We recently had a look at the Xeon W-2155 10-core processor and today we’re moving up to almost double the cores. I have Intel‘s Xeon W-2195 (product link) on the testbench today and this is an 18-core/36-thread processor. We still stay within the LGA2066 package and socket which means that we get a huge amount of performance on a very little real-estate. Intel Xeon W processors deliver an optimised performance for the needs of mainstream workstation professionals. Among the highlights are hardware-enhanced workload performance, security, and reliability.
Xeon W Processor Series
The new Xeon W processor series consists of 8 SKUs, starting with a simple 4-core/4-thread version. From there, it moves over six, eight, ten and fourteen-core versions until it reaches the top of the series. At the top is the impressive W-2195 which has 18 cores and 36 threads as well as 24.75MB cache. It is the processor we’ll be putting through its paces today and it is quite the insane amount of cores. Especially when we consider that we are staying in a reasonable form factor.
The W-series is normal sized and not as large as the LGA3647 or TR4 sockets which we’ve seen lately. Instead, we are operating on the LGA2066 platform and that makes compatibility a lot easier. At least at this point in time. You won’t have trouble finding compatible cooling products, they are already widely available from most manufacturers.
We recently had a look at the Intel Xeon Scalable platform which goes scales with multiple CPUs. That’s not the case here and we’re stuck with one socket per system. However, with 18 cores and 36 threads, that’s okay. We’ll have plenty of threads available for extreme multitasking and high-performance calculations.
Intel Xeon W-2195 18-Core 36-Thread Processor
The Xeon W-2155 is a Skylake-series processor built on a 14nm process. It has a base clock of 2.3GHz and a Turbo mode up to 4.3GHz. The base clock is a bit lower than it was on the 10-core, but we get close to the same turbo mode. We also stay within the same 140W TDP which makes this CPU very impressive! That’s a great speed across 36 threads.
Within the 140W TDP package of the W-2195 are 24.75MB cache and a quad-channel memory controller. The processor supports up to 512GB DDR4 ECC memory with a 2666MHz speed. The memory controller has a bandwidth ability of 85.3 GB/s on a 46-bit physical address extension.
That’s all very nice and I’m sure you already love what you’ve learned about the series and this processor. What really can make a difference though, are the 48 available PCI-Express lanes. With that amount of lanes at your disposal, you can install a lot of high-performance features without bottleneck worries. 4-GPU setups, for example, are no problem at all.
Intel C422 Chipset
You no longer have multiple options when it comes to the chipset. The X299 motherboards sadly won’t run the new Xeon processors, so you’ll have to get a C400 series instead. The processors are designed to be workstation processors, and as such, perfect for the C422 chipset to be the base for those. Intel’s C422 chipset is purpose-built for the same thing as the Xeon-W series which makes it a perfect match.
The Intel C422 chipset is built on a 22nm process and comes in a 6W TDP. It supports up to 24 PCI-Express lanes, 8 SATA 6Gbps ports with RAID, and integrated LAN. Up to 14 USB ports can be connected to it and they can be a mix of up to 10 USB 3.0/USB 3.1 Gen.1 and up to 14 USB 2.0 ports. Both Intel Trusted Execution Technology and Intel Boot Guard are supported too along with vPro technology and VT-d.
Advanced Processor Technologies
The basic features such as thread count, speed, and available PCI-Express lanes might already sell this product, but there’s more in the details. Intel baked a lot of their various technologies into this chip, offering an even better performance at situation-intended tasks.
With up to 512GB memory and 36 threads in the series, VM-setups is not out of the question. Even with just 10 cores 20 threads you can set up a lot of VMs. For this, you get basic virtualization with Intel VT-x support and directed I/O virtualization with Intel VT-d. All these cores are able to talk well with each other thanks to TSX-NI. Intel Transactional Synchronization Extension New Instructions help with the multi-threaded scaling. This technology helps make parallel operations more efficient via improved control of locks in software.
If the above sounds like jibberish to you, then this might sound more familiar. The supported instruction set extensions are Intel SSE4.2, Intel AVX, Intel AVX2, and Intel AVX-512 with 2 AVX-512 FMA Units.
Does Security Matter to You?
If security matters to your setup, then you’ll be happy with the next set of features. With AES-NI hardware support, encryption becomes faster and less resource hogging. After all, there should never be a situation where you need to lower your security in order to gain the performance you need. That’s not an acceptable scenario for any enterprise. AES-NI are valuable for a wide range of cryptographic applications. It includes applications that perform bulk encryption/decryption, authentication, random number generation, and authenticated encryption.
The Intel Secure Key feature consists of a digital random number generator that creates truly random numbers to strengthen encryption algorithms. That’s another great layer as software solutions are far more vulnerable. Last but certainly not least, you also get Intel MPX for memory protection.
Xeon W Feature Highlights
- Up to 18 cores and 36 threads
- Four-channel DDR4-2666 ECC memory support
- Intel Turbo Boost 2.0 technology
- Intel ACX-512 acceleration with up to 2 FMA
- LGA2066 socket
- 48 PCI-Express 3.0 lanes
- Intel Mesh Architecture
- Rebalanced Intel smart cache hierarchy
- Intel vPro Technology
- Intel Virtual RAID on Chip (Intel VROC)
- One integrated Gigabit Ethernet
Package and Accessories
Most desktop processors come in the same type of box. The processor is located in a small plastic box inside a larger cardboard box. That’s not the case with the Intel W-2195 that I received. It came in this really cool plastic box. While the package doesn’t matter as such, I like it. When you spend that kind of money on a processor, it’s cool to get a small recognizable collectable.
The processor itself is sitting well protected on the inside, between two pieces of soft foam.
The Setup and Test Methodology
It is not an easy job to come up with fair tests for a processor. After all, you can’t run a CPU without a lot of other hardware, and all that hardware will affect the results. Still, I’ve tried to dig out as many tests as possible which show what the processor is capable of. Hopefully, it will show us where the strengths and weaknesses are located.
Since the last review, we had to make a few changes to the methodology due to changes in technology. Our old graphics card is no longer supported by many motherboards, so that had to be replaced. At the same time, SATA has turned into a true legacy device when it comes to booting. As a result, we’ve switched to an NVMe PCIe drive as OS drive. We’ve also switched the OS to Windows 10 for compatibility reasons. Windows Server was creating more and more issues with driver availability and benchmark software, which is the reason.
- Motherboard: Gigabyte MW51-HP0
- Cooler: Noctua NH-U12DX i4
- RAM: Crucial DDR4 2133 MHz (8x 16GB)
- OS Drive: Intel Optane 900p
- GPU: ASUS ROG Strix RX570
- PSU: be quiet! Dark Power Pro 11 1200W
Suites are popular too as they offer up a whole lot of results with very little effort. We’ll be using AIDA64, GeekBench4, and PerformanceTest 8 here.
- AIDA64 Engineering Edition
- GeekBench 4
- PerformanceTest 8
Having the CPU calculate is an easy way to show how well it does. The quicker the test is performed, the better the performance. For these tests, I’ve picked Super Pi and wPrime. Both are applications familiar to most, and the results are easily comparable.
- SuperPi Mod 1.9 WP
- wPrime 2.10
Rendering tests are mostly handled by your graphics card, but not solely. While these aren’t exactly intended areas of operation, they do offer us comparable figures between our reviews. We’ll be using HandBrake for H.265 conversion of 4K video as well as Cinebench 11.5 and 15 here.
- Cinebench R11.5
- Cinebench R15.0
- HandBrake 4
SiSoft Sandra and SpecWPC
SiSoftware’s Sandra is technically a benchmark suite too, but it offers a lot more details. It tests more and different things, so it deserves its own category. SpecWPC is also a whole set of tests and an industry standard. The results are directly comparable with other reviews which makes it great. Most professionals won’t look at anything else.
- SiSoftware Sandra Platinum SP4
- SpecWPC 1.2
Memory is another thing which is affected by a processors capabilities and performance. For that reason, I’ve chosen to include those test results too.
- AIDA64 Engineering Edition
- GeekBench 4
- PerformanceTest 8
- SiSoftware Sandra Platinum SP2/SP4
It should be noted that some charts don’t make a whole lot of sense right now. The previous generation was tested with different software and there has been a hardware upgrade since we started our LGA3647 reviews. We no longer have the predecessors available, but we will retest the current generation Xeon processors and update them with new and comparable values. So please keep that in mind for the comparison charts where GPU and storage has an effect. Pure calculation benchmarks should not be affected as they rely on the CPUs hardware capabilities alone.
AIDA64 Engineer provides several methods to measure system performance. These benchmarks are synthetic, i.e. the results show only the theoretical (maximum) performance of the system. In contrast to application tests, synthetic benchmarks do not tend to reflect the “real world” performance of the computer. These benchmarks provide a quick and easy comparison between computer states, e.g. when certain parameters (CPU clock speed, memory timings, etc.) change in system configuration.
Geekbench 4 measures your system’s power and tells you whether your computer is ready to roar. How strong is your mobile device or desktop computer? How will it perform when push comes to crunch? These are the questions that Geekbench can answer.
PerformanceTest is a fast, easy to use software benchmarking tool that allows everybody to quickly assess the performance of their PC and compare it to a list of standard ‘baseline’ computer systems. With it, you can compare the performance of your machine, measure the effect of configuration changes and upgrades, and make objective, independent measurements on which to base your purchasing decision.
Super PI is a computer program that calculates pi to a specified number of digits after the decimal point—up to a maximum of 32 million. It uses Gauss–Legendre algorithm and is a Windows port of the program used by Yasumasa Kanada in 1995 to compute pi to 232 digits.
wPrime is a leading multithreaded benchmark for x86 processors that tests your processor performance by calculating square roots with a recursive call of Newton’s method for estimating functions, with f(x)=x2-k, where k is the number we’re square rooting, until Sgn(f(x)/f'(x)) does not equal that of the previous iteration, starting with an estimation of k/2. It then uses an iterative calling of the estimation method a set amount of times to increase the accuracy of the results. Then it confirms that n(k)2=k to ensure the calculation was correct. It is repeated for all numbers from 1 to the requested maximum.
Please note that items marked with an asterisk are measured with 4 threads. Those not marked are tested with as many threads as the CPU has available. We will retest anything that is possible and update with better-comparable numbers.
Cinebench R11.5 & R15.0
Cinebench is a real-world cross platform test suite that evaluates your computer’s performance capabilities. It is a perfect tool to compare CPU and graphics performance across various systems and platforms, and best of all: It’s completely free.
HandBrake is a tool for converting video from nearly any format to a selection of modern, widely supported codecs. There are three reasons to love this tool: It converts video from nearly any format, it is free and open-source, and it is available on Windows, Mac, and Linux.
SiSoftware Sandra Platinum
SiSoftware Sandra is a system analyser that includes benchmarking, testing and listing modules. It tries to go beyond other utilities to show you more of what is going on under the hood, so you draw comparisons at both a high and low-level in a single product.
Multimedia SIMD Units
SPECwpc is the first benchmark to measure all key aspects of workstation performance based on diverse professional applications. The latest version is SPECwpc V1.2 (introduced on June 3, 2015), which extends performance measurement from physical to virtualised workstations. Results from SPECwpc V1.2 are comparable to those from SPEC V1.0.4, but not to any version before that.
More than 30 workloads are included in SPECwpc V1.2 to test CPU, graphics, I/O and memory bandwidth. The tests are divided into the application categories listed below. Individual scores are generated for each test, and a composite score for each category is generated based on a reference machine, yielding a “bigger is better” result. The reference machine has the following configuration: x3430 processor, 8GB (4x2GB) memory, V4800 graphics, and SATA 7.2k rpm hard disk drive.
Media and Entertainment
Memory Bandwidth and Latency
The memory tests we use are part of the various benchmark apps seen on the previous pages. This includes AIDA64 Engineer, Geekbench 4, PerformanceTest, and SiSoftware Sandra Platinum. A wider array of tests will give us a wider array of results, and while the memory isn’t the CPU, the performance depends on it.
AIDA64 Engineer provides several methods to measure system performance. These benchmarks are synthetic, i.e. the results show only the theoretical (maximum) performance of the system.
Geekbench 4 measures your system’s power and tells you whether your computer is ready to roar.
PerformanceTest is a fast, easy to use software benchmarking tool that allows everybody to quickly assess the performance of their PC and compare it to a list of standard ‘baseline’ computer systems.
SiSoftware Sandra Platinum
Sandra is a system analyser that includes benchmarking, testing and listing modules. It tries to go beyond other utilities to show you more of what is going on under the hood
Final Thoughts and Pricing
The Xeon W series is quite new and as such it might not be easy to find. At the time of this review, I could not find any listing at our usual shopping places. However, the official MSRP for the Xeon W-2195 is $2553.00.
Today we had a look at the flagship Xeon W-series processor, and boy, what a beast. With 18 cores packed in an LGA2066 form factor, it isn’t hard to see why this is the flagship. It is an unprecedented amount of cores and threads for a CPU of this type. Not only that, it delivers an incredible performance. It doesn’t really matter what you’re going to use this type of processor for, whether you’re dealing with graphics, financial calculation, or scientific research. It shines in all areas. Well, it isn’t a gaming CPU, but I doubt there is anyone out there who would purchase it for that anyway.
Is the CPU worth the price?
The asking price certainly isn’t cheap, but is it worth it? Yes, in my opinion, it is! And also sure that a lot of people will think about a comparison to Threadripper now – but there really isn’t any. Yes, they’re powerful processors, but they’re far from what Intel delivers here. At least that’s my take on it and I have to admit, I put out that statement without having tested them in the same scenarios. Sadly, I don’t expect to be able to either, but that’s how it goes sometimes.
The pure numbers truly speak for themselves. Not only does the Xeon W-2195 deliver, it’s an impressive step up from the previous generations. The official statements of 20% IPC gain might, in fact, be an understatement in many situations. The processor chews through the numbers as hot water does through ice.
Intel packed a lot of advanced technologies into this processor, we already know that from our review of the 10-core W-2155 processor. The W-2195 has the same features, but more cores, more threads, and more cache. Depending on the task and whether your application setup can take full advantage of all the threads is something else. It isn’t every situation that will benefit from the extra cores, but it will never hurt. Especially when it comes to encryption, we see a very strong performance and that is no doubt due to the integrated hardware functions.
- 36 Threads, 18 Cores
- Quad-Channel DDR4 ECC memory up to 2666MHz
- AVX-512 with up to 2 FMA units and Intel Mesh Architecture
- Great compatibility through the LGA2066 socket
- 48 PCIe Lanes
- vPro and VROC Technologies
- Availability at the time of review
- The price might be an issue for some interested parties
“Intel’s Xeon W-2195 is a true monster and a worthy flagship product. Those who need the performance and features within will love it!”
Thank You, Intel, for this review sample.