The Intel Skylake-X Review: Core i9 7900X, i7 7820X and i7 7800X Tested
by Ian Cutress on June 19, 2017 9:01 AM ESTComparing Skylake-S and Skylake-X/SP Performance Clock-for-Clock
If you’ve read through the full review up to this point (and kudos), there should be three things that stick in the back of your mind about the new Skylake-SP cores: Cache, Mesh and AVX512. These are the three main features that separate the consumer grade Skylake-S core from this new core, and all three can have an impact in clock-for-clock performance. Even though the Skylake-S and the Skylake-SP are not competing in the same markets, it is still poignant to gather how much the changes affect the regular benchmark suite.
For this test, we took the Skylake-S based Core i5-6600 and the Skylake-SP based Core i9-7900X and ran them both with only 4 cores, no hyperthreading, and 3 GHz on all cores with no Turbo active. Both CPUs were run in high performance modes in the OS to restrict any time-to-idle, so it is worth noting here that we are not measuring power. This is just raw throughput.
Both of these cores support different DRAM frequencies, however: the i5-6600 lists DDR4-2133 as its maximum supported frequency, whereas the i9-7900X will run at DDR4-2400 at 2DPC. I queried a few colleagues as to what I should do here – technically the memory support is an extended element of the microarchitecture, and the caches/uncore/untile will be running at different frequencies, so how much of the system support should be chipped away for parity. The general consensus was to test with the supported frequencies, given this is how the parts ship.
For this analysis, each test was broken down in two ways: what sort of benchmark (single thread, multi-thread, mixed) and what category of benchmark (web, office, encode).
For the single threaded tests, results were generally positive. Kraken enjoyed the L2, and Dolphin emulation had a good gain as well. The legacy tests did not fair that great: 3DPM v1 has false sharing, which is likely taking a hit due to the increased L2 latency.
On the multithreaded tests, the big winner here was Corona. Corona is a high-performance renderer for Autodesk 3ds Max, showing that the larger L2 does a good job with its code base. The step back was in Handbrake – our testing does not implement any AVX512 code, but the L3 victim cache might be at play here over the L3 inclusive cache in SKL-S.
The mixed results are surprising: these tests vary with ST and MT parts to their computation, some being cache sensitive as well. The big outlier here is the compile test, indicating that the Skylake-SP might not be (clock for clock) a great compilation core. This is a result we can trace back to the L3 again, being a smaller non-inclusive cache. In our results database, we can see similar results when comparing a Ryzen 7 1700X, an 8-core 95W CPU with 16MB of L3 victim cache, is easily beaten by a Core i7-7700T, with 4 cores at 35W but has 8MB of inclusive L3 cache.
If we treat each of these tests with equal weighting, the overall result will offer a +0.5% gain to the new Skylake-SP core, which is with the margin of error. Nothing too much to be concerned about for most users (except perhaps people who compile all day), although again, these two cores are not in chips that directly compete. The 10-core SKL-SP chip still does the business on compiling:
If all these changes (minus AVX512) offer a +0.5% gain over the standard Skylake-S core, then one question worth asking is what was the point? The answer is usually simple, and I suspect involves scaling (moving to chips with more cores), but also customer related. Intel’s big money comes from the enterprise, and no doubt some of Intel’s internal metrics (as well as customer requests) point to a sizeable chunk of enterprise compute being L2 size limited. I’ll be looking forward to Johan’s review on the enterprise side when the time comes.
Facebook
Twitter
RSS
264 Comments
View All Comments
Soheil - Sunday, June 25, 2017 - link
Anyone knows why 1600X better than 1800X?OddFriendship8989 - Thursday, June 29, 2017 - link
I'm late here as usual but why are you not comparing against the 7700k and 7600k? I get that these are HEDT chips, but it's worth comparing against the high end mainstream especially when the 7800x and 7700k are priced similarly that someone MIGHT consider jumping over.I hate to say it but this is the typical stuff you guys used to do, and I know it takes more time to put together more CPUs, but logical comparisons MUST be made and these charts show a bit of laziness.
ashlol - Friday, June 30, 2017 - link
can we have the GPU tests pleaseOxford Guy - Saturday, July 1, 2017 - link
"The discussion on whether Intel should be offering a standard goopy TIM or the indium-tin solder that they used to (and AMD uses) is one I’ve run on AnandTech before, but there’s a really good guide from Roman Hartung, who overclocks by the name der8auer. I’m trying to get him to agree to post it on AnandTech with SKL-X updates so we can discuss it here, but it really is some nice research. You can find the guide over at http://overclocking.guide."If you have a point to make then make it. After all, you said you've already "run" this discussion before. Tell us why polymer TIM is a better choice than solder (preferably without citing cracks from liquid nitrogen cooling).
ashlol - Monday, July 3, 2017 - link
Anyway both are bad since you have to delid it to achieve good cooling. I have delidded a 4770k and a 6700k and put liquid metal TIM between the die and the IHS and they both run 15°C cooler at 4.6-4.7GHz@60°C with custom loop. And from seeing the temperature under overclock I will have to delid those skylake-x too.parlinone - Tuesday, July 4, 2017 - link
What I find most shocking is a $329 Ryzen 1700 outperforms a $389 7800X at Cinebench...for less than half the power.The performance to power ratio translates to 239% in AMD's advantage. That's unprecedented, and I never imagined to see that day.
dwade123 - Thursday, July 6, 2017 - link
Only in Cinebench and AES is where Ryzen look good. 7800x beats the 1800x in everything else in this review. Ryzen is too inconsistent in both productivity and gaming. It is priced accordingly to that and not out of good faith from AMD. This is also the reason why Coffee Lake will only top out at 6 cores. because it can consistently beat the best Ryzen model.IGTrading - Friday, July 14, 2017 - link
I absolutely disagree with the conclusion. The correct conclusion can only drawn when comparin apples to apples. Oh, if you want to be objective and compare apples to oranges, you can't just take into considerantion today's benchmark results and price. Have we forgotten about the days we REVIEWERS were complaining about the high power consumption of Pentium 4 and Pentium D ?! What about the FX 8350 ?! Is power consumption not an objective metric anymore?! What about platform price ?! What about price/performance?! Why do some people get suddenly blinded by marketing money?! Conclusion: i-7900X is the highest performance the home power user can get today if money for the CPU , mobo and subsequent power consumption are not an issue. Comparing apples to apples or core for core, the i7820X clearly shows Intel's anxiety with Zen. The i7820X consumes 40% more than the AMD 1800X and costs 20% more while its motherboard is 200% the price. So paying all these heaps of money, CORE for CORE the Intel 7820X is a bit faster in some benchmarks, as it should be considering the power consumption and price you pay, EQUAL in a few benchmarks and SLOWER in a few other benchmarks. Would you pay the serious extra money for this ?! And put up with the 40% higher power consumption and heat generation ?! Come ooooon ...azulon1 - Sunday, July 16, 2017 - link
Wow how exactly is this fair that Intel gets a pass for gaming, because there were problems with the problem with the platform. If I remember Rison also had a problem with gaming. But it didn’t stop you guys then did it. don’t group me into AMD fanboy, But why such a bais?Soheil - Saturday, July 29, 2017 - link
no one answer to me? why 1600X better than 1700 and 1700X and some time better than 1800X?what about 1600? is good like as 1600X for gaming or not?