Intel Core i9-7980XE and Core i9-7960X Conclusion

In the 2000s, we had the frequency wars. Trying to pump all the MHz into a single core ended up mega-hurting one company in particular, until the push was made to multi-core and efficient systems. Now in the 2010s, we have the Core Wars. You need to have at least 16 cores to play, and be prepared to burn power like never before. With today’s launch, Intel has kicked their high-end desktop offerings up a notch, both in performance and power consumption.

Pun density aside, both Intel and AMD are pursuing a many-core strategy when it comes to the high-end desktop. Both manufacturers have the same story in their pockets: users need more multi-threaded performance, either for intense compute or for a heavy scoop of compute-based multi-tasking, to include streaming, encoding, rendering, and high-intensity minesweeper VR all at the same time. Due to having so many cores, single threaded performance while under load can be lower than usual, so both companies focus on throughput rather than responsiveness.

The Core i9-7980XE is Intel’s new top-of-the-line HEDT processor. Coming in at 18-cores and drawing 165W, this processor can reach 4.4 GHz at top turbo or 3.4 GHz on all-core turbo. At $1999, it becomes the most expensive (?) consumer focused processor on the market. It is priced at this level for two reasons: first such that it doesn’t cannibalize Intel’s enterprise sales which have a higher profit margin, but also due to Intel’s product stack it fills up several price points from $300 all the way to $2000 now, and with it being Intel’s best consumer processor, they are hoping that it will still sell like hot cakes.

Our performance numbers show that Intel’s best multi-core consumer processor is deserving of that title. In most of our multi-core tests, Intel has a clear lead over AMD: a combination of more cores and a higher single threaded performance compensates for any frequency difference. For anyone with hardcore compute, Intel gets you to the finishing line first in almost every scenario. AMD does win on a few benchmarks, which is something we saw when Johan tested the enterprise versions of Intel's and AMD's CPUs in his review, where he cited AMD’s FP unit as being the leading cause of the performance improvement.

There are going to be three cautionary flags to this tale based on our testing today, for anyone looking at Skylake-X, and they all start with the letter P: Power, Platform, and Price.

Power: In our testing, Intel’s cores can consume between 20W and 7.5W per core, which is a large range. When all cores are at load, as well as the mesh and DRAM controllers, the Core i9-7980XE draws over 190W, well above the TDP rating of 165W. This will cause concern for users that take the TDP value as rote for power consumption – and for any users thinking of overclocking it might also be worth investing in custom cooling loops. The processors from AMD consume ~177W at load, which for two cores less is about the same ballpark.

Platform: X299 motherboards are designed to handle Kaby Lake-X, Skylake-X LCC and Skylake-X HCC processors. Almost all the motherboards should be geared towards the high-end processors which makes platform compatibility a bit of a non-issue, however testing by others recommends some form of active cooling on the power delivery. When investing $1999 in a processor, especially if a user is considering overclocking, it is likely that a good motherboard is needed, and not just the bargain basement model. Some users will point to the competition, where AMD's processors offer enough lanes for three x16 GPUs and three PCIe 3.0 x4 storage devices from the processor at the same time, rather than reduced bandwidth for 3-way and requiring storage to go through the chipset.

Price: $1999 is a new record for consumer processors. Intel is charging this much because it can – this processor does take the absolute workstation performance crown. For high performance, that is usually enough – the sort of users that are interested in this level of performance are not overly interested in performance per dollar, especially if a software license is nearer $10k. However for everyone else, unless you can take advantage of TSX or AVX-512, the price is exorbitant, and all arrows point towards AMD instead. Half the price is hard to ignore.

Users looking at the new processors for workstation use should consider the three Ps. It’s not an easy task, and will highly depend on the user specific workflow. The recommendations ultimately come down to three suggestions:

  • If a user needs the top best workstation processor without ECC, then get Skylake-X.
  • If a user needs ECC or 512GB of DRAM, Xeon-W looks a better bet.
  • If a user has a strict budget or wants another GPU for compute workloads, look at Threadripper.

For those weighing up the Core i9-7960X compared to the Core i9-7980XE, part of me wants to say ‘if you’re going for cores and prepared to spend this much, then go all the way’. If the main purpose is throughput, for the benchmarks that matter, the 7980XE is likely to provide benefit. For multi-taskers, the benefits are less clear, and it would be interesting to get the Core i9-7940X and Core i9-7920X in for testing.

Performance Per Dollar Analysis
Comments Locked

152 Comments

View All Comments

  • CrazyHawk - Tuesday, September 26, 2017 - link

    "Intel also launched Xeon-W processors in the last couple of weeks."

    Just where can one purchase these mythical Xeon-W processors? There hasn't been a single peep about them since the "launch" week. I've only heard of two motherboards that will support them. They seem to be total vaporware. On Intel's own site, it says they were "Launched" in 3Q2017. Intel had better hurry up, 3Q will be up in 4 days!
  • samer1970 - Tuesday, September 26, 2017 - link

    I dont understand why intel disables ECC on their i9 CPU , they are losing low budget workstation buyers who will 100% choose AMD threadripper over intel i9..

    Even if they are doing this to protect their xeons chips ,they can enable non buffered ECC and not allow Registered ECC on the i9 problem solved. unbuffered ECC has Size limitation and people who want more RAM will go for xeons.

    Remember that their i3 has ECC support , but only the i3 ...

    intel , you are stupid.
  • vladx - Wednesday, September 27, 2017 - link

    Newsflash, these chips don't target "low budget workstation buyers". Golden rule is always: "If you can't afford it, you're not the target customer.".
  • samer1970 - Wednesday, September 27, 2017 - link

    Thats not a Golden Rule anymore with the Threadripper chips around . it is called "Stupid rule" ...

    They are allowing AMD to steal the low budget workstation buyers by not offering them an alternative to choose from.
  • vladx - Wednesday, September 27, 2017 - link

    The "low budget workstation buyers" as you call them are a really insignificant percentage of an already really small piece of the huge pie of Intel customers.
  • samer1970 - Wednesday, September 27, 2017 - link

    who told you so ? Most engineering students at universities need one , and Art Students who render alot as well. all these people will buy threadripper CPU and avoid intel , for intel xeon are 50% more expensive .

    andI dont cae about the percentage in intel Pie ... hundreds of thousands student enter uiviersites around the world each year . Low percentage or not they are alot ...

    how much do you think a low budget workstation costs ? they start from $3000 ... and with xeon Pricing , it will be very difficult to add alot of RAM and a good workstation card and fast SSD .
  • esi - Wednesday, September 27, 2017 - link

    What's the explanation for some of the low scores of the 7980XE on the SPECwpc benchmarks? Particularly Poisson, where the 6950X is 3.5X higher.
  • ZeDestructor - Wednesday, September 27, 2017 - link

    Most likely cache-related
  • esi - Wednesday, September 27, 2017 - link

    Maybe. But one that really makes no sense is the Dolphin 5.0 render test. How can the 7980XE take nearly twice as long as the 7960X?
  • esi - Wednesday, September 27, 2017 - link

    So I ran the Poisson benchmark on by 6950X. It uses all 10 cores (20 h/w threads), but can be configured to run in different ways: you can set the number of s/w threads per process. It then creates enough processes to ensure there's one s/w thread per h/w thread. Changing the s/w threads per processes significantly effects the result:

    20 - 1.34
    10 - 2.5
    5 - 3.31
    4 - 3.47
    2 - 3.67
    1 - 0.19

    Each process only uses about 2.5MB of RAM. So the 1-thread per process probably has a low result as this will result in more RAM usage than L3 cache, whereas the others should all fit in.

    Would be interesting to see what was used for the 7980/7960. Perhaps the unusual number of cores resulted in a less than optimal process/thread mapping.

Log in

Don't have an account? Sign up now