Sign in with
Sign up | Sign in

Intel Xeon E5-2600 v2: More Cores, Cache, And Better Efficiency

Intel Xeon E5-2600 v2: More Cores, Cache, And Better Efficiency
By

Intel recently launched its Xeon E5-2600 v2 CPU, based on the Ivy Bridge-EP architecture. We got a couple of workstation-specific -2687W v2 processors with eight cores and 25 MB of L3 cache each, and are comparing them to previous-generation -2687Ws.

Intel has a lot of irons in its proverbial fire. The company is merrily hammering away on the mobile space with its Haswell and Silvermont architectures. It’s targeting everything from tablets to notebooks with very powerful integrated graphics, and doing a really good job, in my opinion. We have Bay Trail-based tablets in the lab that look promising, and I have a bunch of data from the Iris Pro 5200 graphics engine that hasn’tevenbeen published yet.

However, the desktop portfolio is cooling off to the side after more than two years of fairly modest evolution. Because the Tom’s Hardware team spends so much time at work and play on stationary workstations, our disappointment with Intel’s efforts in this space is woven through much of the site’s content, even if Intel does sell the fastest CPUs. Without AMD competing at a high enough level, we’ve had very little reason to recommend upgrading your host processor since the Sandy Bridge days.

Not so in the server and workstation space. There, Intel leverages its manufacturing strength to build CPUs that deftly cut through professional applications and carve up our efficiency measurements. A few months back, we previewed the Xeon E5-2697 v2 in Intel's 12-Core Xeon With 30 MB Of L3: The New Mac Pro's CPU? and determined the chip to be more efficient than either the eight-core Xeon E5-2687W or Core i7-3970X.

And now I have a pair of Xeon E5-2687W v2 processors based on the same Ivy Bridge-EP design. Although they’re 150 W CPUs (the -2697 v2 is a 130 W part), the workstation-specific chips sport the same eight cores as the first-gen -2687W. Intel added shared L3 cache, though. It also increased the peak base and Turbo Boost frequencies at a lower peak VID. And I know both CPUs bear the same TDP, but the newer architecture is simply lower-power.  

Meet Ivy Bridge-EP

After today, we’ll have benchmarked the 12- and eight-core Xeon E5-2600-series CPUs. But Intel also sells four-, six-, and 10-core models as well. In fact, there are 18 total SKUs up and down the v2 stack. The diverse line-up is derived from three physical dies sporting six, 10, and 12 cores. As you can imagine, each set of resources is intentionally modular to help simplify the creation of these different designs.

12-core die12-core die

The most complex version, which is what we previewed in the Xeon E5-2697 story, employs three columns of building blocks, consisting of cores and 2.5 MB last-level cache slices, and four rows of those resources. Multiple ring buses facilitate communication across the die, and multiplexers ensure information gets to the stop where it’s needed. There’s a single QPI agent communicating at up to 9.6 GT/s (though existing models are capped at 8 GT/s), and 40 lanes of third-gen PCI Express connectivity split into two x16 links and an additional eight-lane link. The 12-core die utilizes two memory controllers, each responsible for two channels of up to DDR3-1866.

10-core die10-core die

Stepping back to 10 cores reduces complexity quite a bit. The configuration shrinks to two columns, but is now five rows long. The QPI agent remains intact, but maxes out at 8 GT/s, while the PCI Express controller doesn’t change at all. Intel’s 10-core configuration sports a single memory controller that hosts all four channels. And there’s just one ring bus to shuttle data between the various stops, too.

Eight-core Xeon E5s are sourced from this same die, so a couple of the core/cache slices are disabled, leaving everything else functional. Incidentally, that’s how the Xeon E5-2687W v2 we’re testing today can include eight cores but also a 25 MB shared L3 cache—two cores are disabled, but the corresponding L3 remains active.

Six-core dieSix-core die

Once you drop to six cores, it’s cheaper to create a third die configuration than to create higher-volume parts from the pricey 10-core arrangement. Also a two-column part, the six-core CPU is three rows long with an 8 GT/s-capable QPI agent and the same PCI Express connectivity. Again, one memory controller is responsible for all four 64-bit DDR3 channels.

Intel uses those three dies to create a stack broken up into Advanced, Standard, Basic, and Segment-Optimized models ranging from 60 up to 150 W, and base clock rates from 1.7 up to 3.5 GHz. From the top to the bottom, the entire portfolio is compatible with the same LGA 2011 (Socket R) interface as before. That means upgrading an existing server or workstation is as easy as updating the platform’s firmware. On our Intel W2600CR2 motherboard, we simply did this with the Sandy Bridge-EP-based Xeon E5-2687Ws installed.  


Cores
LLC
QPI
Memory
Base Clock
TDP
Price
Advanced
Xeon E5-2690 v2
10
25 MB
8 GT/s
DDR3-1866
3.0 GHz
130 W
$2057
Xeon E5-2680 v2
10
25 MB8 GT/sDDR3-18662.8 GHz
115 W
$1723
Xeon E5-2670 v2
10
25 MB8 GT/sDDR3-18662.5 GHz
115 W
$1552
Xeon E5-2660 v2
10
25 MB8 GT/sDDR3-18662.2 GHz
95 W
$1389
Xeon E5-2650 v2
8
20 MB
8 GT/sDDR3-18662.6 GHz
95 W
$1166
Standard
Xeon E5-2640 v2
8
20 MB
7.2 GT/s
DDR3-1600
2.0 GHz
95 W
$885
Xeon E5-2630 v2
6
15 MB
7.2 GT/sDDR3-16002.6 GHz
80 W
$612
Xeon E5-2620 v2
6
15 MB
7.2 GT/sDDR3-16002.1 GHz
80 W
$406
Basic
Xeon E5-2609 v2
4
10 MB
6.4 GT/s
DDR3-1333
2.5 GHz
80 W
$294
Xeon E5-2603 v2
4
10 MB
6.4 GT/s
DDR3-1333
1.8 GHz
80 W
$202
Segment-Optimized
Xeon E5-2697 v2
12
30 MB
8 GT/s
DDR3-18662.7 GHz
130 W
$2614
Xeon E5-2695 v2
12
30 MB
8 GT/s
DDR3-18662.4 GHz
115 W
$2336
Xeon E5-2687W v2
8
20 MB
8 GT/s
DDR3-18663.4 GHz
150 W
$2108
Xeon E5-2667 v2
8
25 MB
8 GT/sDDR3-18663.3 GHz
130 W
$2057
Xeon E5-2643 v2
6
25 MB
8 GT/sDDR3-18663.5 GHz
130 W
$1552
Xeon E5-2637 v2
4
15 MB
8 GT/sDDR3-18663.5 GHz
130 W
$996
Xeon E5-2650L v2
10
25 MB
8 GT/sDDR3-16001.7 GHz
70 W
$1219
Xeon E5-2630L v2
6
15 MB
7.2 GT/s
DDR3-16002.4 GHz
60 W
$612

CPUs in the Advanced bin are mostly 10-core models with 25 MB of L3, though there’s an eight-core CPU with 20 MB in there as well. They all feature 8 GT/s QPI links, Hyper-Threading and Turbo Boost support, and a memory controller capable of 1866 MT/s transfer rates.

The Standard stack is smaller, with one eight-core SKU boasting 25 MB of LLC and two six-core chips complemented by 15 MB. Intel’s QuickPath Interface is deliberately slowed to 7.2 GT/s, as is the quad-channel memory controller’s maximum speed (all three processors accommodate up to DDR3-1600 modules). Hyper-Threading and Turbo Boost are both retained, though.

Both members of the Basic segment are quad-core CPUs with 10 MB of shared L3. QPI performance is pared back to 6.4 GT/s, while the memory controller tops out at DDR3-1333. That’s still arguably plenty for the lower-power applications those processors will find themselves in, though Intel does deactivate Hyper-Threading and Turbo Boost, unfortunately.

Intel Xeon E5-2600 v2: More Cores, More Cache, And Better Efficiency

Xeon E5-2687W v2: Bringing Out The Big Guns

Of course, Intel’s Xeon E5-2687W v2 doesn’t fit into any of those three categories. It’s a workstation-specific member of the Segment-Optimized line-up, purpose-built for roomy pedestal/4U enclosures where dissipating 2 x 150 W isn’t a problem, and a balance between parallelism and clock rate takes precedent over more lower-frequency cores.

Like the Xeon E5-2687W before it, -2687W v2 is an eight-core part. Its base clock rate increases from 3.1 GHz the generation prior up to 3.4 GHz, and the maximum Turbo Boost frequency similarly jumps from 3.8 to 4 GHz. An extra 5 MB of shared L3 cache typically won’t confer significant gains. However, you will see benchmark situations where it makes a difference.

Each of the -2687W v2’s QPI links operate at a full 8 GT/s. And the processor’s quad-channel memory controller supports 1866 MT/s data rates. In theory, that’s up to 59.7 GB/s per processor, though real-world throughput is always going to be lower.

The 10-core die on which our Xeon E5-2687W v2 is basedThe 10-core die on which our Xeon E5-2687W v2 is based

Of course, the second-gen Xeon E5 is built using Intel’s Ivy Bridge architecture, so it gets the subtle tweaks introduced back in April 2012 alongside the company’s desktop Core CPUs, including a handful adjustments to the core, cache, and memory controller that improve IPC throughput by a few percent compared to Sandy Bridge.

When you combine the architectural evolution, higher clock rates, and more shared L3 cache, you know what to expect going from Xeon E5-2687W to -2687W v2. But that’s not the whole story. When Intel made the switch from Sandy to Ivy Bridge, its emphasis was on transitioning from 32 to 22 nm manufacturing. The company does successfully push the Xeon E5 family’s performance story forward. However, it also cuts power consumption. That combination is great for boosting efficiency. So, we’re going to start by digging into the benchmarks, fold in power consumption, and then wrap with an energy comparison.

Ask a Category Expert

Create a new thread in the Reviews comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 47 comments.
This thread is closed for comments
  • 1 Hide
    GL1zdA1 , January 7, 2014 12:54 AM
    Does this mean, that the 12-core variant with 2 memory controllers will be a NUMA CPU, with cores having different latencies when accessing memory depending on which MC is near them?
  • 1 Hide
    Draven35 , January 7, 2014 1:43 AM
    The Maya playblast test, as far as I can tell, is very single-threaded, just like the other 3d application preview tests I (we) use. This means it favors clock speed over memory bandwidth.

    The Maya render test seems to be missing O.o
  • 1 Hide
    Cryio , January 7, 2014 2:43 AM
    Thank you Tom's for this Intel Server CPU. I sure hope you'll make a review of AMD's upcoming 16 core Steamroller server CPU
  • 2 Hide
    Draven35 , January 7, 2014 3:08 AM
    Tell AMD that.
  • 4 Hide
    cats_Paw , January 7, 2014 3:09 AM
    Dat Price...
  • 0 Hide
    voltagetoe , January 7, 2014 3:10 AM
    If you've got 3ds max, why don't you use something more serious/advanced like Mental Ray ? The default renderer tech represent distant past like year 1995.
  • 0 Hide
    lockhrt999 , January 7, 2014 3:13 AM
    "Our playblast animation in Maya 2014 confounds us."@canjelini : Apart from rendering, most of tools in Maya are single threaded(most of the functionality has stayed same for this two decades old software). So benchmarking maya playblast is as identical as itunes encode benchmarking.
  • 2 Hide
    daglesj , January 7, 2014 3:33 AM
    I love Xeon machines. As they are not mainstream you can usually pick up crazy spec Xeon workstations for next to nothing just a few years after they were going for $3000. They make damn good workhorses.
  • 1 Hide
    InvalidError , January 7, 2014 6:20 AM
    @GL1zdA1: the ring-bus already means every core has different latency accessing any given memory controller.Memory controller latency is not as much of a problem with massively threaded applications on a multi-threaded CPU since there is still plenty of other work that can be done while a few threads are stalled on IO/data. Games and most mainstream applications have 1-2 performance-critical threads and the remainder of their 30-150 other threads are mostly non-critical automatic threading from libraries, application frameworks and various background or housekeeping stuff.
  • 2 Hide
    mapesdhs , January 7, 2014 8:03 AM
    Small note, one can of course manually add the Quadro FX 1800 to the relevant file
    (raytracer_supported_cards.txt) in the appropriate Adobe folder and it will work just
    fine for CUDA, though of course it's not a card anyone who wants decent CUDA
    performance with Adobe apps should use (one or more GTX 580 3GB or 780Ti is best).

    Also, hate to say it but showing results for using the card with OpenCL but not
    showing what happens to the relevant test times when the 1800 is used for CUDA
    is a bit odd...

    Ian.

    PS. I see the messed-up forum posting problems are back again (text all squashed
    up, have to edit on the UK site to fix the layout). Really, it's been months now, is
    anyone working on it?

  • 0 Hide
    Draven35 , January 7, 2014 1:48 PM
    Quote:
    If you've got 3ds max, why don't you use something more serious/advanced like Mental Ray ? The default renderer tech represent distant past like year 1995.


    The 3dsMax test does use mental ray. Our Maya render test also uses mr, and the other Max render test uses VRay.

  • 0 Hide
    ddpruitt , January 7, 2014 7:22 PM
    Quote:
    Our Core i7 and dual Sandy Bridge-EP-based Xeons score similarly. Meanwhile, the -2687W v2 crushes this test.
    There are a number of reasons this could occur, without knowing what the exact input and output are any attempts to explain this are guessing (my first guess is that the output is different, bug perhaps). The fact that these benchmarks are run the same way as Tom's runs desktop CPU benchmarks shows that whomever was running these really didn't know what they were doing. These CPUs are fast but this article does them a disservice. I would like to see power measurements when the systems run for the same workload and are running for the same period of time. True they are rarely loaded 24x7 but they are generally running 24x7. The workloads should also match the hardware. These things had 64Gb RAM and not a single one of the benchmarks would have stressed that amount of RAM appropriately. The problem is that if your actually reading and writing this much RAM it can have an impact on power consumption.
  • 0 Hide
    Draven35 , January 7, 2014 7:46 PM
    We don't have any tests specifically oriented towards 64 GB of RAM, but the AE test will use as much as you throw at it. At some point though, the AE test runs into the storage limitation, which is what i suspect was happening here. I also don't know if Chris was using the SD or HD version of the AE test. If the machine had 128 GB of RAM, he could have run the AE test from a RAM disk, thus removing storage from the equation. Without just throwing random items into a scene, it is difficult to create a test meant for 64 GB systems without a 64 GB system to create the test on- some of these tests were originally created on the HP z400 we reviewed a few years back, and the newer Maya and Max tests I use on workstations were authored on our baseline workstation. I am now developing updated test s on an HP z600 with 12 GB of RAM... but the extra RAM will be nearly inconsequential in the actual tests because I'll still need to be able to run them on the baseline workstation (8 GB RAM). My personal feeling is that the Premiere test is getting a little long in the tooth, as is the AE test (hence why i went back and redid it for HD).
  • 2 Hide
    oxiide , January 7, 2014 9:12 PM
    I know this chip is absolutely inappropriate for gaming, but it would still be somewhat interesting to me to see how it stacks up in the usual gaming benchmarks. Multithreading in games is getting a little better, and the results might show Intel's potential if they weren't so heavily focused on power consumption in their recent consumer-level products. On the other hand, it might be completely predictable and boring. But I think it would still be interesting to find out for sure.

    Again, obviously, I know the productivity benches are what's important here. I know no one's gaming on a server processor, like ever. But while you've got a review sample, why not experiment a little? :) 

    Great review as always.
  • 0 Hide
    puppetMaster3 , January 7, 2014 9:27 PM
    What is 'segmented' models? It is all over the article and not explained.
  • 0 Hide
    Crass Spektakel , January 8, 2014 2:07 AM
    Right now I am using a system build in 2008 with two Xeon 5450 (thats basically eight cores at 3.0Ghz from the last Core2 Q9000 series), 16GB RAM and a Radeon 280X for Work, Development and also Gaming. Runs quite well, even a slightly overclocked i7-4770 doesn't beat this old war horse (at least not always). The i7-49xx on the other hand is quite superior to my old Xeon, especially when overclocked (My Xeon board offers not overclocking at all).
  • 0 Hide
    daglesj , January 8, 2014 3:30 AM
    Yes I'm running a 2008 Dell Precision T5400 with a single 2.8GHz quad Xeon (soon to be a double 2.8Ghz quad setup this week) with 16GB of ECC ram and 1TB SSHD drives.Fantastic machine. Takes everything I throw at it and will take even more once I get up to 8 cores.Might treat it to a matched pair of 3.3GHz chips in the summer.
  • 1 Hide
    Draven35 , January 8, 2014 4:01 AM
    Quote:
    I love Xeon machines. As they are not mainstream you can usually pick up crazy spec Xeon workstations for next to nothing just a few years after they were going for $3000. They make damn good workhorses.


    I have an HP z600 with 2x 2.26 Ghz Xeon 5520s and 12 GB RAM, 2x 500 GB hard drives... total invested: $550. Its my personal 3d machine and benchmark development machine. Going to put it up to 24 GB shortly.
  • 1 Hide
    DoDidDont , January 8, 2014 4:10 AM
    I think for people that already own a pair of E5-2687W processors, the upgrade to the V2 version isn’t really worth it from a price point. If its purely for work purposes and multi-threaded apps then the E5-2690V2 being in the same price bracket, would be a better choice. Very slight drop in clocks, but more cores. Shame Tom’s didn’t include the 2690V2 and 2697V2 in the multithreaded tests, for people considering whether or not to upgrade their SandyBridge platforms. Definitely something wrong with Tom’s CineBench results.Every benchmark I have seen online scores a pair of E5-2687w’s at around the 26 mark for the multiple test, and my own E5-2687w’s lowest score has been 25.6 and highest 26.22, so I’m guessing the results should be 26 for the E5-2687w and 28 for the V2.
  • 0 Hide
    Michael Robinson , January 8, 2014 5:13 AM
    I still find it difficult understanding where you'd use them though. Sure the idea of 12 cores on two CPU's (and I assume 48 threads) sounds like a geek dream but with so much graphic work capable of being offloaded onto the GPU I'm not sure what the point is. It would of been interesting to see some figures for CPU bound games as well - I know these things aren't for gaming but it allows us non-workstation users to be able to contrast and compare. It would also of been interesting to try and run tests with various programs (and limiting the cores being used) to see whether huge cache sizes make much of a difference and how. A similar thing could be done for the number of cores to see what games really can use each and every core.
Display more comments