Intel Xeon W-2295 – Review 2020
Want to build a long-lasting, eminently powerful desktop PC that can serve as a platform for demanding tasks like multimedia editing or financial modeling? An enthusiast-class Intel Core X-Series or AMD Ryzen Threadripper are your most obvious choices for a CPU to serve as its brains. A third option, reviewed here, is a workstation-class CPU like the Intel Xeon W-2295. This $1,333 chip is very similar to the Intel Core i9-10980XE Extreme Edition, but it adds niche security and system stability features that some professional software applications might require. Unless these features are essential, however, we recommend that builders of high-end desktop (HEDT) PCs choose the Ryzen Threadripper 3960X or 3970X instead.
Xeon Versus Core
Introduced last fall, the Xeon W-2200 CPU family uses the same “Cascade Lake” microarchitecture as the 10th Generation Core X-Series chips, such as the Core i9-10980XE. That means the same 14-nanometer production process and the same basic features, like Hyper-Threading to double the number of instructions per core, and support for PCI Express Gen 3 (but not Gen 4) for interfacing with add-in cards.
In the Xeon W-2295’s case, the similarities also extend to performance specs, too. Both it and the Core i9-10980XE have 18 cores, support for up to 36 concurrent processing threads, and a base clock speed of 3GHz. They also share a boost clock speed of 4.6GHz and a total L3 cache size of 24.75MB. The W-2295 supports up to 72 PCI Express lanes, for extreme configs involving multiple video cards or banks of PCI Express SSDs.
Similarities aside, the Xeon chips offer a few specialized features that Core X-Series CPUs do not. The Xeon W-2295 has support for error-correcting code (ECC) DDR4 memory. ECC is a workstation-grade feature that reduces the possibility of errors while your computer is reading and writing data from memory.
That sounds great on the surface—who wouldn’t want to reduce errors? In practice, however, using ECC may result in a small performance hit, and the advantages vary greatly. Certain specialized applications, especially financial and scientific modeling programs, are extremely sensitive to memory errors, while most everyday apps like web browsers will see no benefit. (If you’re a civil architect doing stress calculations on a highway-bridge design: ECC all the way.)
In the case of the W-2295, the memory support is quad-channel at up to 2,933MHz, and the CPU supports up to 1TB assuming the motherboard in question does, too. The Xeon W-2295 also supports Intel’s vPro remote management and security capabilities. Like ECC, vPro can be essential in some situations, especially in large deployments of PCs by an IT department. If you’re just building a single workstation for your home office, however, vPro is likely less critical.
As niche as the Xeon W-2295’s added features may be, they’re not available on the enthusiast-class Core X chips. If you know you need vPro or ECC, a Xeon CPU is one of your only paths to getting them on an HEDT platform. It is possible to pair a third-generation Ryzen Threadripper CPU with ECC memory, but few motherboards currently support that configuration. As a result, Intel is marketing the Xeon W-2200 processor family to data scientists, engineers, and visual-effect artists, and other professionals who need a powerful, stable chip to perform simulations and renderings.
A Word About Motherboard Differences
In addition to sharing the same microarchitecture, the Xeon W-2295 and Intel Core i9-10980XE also share the same LGA 2066 socket for connecting to a desktop motherboard. The name comes from the 2,066 pins inside the socket. However, the two chips do not share the same chipset—the hardware that enables information to flow between the CPU and the rest of a PC’s components like the hard drive, graphics card, and memory. The Xeon W-2295 uses the C422 chipset, while the Core i9-10980XE uses the X299 chipset. This is important when you’re choosing a motherboard for your new PC, since motherboards using different chipsets on the same socket are sometimes all compatible with a given CPU. Not so here.
Motherboard compatibility is also important for two more reasons: keeping the cost of your system under control, and ensuring you have a wide range of board options from which to choose. A quick glance at current prices shows that plenty of C422 and X299 motherboard models start at around $300, though many of the cheapest do not support ECC memory. The range of board options is one advantage that Xeon and Core X-Series chips currently have over their Ryzen Threadripper competitors, which require expensive and still relatively scarce TRX40 motherboards for the newest (third-generation) chips.
Apart from the motherboard, the cooling system and power supply are the two other major factors to consider when shopping for a new CPU. The Xeon W-2295 has the same 165-watt thermal design power (TDP) requirement as the Core i9-10980XE. While Intel does include a cooling fan with the Xeon W-2295, people who want to ensure that their CPU runs at peak performance consistently will probably want to use an upgraded, third-party liquid cooling system.
The Xeon W-2295 lacks an integrated graphics processor, but most prospective owners will want to invest in a more powerful workstation-class graphics card, in any case, like an Nvidia Quadro, so its absence is no great deficiency. (See our picks for the best graphics cards.)
Computing Performance
To test the Xeon W-2295’s performance, we used a workstation supplied by Boxx Technologies. In addition to the Xeon CPU, the system has 64GB of memory, a 960GB Intel Optane SSD 905P, and an Nvidia Quadro RTX 5000 graphics card. For the reasons mentioned above, the X299 testbed that we used to evaluate the Core i9-10980XE and other Core X-Series CPUs over the last few years isn’t compatible with the Xeon W-2295.
As a result, the performance numbers that follow should be taken with a grain of salt, especially when comparing the two Intel chips. In addition to the Xeon and the X-Series Core i9, I’ve also included some competing AMD Ryzen CPUs, including our current Editors’ Choice among HEDT CPUs, the Ryzen Threadripper 3970X. Two less-powerful and less-expensive mainstream CPUs, the AMD Ryzen 9 3950X and the Core i9-9900KS, are also included to help illustrate the performance differences among mainstream, enthusiast, and workstation CPUs.
Overall, the Xeon W-2295 offered very similar performance results on our benchmark tests to those of the Core i9-10980XE, with one chip performing better on some tests and the other on the remainder. The Xeon W-2295 also offers good performance when only a single CPU core is required—in some cases, it outpaced its Ryzen competitors on single-core tests. However, it falls significantly behind both Ryzen Threadripper CPUs on essential multi-core workloads like rendering 3D images.
Our first test, Cinebench R15, demonstrates performance on demanding multimedia workflows. It’s a CPU-centric test that gauges both the single-core performance and the multi-core performance of a processor. The resulting scores are proprietary numbers that represent the CPU’s capabilities while rendering a complex 3D image. They show that the Core i9-10980XE and Xeon W-2295 perform roughly equally, but that both are far slower than the Ryzen Threadripper chips when it comes to multi-core performance.
This is unsurprising, because the Ryzen Threadripper 3970X has a whopping 32 cores and 64 threads, far more than the Xeon W-2295 has.
The results of our 7-Zip file-compression test, which is less demanding than Cinebench but also depends heavily on the number of CPU cores, mirror those of Cinebench.
Different Apps, Different Results
The results of our Handbrake video conversion test are much more surprising. It’s another tough, threaded workout that’s highly CPU-dependent and scales well with cores and threads. In this case, the Xeon W-2295 performed significantly worse than all of the other chips, including its Core i9-10980XE sibling.
While Handbrake is a widely-used utility, it’s not the sort of task for which Xeon processors are designed, and we’ve frequently seen large discrepancies when testing other workstation-class systems and CPUs on this test. It’s possible that the unique features of the Xeon platform, especially ECC memory, are hampering performance here.
Our ray-tracing test has multi-core and single-core components, but it’s less susceptible to core and thread counts than Cinebench and Handbrake are. The test uses the built-in benchmark of the POV-Ray program, which digital artists can use to create ray-traced scenes. The differences on the all-cores setting are very minor, with an 8 percent difference between the lowest and highest results.
The differences on the single-core test are far more pronounced, however, and both of the Threadripper chips come out significantly ahead.
Unsurprisingly, on less-demanding tests like rendering a simple graphic of a squirrel in Blender, or converting audio files using an old version of Apple’s iTunes test, nearly all of the CPUs perform roughly the same.
As a final test of the Xeon W-2295’s performance, I ran SPECviewperf 13, the most realistic and challenging workstation benchmark we run and the one we give greatest weight. This test uses viewsets from actual independent software vendor (ISV) apps to render, rotate, and zoom in and out of wireframe and solid 3D models, with results listed in frames per second (higher is better). The viewsets we use are from PTC’s Creo computer-aided design (CAD) platform; Autodesk’s Maya modeling and simulation software for film, TV, and games; and Dassault Systemes’ SolidWorks 3D rendering package.
We don’t typically run these tests on non-workstation CPUs, so instead I’ll compare these numbers to the most theoretically powerful workstation we’ve ever tested: the Dell Precision 7920 Tower. It is an unfair comparison, since the Precision 7920 includes not one but two Intel Xeon Platinum 8260 processors (sporting 24 cores and 48 threads each) and 96GB of memory. But it actually turns out quite well for the Xeon W-2295, which scored 305 frames per second (fps) on the Creo viewset, 343fps on Maya, and 182fps on SolidWorks. The Precision 7920 tower scored 258fps on Creo, 329fps on Maya, and 157fps on SolidWorks.
Highly Specialized
The main reason to buy the Xeon W-2295 over its Core i9-10980XE sibling is if you need to run specialized software that requires unique features like ECC support. If you do, you’ll need to check with the maker of the software to see what its requirements are. It’s clear that the Xeon can perform quite well on demanding workstation tasks, while it falls far short on other workflows like video conversions using open-source consumer-level software.
If you don’t need the extra security or stability that the Xeon platform offers, it’s actually better to skip the Core i9-10980XE and choose the Ryzen Threadripper 3960X or 3970X instead for your next HEDT build. The performance results above help explain a bit why that’s the case, but make sure to read our full reviews of those CPUs for more on why they’re superior to the Core i9-10980XE.
Ultimately, the Xeon W-2295’s price-to-performance ratio makes it difficult to justify unless you specifically use software that requires it.
Intel Xeon W-2295 Specs
Core Count | 18 |
Thread Count | 36 |
Base Clock Frequency | 3 GHz |
Maximum Boost Clock | 4.6 GHz |
Unlocked Multiplier? | No |
Socket Compatibility | Intel LGA 2066 |
Lithography | 14 nm |
L3 Cache Amount | 24.75 MB |
Thermal Design Power (TDP) Rating | 165 watts |
Integrated Graphics | None |
Bundled Cooler | Intel Stock Cooler |