident xeon experts, and will remain in the peripheral, so as to get accurate and exacting information on real performance results between the two processors. Because no doubt, I want the truth here as well, so I will wait for some other expert forum members to offer perspective here. That's all I'm looking for. I will be glad to share my perspective on the matter when the time is right.


Yes, the HP z820 supports up to two LGA 2011 processors.

Total side conversation here, but you bring up an interesting point, I am well aware of the 8 core chips and their boost clock advantage, that you speak of here, and my response to that is my second z820 has two 2673 v2 processors, which from my research is the best all around 8 core chip in the 2600 series family due to highly desireable 4.0GHz single core turbo, and 3.3Ghz base turbo, and a very low TDP of 110W. Now you could say that the 2687w is better, and I would be forced to agree, considering it has a 100mhz advantage in base turbo, at 3.4 GHz, but that is a 150w TDP chip, so your gains quite literally go up in smoke, if you get what I'm putting down! Plus that requires the mandatory liquid cooling system which on the z820 is probably just a bunch of hocus pocus anyway! Now the common middle ground and most popular choice for the uninformed would definitely be the 2667 v2, sporting identical turbo specs but again that chip still has a substantial TDP disadvantage when compared with the 2673v2. So those are my thoughts on the 8 core chips from the 2600 series family, just for reference purposes I guess. my logic is I have to have the best of both worlds from intels 2600 series chips, even if that involves two z820s, four processors and four liquid coolers to keep everything nice and cool (I was kidding earlier, z820 liquid coolers are the obvious choice here) so now I can sleep soundly at night knowing I have taken full advantage of intels 2600 series CPU both in core count and boost frequency. I guess the only question remaining will be hearing more about single core boost clock advantages, and whether those advantages lie with the 2696 v2 or the 2697 v2. Because thats the last question we need answered here, since someone already established the fact that the 2696 v2 has a 100mhz all core boost clock advantage over the flagship 2697 v2. Which is slightly odd and paradoxical
....
when you consider it's coming into the game with a lower TDP of 120 watts. It's gotta make you wonder if perhaps even in the retail space we should be buying oem CPUs? Well yeah, I think that is totally clear cut, at least within this cpu family, as it's clear nearly all the stand out cpus are oem only! And yes the 2696 and 2673 are both oem for the record. There has to be some corellation here, does anyone else see it?

Lot's of baiting here OP and I know exactly who these words are intended for, referencing the linked thread on my first post. BLUF (bottom line up front) you need to stop with the "I know more than you" otherwise what is the point of this thread?

 

[storm-chaser]

Lot's of baiting here OP and I know exactly who these words are intended for, referencing the linked thread on my first post. BLUF (bottom line up front) you need to stop with the "I know more than you" otherwise what is the point of this thread?

No, I just want to get the proper information to the masses, the single core turbo speed of the 2696 v2 is 3.5ghz, making it equivalent to the flagship 2697 v2 in single core performance and better in base turbo mode, just wanted to confirm that most of the CPU spec sites online have this point wrong. That's why I made a point of asking for double checking your specs before you post what you think it is.

Yes, the all core turbo is just that, the max turbo when all the cores are under load. But like I said, that is a rather rare in most use cases. It isn't just single core max turbo, but the 2697 v2 maintains a higher turbo speed any time 4 cores or less are under load. Which will be most times on a computer under normal use or playing games. Which is why the 2697 v2 would be faster in most use cases. The only time the 2696 v2 would be faster would be in uses where you are using 6 or more cores under load. But with this level of machine, it is admittedly more likely that the person using them would be putting the processor under significant load rather often. Otherwise it'd really a waste of a system like this.

Thank you these are the kind of details I was hoping for with this thread. I will be back later with more details on my own research into the two processors later.

 

[Selaya]

I mean, I/we are of course open to being proven worng, but every single citation online that I could find claims that the 2696 v2's 1T turbo's 3,300 MHz and not 3,500.
Could you please post proof(s) of your claims (CPU-Z, HWinfo or something)?

 

[Zyll Goliat]

I mean, I/we are of course open to being proven worng, but every single citation online that I could find claims that the 2696 v2's 1T turbo's 3,300 MHz and not 3,500.
Could you please post proof(s) of your claims (CPU-Z, HWinfo or something)?

There is No point to telling him this he knows better and all of them are"lying"....also Intel put 2697 V2 as their flagship to deceive people because clearly according to @storm-chaser 2696 V2 is better.....In Fact he opened this thread after I was saying the very same thing as you.....

I even told him that maybe he have some ES sample or something that boosting up to the 3,5Ghz but NO he will continue to claim that all of this is false that online stats are false and that everybody is wrong except him....

 

[phill]

I'll just leave the link to Intels Ark data sheets here.....

It's a CPU guys, it's not going to make a massive difference... What's 200MHz between PC forum members anyways??...... I very much doubt that Cinebench will be that much quicker.....

 

[Zyll Goliat]

I'll just leave the link to Intels Ark data sheets here.....

View attachment 213993

It's a CPU guys, it's not going to make a massive difference... What's 200MHz between PC forum members anyways??...... I very much doubt that Cinebench will be that much quicker.....

That's the stats for 2697 V2 don't see your point @phill ?

And yeah sure the reality is that both of those CPU's are very close in performance I personally picked 2697 V2 as I assumed that they could be better binned & cherry picked CPU's as they are Intel flagship + they have higher Turbo boost and I needed that as I am doing BCLK overclocking and wanted to put this CPU to the limits also the price difference was not even 20$ or something....Well after all I managed 115 on BCLK 3,45Ghz on all cores + Turbo Boost 4,03Ghz...I guess I did OK....maybe If I have some server motherboard want CPU' to keep on stock or I needed those 20$ for a beer...then yeah why not maybe I will go for 2696 V2 instead....

 

[phill]

That's the stats for 2697 V2 don't see your point @phill ?

And yeah sure the reality is that both of those CPU's are very close in performance I personally picked 2697 V2 as I assumed that they could be better binned & cherry picked CPU's as they are Intel flagship + they have higher Turbo boost and I needed that as I am doing BCLK overclocking and wanted to put this CPU to the limits also the price difference was not even 20$ or something....Well after all I managed 115 on BCLK 3,45Ghz on all cores + Turbo Boost 4,03Ghz...I guess I did OK....maybe If I have some server motherboard want CPU' to keep stocked or I need those 20$ for beer...then yeah why not maybe I will go for 2696 V2 instead....

Ah my apologies, I thought that was the information in need, I can't see any info on Intels site for the E5 2696 V2, I see one for a E5 2695 V2? E5 2695 V2

 

[Zyll Goliat]

Ah my apologies, I thought that was the information in need, I can't see any info on Intels site for the E5 2696 V2, I see one for a E5 2695 V2? E5 2695 V2

Yeah but you can find stats on dozen other site's and all of them saying the same 2,5Ghz-3,3Ghz....The thing is as I already said that to the @storm-chaser that there is some possibility that he have maybe some ES sample or some OEM CPU's or something that possibly have that turbo boost up to the 3,5Ghz.....Even then he get angry at me and opened this thread and continue to claim that all online stats are flawed....

 

[Blue4130]

Any suggestions on which is the better CPU here? From a quick ebay search it looks like the 2696 v2 runs about $100 ish and the 2697 v2 runs about $15-20 more. For very general background, these are both 12 core / 24 thread processors at the top end of the 2600 series family, with the 2697 v2 generally regarded as the "flagship" processor of this series. Feel free to offer up your own opinions and advice, I want a productive conversation and focus on real, attainable facts and specifications (and please double check your specifications, because not everything you read online is accurate), if that is possible here. As I really do want this to be a productive thread, because this has become a hot topic as of late and I've yet to really see exacting information on which CPU is better. I will reserve my opinion for later, since I'm completely certain it will be very controversial to some members here. I want to hear from the forum first, so lets get expert advice, I'm sure there are plenty of Xeon wizards out there who can answer this question for us and settle this once and for all. Which chip is the better performer? Lets hear your thoughts on the matter!

Are you a used car salesman? Every post you make has the most bizarre wording and phrases. They are quite entertaining though...

 

[dgianstefani]

All his threads are like this. He asks advice then does his own thing anyway lmao

 

[phill]

Yeah but you can find stats on dozen other site's and all of them saying the same 2,5Ghz-3,3Ghz....The thing is as I already said that to the @storm-chaser that there is some possibility that he have maybe some ES sample or some OEM CPU's or something that possibly have that turbo boost up to the 3,5Ghz.....Even then he get angry at me and opened this thread and continue to claim that all online stats are flawed....

I never really go for the ES/QS stuff myself, but again I don't see the worry about it all myself. Relax though, its a CPU and some MHz, nothing is worth arguing over that much on a forum. If it gets that bad, there's always an easy fix, ignoring it

 

[Hachi_Roku256563]

Do forgive me if im missing somthing but is it not as simple as ​

Intel® Xeon® Processor E5-2697 (2.7-3.5)​

Faster clocks over ​

Intel Xeon E5-2696 v2 CPU (2.5-3.5) therefore the first chip is faster ​

I cant see any diffrence bar that

 

[storm-chaser]

I mean, I/we are of course open to being proven worng, but every single citation online that I could find claims that the 2696 v2's 1T turbo's 3,300 MHz and not 3,500.
Could you please post proof(s) of your claims (CPU-Z, HWinfo or something)?

Okay, this is real proof of my 2696 v2, boosting to well past the INCORRECTLY REPORTED SPEED OF 3.3GHz Nothing "weird" or wrong with pointing out the accurate information so that other people may benefit down the road, and actually make the better choice and get the 2696 v2 instead.

Yeah but you can find stats on dozen other site's and all of them saying the same 2,5Ghz-3,3Ghz....The thing is as I already said that to the @storm-chaser that there is some possibility that he have maybe some ES sample or some OEM CPU's or something that possibly have that turbo boost up to the 3,5Ghz.....Even then he get angry at me and opened this thread and continue to claim that all online stats are flawed....

If you check CPU world 2696 v2 chip comments, you will see I am NOT the only here who has observed this behavior, here. And for the record, this is not an ES chip, its the standard 2696 v2. This is an OEM chip however, so this should be noted when actually building your system as not all boards and systems will accept a 2696 v2 because it is OEM.

https://www.cpu-world.com/CPUs/Xeon/Intel-Xeon E5-2696 v2.html

Do forgive me if im missing somthing but is it not as simple as ​

Intel® Xeon® Processor E5-2697 (2.7-3.5)​

Faster clocks over ​

Intel Xeon E5-2696 v2 CPU (2.5-3.5) therefore the first chip is faster ​

I cant see any diffrence bar that

You should be able to see now that this is simply a case of CPU spec websites having incorrect information regarding single core turbo speeds, and the 2696 v2 is actually the better processor when you analyze with the correct turbo multipliers, which are listed below:

Turbo Ratio Limits (2696 v2):
35x (1c), 34x (2c), 33x (3c), 32x (4c), 31x (5-12c)

Turbo Ratio Limits (2697 v2)
35x (1c), 34x (2c), 33x (3c), 32x (4c), 31x (5c), 30x (6 or more cores)

More over, this is all the more impressive considering the 2696 v2 has a 100 MHz all core turbo advantage, when you consider processor TDP, coming in at 10 watts less than the 2697 v2.

There is No point to telling him this he knows better and all of them are"lying"....also Intel put 2697 V2 as their flagship to deceive people because clearly according to @storm-chaser 2696 V2 is better.....In Fact he opened this thread after I was saying the very same thing as you.....

View attachment 213979
I even told him that maybe he have some ES sample or something that boosting up to the 3,5Ghz but NO he will continue to claim that all of this is false that online stats are false and that everybody is wrong except him....

NO, its not an ES or some weird version of the CPU. This is the standard 2696 v2 processors we are working with here. I hope in light of all the facts we can see in hindsight, it was merely the fact that Intel simply screwed up and actually mis-named #2 as the flagship processors for the 2600 series family! lol!!

Are you a used car salesman? Every post you make has the most bizarre wording and phrases. They are quite entertaining though...

Not exactly. Used CPU salesman would be slightly more accurate. And that word has slightly more significance, given the inaccuracies we have now established, surrounding the listed turbo speeds of the 2696v2.