Intel’s Core i9-9900K Finally Available at Eye-Watering Prices
When Intel launched the Core i9-9900K, we noted that while the CPU was easily the best-performing (and best-priced, in terms of price-per-core) Intel chip we’d ever tested, its price/performance ratio didn’t hold up very well compared to AMD’s eight-core Ryzen 7 2700X. When we checked prices back on October 26, we saw the Core i9-9900K listing at $580 and said that this might not actually have much impact on the CPU’s attractiveness to its target market. Because the Core i9-9900K was easily the fastest and most capable CPU Intel had previously launched, even besting the 10-core Core i9-7900X in some tests, the company had some room on price when addressing customers who work primarily within the Intel ecosystem.
This kind of price inflation puts the Core i9-9900K squarely into the Core i9-7900X’s turf, and there’s not a clear winner between them — it varies, depending on which tests you run. The Core i9-99900K is generally equal to the 7900X, but there are certain tests, like Qt compiling, where the 10-core pulls ahead. Feel free to compare them in the slideshow from our 9900K review below.
We open with Cinebench R15 and some immediately interesting results. Chips like the Core i7-7700K are relatively competitive in single-threaded mode, but fall to less than half the Core i9-9900K's multi-threaded performance.
We've added 7zip's compression test by popular request in addition to the more-common decompression scenario. We have fewer data points for this graph — I only had time to retest just so many platforms — but this test is a decisive overall Intel win. While the Ryzen 7 2700X matches the Core i7-8086K well, Threadripper is only modestly ahead of the Core i9-9900K, despite packing twice the cores.
When it comes to decompression, AMD's CPUs are on vastly stronger footing. Ryzen 7 2700X is within six percent of the Core i9-9900K in this test, while the Threadripper 2950X leaves every Intel chip in the dust.
The Ryzen 7 2700X is roughly 12 percent behind the Core i9-9900K in H.264, with the 8086K improving on the 8700K by a few percent. The 9900K is faster than even the 7900X courtesy of its much higher clock speeds and diminishing marginal returns from scaling. It's a full 30 percent faster than the Core i7-6900K. Only Threadripper is faster.
We have fewer samples in Handbrake H.265, but we see similar patterns. AMD's Ryzen isn't as strong in H.265 as it is in H.264 and it doesn't scale as well, either — Intel sees greater relative improvement moving from the 8086K to the 9900K than AMD does when shifting from Ryzen 7 2700X to Threadripper 2950X. That's not a result we expected, but we haven't had time to dig into the potential reasons why (and the 2950X's score in Handbrake 1.12 is not unusually slower than the other performance degradations we observed). The 9900K encodes H.265 video in 75 percent the time it takes the Ryzen 7 2700X.
We're transitioning to the new Blender benchmark, which includes several of the performance tests we've previously used. Gooseberry, however, is not part of that benchmark, so we're breaking it out separately. The Core i9-9900K shines here, outperforming the older 6900K by 21 percent, passing Ryzen 7 by 13 percent, and even matching the Core i9-7900X — not a result we expected. Only Threadripper and its commensurately higher price tag is faster.
Blender's benchmark is still in beta — we had one render crash on the 9900K but confirmed it was an application issue rather than a problem with the CPU. While the 9900K beats Ryzen 7 in all of these tests, the gap varies from as little as five percent to as much as 11 percent depending on the specifics of the scene in question. The 2700X is in-between the 8086K and the 9900K (closer to the latter than the former) while Threadripper takes the lead you'd expect a 16-core CPU to command.
We compile Qt using Microsoft Visual Studio Community Edition 2017. AMD's second-generation Ryzen performance improved markedly in this test compared to its first, with Ryzen 7 landing squarely between the 8086K and the 9900K. The 7900X is slightly faster than the 9900K, but not nearly enough to justify its price tag, while Threadripper takes the overall performance crown.
PCMark 10's Extended test shows a tight competition between Ryzen 7 2700X and the two six-core Intel CPUs, with the 9900K holding 11 percent overall lead over AMD. The test clearly doesn't scale past eight cores, with Threadripper turning in identical or slightly slower scores than the 2700X.
Dolphin is a Gamecube and Wii emulator and its CPU benchmark tests JIT compiler performance. Ryzen has never been a strong performer here — Dolphin is single-threaded and quite latency sensitive — and that opens a major window for the 9900K, which finishes the test in 69 percent of the time it takes the Ryzen 7 2700X. Broadwell-E is almost as poorly off as Ryzen here, while the 7900X, 7700K, 8700K, and 8086K are all closer to the 9900K.
We open with Cinebench R15 and some immediately interesting results. Chips like the Core i7-7700K are relatively competitive in single-threaded mode, but fall to less than half the Core i9-9900K's multi-threaded performance.
We've added 7zip's compression test by popular request in addition to the more-common decompression scenario. We have fewer data points for this graph — I only had time to retest just so many platforms — but this test is a decisive overall Intel win. While the Ryzen 7 2700X matches the Core i7-8086K well, Threadripper is only modestly ahead of the Core i9-9900K, despite packing twice the cores.
When it comes to decompression, AMD's CPUs are on vastly stronger footing. Ryzen 7 2700X is within six percent of the Core i9-9900K in this test, while the Threadripper 2950X leaves every Intel chip in the dust.
The Ryzen 7 2700X is roughly 12 percent behind the Core i9-9900K in H.264, with the 8086K improving on the 8700K by a few percent. The 9900K is faster than even the 7900X courtesy of its much higher clock speeds and diminishing marginal returns from scaling. It's a full 30 percent faster than the Core i7-6900K. Only Threadripper is faster.
We have fewer samples in Handbrake H.265, but we see similar patterns. AMD's Ryzen isn't as strong in H.265 as it is in H.264 and it doesn't scale as well, either — Intel sees greater relative improvement moving from the 8086K to the 9900K than AMD does when shifting from Ryzen 7 2700X to Threadripper 2950X. That's not a result we expected, but we haven't had time to dig into the potential reasons why (and the 2950X's score in Handbrake 1.12 is not unusually slower than the other performance degradations we observed). The 9900K encodes H.265 video in 75 percent the time it takes the Ryzen 7 2700X.
We're transitioning to the new Blender benchmark, which includes several of the performance tests we've previously used. Gooseberry, however, is not part of that benchmark, so we're breaking it out separately. The Core i9-9900K shines here, outperforming the older 6900K by 21 percent, passing Ryzen 7 by 13 percent, and even matching the Core i9-7900X — not a result we expected. Only Threadripper and its commensurately higher price tag is faster.
Blender's benchmark is still in beta — we had one render crash on the 9900K but confirmed it was an application issue rather than a problem with the CPU. While the 9900K beats Ryzen 7 in all of these tests, the gap varies from as little as five percent to as much as 11 percent depending on the specifics of the scene in question. The 2700X is in-between the 8086K and the 9900K (closer to the latter than the former) while Threadripper takes the lead you'd expect a 16-core CPU to command.
We compile Qt using Microsoft Visual Studio Community Edition 2017. AMD's second-generation Ryzen performance improved markedly in this test compared to its first, with Ryzen 7 landing squarely between the 8086K and the 9900K. The 7900X is slightly faster than the 9900K, but not nearly enough to justify its price tag, while Threadripper takes the overall performance crown.
PCMark 10's Extended test shows a tight competition between Ryzen 7 2700X and the two six-core Intel CPUs, with the 9900K holding 11 percent overall lead over AMD. The test clearly doesn't scale past eight cores, with Threadripper turning in identical or slightly slower scores than the 2700X.
Dolphin is a Gamecube and Wii emulator and its CPU benchmark tests JIT compiler performance. Ryzen has never been a strong performer here — Dolphin is single-threaded and quite latency sensitive — and that opens a major window for the 9900K, which finishes the test in 69 percent of the time it takes the Ryzen 7 2700X. Broadwell-E is almost as poorly off as Ryzen here, while the 7900X, 7700K, 8700K, and 8086K are all closer to the 9900K.
The results speak for themselves. AVX-512 support in the Core i9-7900X might make it a must-have if you need that compatibility and the Core i9-9900K has slightly lower motherboard costs, but neither chip holds a candle in multi-threading to Threadripper in anything but H.265 encoding, where our results do show the Core i9-9900K still holding an edge over Threadripper, mostly because H.265 encoding doesn’t appear to scale all that well above eight cores. In every test that does scale well above eight cores, including some tests that nominally favor Intel, Threadripper wins.
When we spoke to Intel before the Core i9-9900K launch, the company assured us that it would prioritize high-end, high-value chips like the Core i9-9900K to keep them flowing, even though the company has been working through supply constraints. This type of availability isn’t exactly what we had in mind, and the situation risks turning into a repeat of the Core i7-8700K launch last year, when that chip was technically available but practically impossible to find in-channel at reasonable prices or in anything approaching volume.
Continue reading
Intel’s Raja Koduri to Present at Samsung Foundry’s Upcoming Conference
Intel's Raja Koduri will speak at a Samsung foundry event this week — and that's not something that would happen if Intel didn't have something to say.
Ryzen 9 5950X and 5900X Review: AMD Unleashes Zen 3 Against Intel’s Last Performance Bastions
AMD continues its onslaught on what was once Intel's undisputed turf.
Leaked Benchmarks Paint Conflicting Picture of Intel’s Rocket Lake
Rumors about Rocket Lake have pointed in two opposite directions recently, but the more competitive figures are more likely to be true.
Intel’s Iris Xe Max Discrete GPU Is Slower Than the Integrated Version
Intel's Iris Xe Max has debuted, but the discrete GPUs performance is a bit odd, as new benchmarks show.