Romanian publication Lab501 has published the first review of the Intel Core i9-10980XE, and it is apparently unsanctioned by Intel. The 18-core, 36-thread processor is the flagship of the chipmaker's Cascade Lake-X family, which Intel announced earlier this month.
It's uncertain how the publication obtained the unreleased processor, or whether it's an engineering sample (ES) or retail chip. It's also unlikely that this sample is using production-class firmware. Lab501 paired the Core i9-10980XE with a Gigabyte X299 Aorus Master motherboard, 32GB (4x8GB) of G.Skill Sniper X DDR4-3200 memory with CL14-14-14-36 timings, and an Nvidia GeForce RTX 2080 Ti graphics card.
Core i9-10980XE Application Performance
Benchmark | Core i9-10980XE | Ryzen 9 3900X | Core i9-9900K |
---|---|---|---|
PCMark 10 | 7,214 points | 7,571 points | 7,112 points |
WinRAR 5.8 | 24,656 KB/s | 28,156 KB/s | 28,804 KB/s |
7-Zip | 137,310 MIPS | 110,845 MIPS | 68,123 MIPS |
Lab501 inexplicably did not include Threadripper processors in its test pool. Bear in mind, those are the -10980XE's natural competitors.
Lab501's results show the Core i9-10980XE falling behind the Ryzen 9 3900X in PCMark 10. When it comes to compression workloads, WinRAR 5.8 apparently favors the Ryzen 9 3900X, while 7-Zip puts the Core i9-10980XE ahead.
Benchmark | Core i9-10980XE | Ryzen 9 3900X | Core i9-9900K |
---|---|---|---|
3ds Max 2020 | 518 seconds | 745 seconds | 934 seconds |
Blender 2.8 | 421 seconds | 483 seconds | 636 seconds |
DaVinci Resolve 15 | 96 seconds | 109 seconds | 165 seconds |
HandBrake 1.2.2 | 45 seconds | 46 seconds | 55 seconds |
Cinebench R20 Single-Core | 458 points | 522 points | 511 points |
Cinebench R20 Multi-Core | 8,563 points | 7,076 points | 4,935 points |
POV-Ray 3.7 Single-Core | 473 points | 492 points | 517 points |
POV-Ray 3.7 Multi-Core | 7,303 points | 6,151 points | 4,276 points |
The Core i9-10980XE has six more cores than the Ryzen 9 3900X, and this difference is enough to give the Intel chip the advantage with software that embraces cores, such as 3ds Max 2020, Blender 2.8, and DaVinci Resolve 15. In HandBrake 1.2.2., however, the Core i9-10980XE is only one second faster than the Ryzen 9 3900X.
On the flipside, the Ryzen 9 3900X has a better single-core performance than the Core i9-10980XE. The Ryzen 9 3900X performs up to 14% and 4% faster than the Core i9-10980XE in Cinebench R20 and POV-Ray 3.7 single-core tests, respectively.
The tides turn in the Core i9-10980XE's favor in the multi-core tests. The six extra cores gives the Core i9-10980XE a 21% and 18.7% lead in Cinebench R20 and POV-Ray 3.7, respectively.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Core i9-10980XE Gaming Performance
Game | Core i9-10980XE | Ryzen 9 3900X | Core i9-9900K |
---|---|---|---|
Grand Theft Auto V | 114 FPS | 124 FPS | 126 FPS |
Far Cry New Dawn | 90 FPS | 106 FPS | 121 FPS |
Shadow of the Tomb Raider | 175 FPS | 173 FPS | 176 FPS |
Middle-earth: Shadow of War | 151 FPS | 146 FPS | 158 FPS |
Warhammer 40,000: Dawn of War III | 116 FPS | 130 FPS | 134 FPS |
Batman: Arkham Knight | 184 FPS | 181 FPS | 184 FPS |
Sleeping Dogs | 249 FPS | 262 FPS | 280 FPS |
Metro Exodus | 91 FPS | 95 FPS | 93 FPS |
Ashes of the Singularity | 98 FPS | 99 FPS | 98 FPS |
Deus Ex: Mankind Divided | 100 FPS | 106 FPS | 114 FPS |
Lab501 has a pretty extensive gaming test suit. For the sake of simplicity, we'll look at the 1920 x 1080 results. It's evident that gaming favors processors with high operating clocks, so the Core i9-9900K continues to be the fastest gaming processor on the planet. Nonetheless, if we compare the Core i9-10980XE and the Ryzen 9 3900X, the results reveal that AMD's chip is better at gaming.
Due to the significant difference in cores, it's a no-brainer that the Core i9-10980XE would come out on top in software that scales with cores and memory bandwidth. The Ryzen 9 3900X undoubtedly put up a respectable fight. At the same time, we're more interested in seeing what the Ryzen 9 3950X can do, although given the segment in which the Core i9-10980XE is in, it would only be fair to compare it against AMD's Threadripper 3000-series offering.
Zhiye Liu is a news editor and memory reviewer at Tom’s Hardware. Although he loves everything that’s hardware, he has a soft spot for CPUs, GPUs, and RAM.
AMD crafts custom EPYC CPU with HBM3 memory for Microsoft Azure – CPU with 88 Zen 4 cores and 450GB of HBM3 may be repurposed MI300C, four chips hit 7 TB/s
AMD-powered El Capitan is now the world's fastest supercomputer with 1.7 exaflops of performance — fastest Intel machine falls to third place on Top500 list
-
justin.m.beauvais Yeah, buuuuuuuut the Ryzen 9 3900X and the 9900K are very much NOT the competition for this CPU... AT ALL. This thing is going to be priced more similarly (I'm saying generously, it is likely to be a lot more) to a 32 core Threadripper if Intel stays true to form. Why not just write an article about its performance? Why compare it to a CPU that isn't out yet? Why compare it to CPUs that are simply NOT the competition?Reply
Just write an article like "The first review of Intel's Core i9-10980XE 18-core chip has popped up in the wild" and talk about the strengths of the chip. You have a huge library of data to pull from to do some general comparisons, use that.
My theory is that Intel sent them one to test, under the table. They probably didn't want a repeat of their last benchmark fiasco. -
kinggremlin justin.m.beauvais said:Yeah, buuuuuuuut the Ryzen 9 3900X and the 9900K are very much NOT the competition for this CPU... AT ALL. This thing is going to be priced more similarly (I'm saying generously, it is likely to be a lot more) to a 32 core Threadripper if Intel stays true to form. Why not just write an article about its performance? Why compare it to a CPU that isn't out yet? Why compare it to CPUs that are simply NOT the competition?
Just write an article like "The first review of Intel's Core i9-10980XE 18-core chip has popped up in the wild" and talk about the strengths of the chip. You have a huge library of data to pull from to do some general comparisons, use that.
My theory is that Intel sent them one to test, under the table. They probably didn't want a repeat of their last benchmark fiasco.
From the review:
"Unfortunately, I couldn't compare it directly to an equivalent processor, for several reasons. Thus, AMD Ryzen 9 3950X and AMD Ryzen Threadripper 3rd gen have not been launched yet, Intel Core i9 9980XE has not been tested according to the new methodology, and those at Intel have refused to provide it to us, and the top of the range AMD , The Ryzen Threadripper 2990WX is currently locked in a long-term test by my colleague Matei. "
You can drop the conspiracy theories. They couldn't get their hands on anything else. If Intel shipped them one for a positive review, the website missed the memo. You should probably read the article conclusion. The only positive thing they had to say about the chip was the price reduction vs the last generation. -
cryoburner
The 3900X is definitely worth testing it against, since it's the closest processor available at this time to the upcoming 16-core 3950X that should be launching in a matter of weeks, as well as upcoming Threadrippers. Assuming these results are indicative of final performance, there's a very good chance that the 3950X will outperform this processor in nearly all software.justin.m.beauvais said:Why compare it to CPUs that are simply NOT the competition?
Of course, you do get things like additional memory channels and PCIe lanes on an HEDT platform, which could allow the 10980XE to pull ahead in certain workloads, but the next-generation Threadripper platform should offer those things as well.
Intel had to slash the prices of these processors to keep them competitive, and as such, this 18-core model will be priced at just $1000, similar to what prior 16-core Threadrippers launched for, and half of what their existing 18-core 9980XE is selling for. So I would not expect a 32-core Threadripper to be priced at a similar level. The 24-core model might be though.justin.m.beauvais said:This thing is going to be priced more similarly (I'm saying generously, it is likely to be a lot more) to a 32 core Threadripper if Intel stays true to form. -
hannibal Yeah! Threatripper 3000 will rip this apart ;)Reply
Competition is good! These will com down in price somewhat... maybe... it is intell, who knows... -
Olle P
It's mentioned. Just not available yet.nofanneeded said:excuse me but where is the
Ryzen 9 3950X in this whole article ???
Available to the reviewer, yes.cryoburner said:The 3900X is definitely worth testing it against, since it's the closest processor available at this time...
As mentioned the closest competitor of the current generation Threadripper was being used for a long-term test.
One irritating detail in the review is that it's not mentioned if the archiving software was used for compression or decompression. Compression is a bit more demanding but decompression is used way more often. Ryzen is known to not handle 7-zip compression that well but excel at decompression. -
jimmysmitty justin.m.beauvais said:Yeah, buuuuuuuut the Ryzen 9 3900X and the 9900K are very much NOT the competition for this CPU... AT ALL. This thing is going to be priced more similarly (I'm saying generously, it is likely to be a lot more) to a 32 core Threadripper if Intel stays true to form. Why not just write an article about its performance? Why compare it to a CPU that isn't out yet? Why compare it to CPUs that are simply NOT the competition?
Just write an article like "The first review of Intel's Core i9-10980XE 18-core chip has popped up in the wild" and talk about the strengths of the chip. You have a huge library of data to pull from to do some general comparisons, use that.
My theory is that Intel sent them one to test, under the table. They probably didn't want a repeat of their last benchmark fiasco.
Well it will depend on pricing. If AMD keeps it similar to TR2 then this would price near AMDs 16 core 32 thread TR3. I doubt they will, since Intels pricing is out on these AMD will price accordingly.
What I want to know is why is the 9980XE is not present to see if any gains have been made if at all. -
PCWarrior Three things worth pointing out:Reply
1. This review from this Romanian website is about an engineering sample of a stock 10980XE tested on a previous gen X299 Gigabyte motherboard with an early bios for the new gen support (one month prior to actual launch). Gigabyte is notorious when it comes to early BIOSes.
2. Some cross checking against some 9980XE benchmarks shows that the reported 10980XE scores are worse than those of the 9980XE by about 3%. The 10980XE is expected to be better than a 9980XE by 5-10%, not worse by 3%.
3. There appears to be something off in this review. I find it very peculiar that in this review in the 10 gaming tests the 9900K won none against both the 9700K and 8700K at the same time. It instead got beaten by either the 8700K and/or 9700K in 8 out of the 10 cases and it only managed to draw twice. The 8700K alone won against the 9900K in 5 games and in 3 it was a draw. On the 5 occasions the 8700K won it was with a bigger margin than on the 2 where it lost.
1080p gaming:GTA V: 9700Κ:129, 8700Κ 128, 9900Κ:126Far Cry New Dawn: 9900K: 121, 8700K: 121, 9700K: 120
Rise of the Tomb Raider: 9900K: 176, 8700K: 176, 9700K: 175
Shadow of War: 8700K: 162, 9900K: 158, 9700K 155
Dawn of War 3: 8700K 145, 9900K: 134, 9700K: 134
Batman Arkham Knight: 9700K: 193, 9900K: 191, 8700K 189
Sleeping Dogs: 9700K: 287, 9900K: 280, 8700K: 276
Metro Exodus: 8700K: 101, 9700K 94, 9900K: 93Ashes of the Singularity: 9700K: 99, 9900K: 98, 8700K: 98
Deus EX:Mankind Divided: 8700K: 117, 9700K 115, 9900K 114I can understand why the 9700K could perform better on some games due to the lack of hyperthreading and 8 physical threads being more than enough for these games. But the 8700K winning against both the 9900K and the 9700K is very strange as it has both lower clock speeds and uses hyperthreading. And has fewer physical cores too. Probably some might say it might be due to the ring bus latency being smaller on a hexacore but I have never seen such a performance delta between a stock 9900K and a stock 8700K in any other review. In fact, the 9900K always prevailed against the 8700K in the vast majority of games. Never were these two cpus shown to trade blows, especially with the 8700K operating at stock and not being overclocked. -
joeblowsmynose PCWarrior said:Three things worth pointing out:
2. Some cross checking against some 9980XE benchmarks shows that the reported 10980XE scores are worse than those of the 9980XE by about 3%. The 10980XE is expected to be better than a 9980XE by 5-10%, not worse by 3%.
Yeah these numbers don't quite look right to me either, best to wait for proper reviews. It can't possibly be this underwhelming.