What does the new line of Intel processor mean for desktop buyers?



[ad_1]

Intel introduced this week a set of new processors, including its 9th generation Intel Core desktop processors, a new Core X series for high quality content creation, a 28-core Xeon processor for workstations, and A new chipset.

Core i9 wafer

It is a very broad and powerful range reflecting Intel's objectives of increasing average selling prices by segmenting the market and counterbalancing AMD's Ryzen desktop processors. , including the 32-core Threadripper. But with the increased segmentation, the complication increases and it will be difficult for most people to determine which processor makes the most sense for their applications.

Anand Srivatsa Intel on October 18th

Consider the announcements today. Anand Srivatsa, vice president of Intel's Client Computing Group, said Intel's goal was to achieve leadership performance across all segments and uncompromising product lines. To this end, the company has launched a portfolio of office products for content players and creators, consisting of three different lines with different architectures, memory structures and even various manufacturing processes (14 nm designs). and 14nm, 10nm processors planned for a year or two).

Intel is focused on what is said of 130 million "creators". Srivatsa said that 44% of professionals refresh their systems every 2 years, while 34% of potential users and major creators update their computers every 2 or 3 years, much faster than most computer users. office.

Xeon with 28 hearts

Architecture of Xeon W-1375

The most powerful processor is the Xeon W-3175W workstation processor. It's a monster chip with 28 cores and 56 wires, designed to get a simple 4.3 GHz turbo frequency, with possible overclocking (which is unusual for a Xeon chip). It offers 125GB / second of memory bandwidth, with support for 6-channel DDR4 memory and up to 512GB of memory at 2666MHz; ECC memory support; 38.5 MB of cache; and support for AVX-512 instructions. This is the latest addition to the Skylake-SP series, built on the 14-nm Intel process, and expected to be delivered in December. Intel has not yet provided its prices.

Intel has positioned this processor as suitable for applications such as movie creation. Tangent Studio showed how it uses Intel Embree ray casting nuclei to create its next Netflix movie, NextGen. A studio representative explained how this allowed developers to quickly make lighting and creative calls, but he added that rendering each image of a movie can take up to 4 hours. Four different versions must be created for each image. the total render time is still huge.

Xeon motherboard

Intel has not officially introduced any new motherboards, but several new Xeon cards have been on display.

Core X-series

Core X-Series Architecture

The Core X series, designed for high-end content creators, includes Core i7, Core i9, and Core i9 Extreme family models, with a total of seven i7-9800X series models with 8 cores and 16 wires at i9 -9980XE with 18 cores and 35 wires. This series offers Intel Turbo Boost Max technology, which allows an individual core to operate at 4.5 GHz (the base speed ranges from 3.8 GHz on the Core i7-9800X to 3.0 GHz for the i9-9980XE ). These processors support 4 channels of DDR4-2666 memory, and the top of the line includes 24.75 MB of cache.

These processors are known as Basin Falls, an update of the Skylake-X platform from last year, with similar counts and base prices, but usually at a higher frequency. (For example, the i9-9980XE has a base frequency of 3.0 and a 4.5 GHz turbo, while the current i9-7980XE has a base frequency of 2.5 GHz and a turbo of 4.4 GHz). They operate on the existing X299 chipset, with a TDP of 165 Watts. All models support a total of 68 PCIe platform channels, including 44 on the processor and 24 on the platform control hub (PCH). These run on the 14+ process, which seems to have relaxed some of the design rules (which means it's not as dense as the early 14 nm process) in order to allow higher speeds.

A new feature in this regard is the Thermal Welding Interface (TIM) material, which lies between the processor chip and the heat sink, and offers increased thermal conductivity; in other words, it distributes heat further and more evenly, which helps to improve overclocking.

demo photogrammetry

The demonstrations here included photogrammetry, using a snapshot by the Smithsonian American Art Museum of many photos on a 10-core version (because the application will not evolve much further).

Unreal Demo

Another demo showed that the Unreal game engine was running on a 18-core 9980XE processor that used both multithreaded rendering and higher-than-thread performance to show what the game would look like.

Prices range from $ 589 to $ 1,979 (amount of 1000). These will be shipped in November.

9th generation base

9th generation architecture

Finally, Intel announced its 9th generation Core processor line, with its Core i5, Core i7 and Core i9 variants, all designed for overclocking. The Core i5-9600K has six cores and six threads, with a maximum frequency of 4.6 GHz, 9 MB cache and a list price of $ 262 (1000 quantity). The Core i7-9700K has eight cores and eight threads, with a maximum storage capacity of 4.9 GHz, 12 MB of cache, and a list price of $ 374. The high-end Core i9-9900K offers eight cores and 16 threads, with a maximum frequency of 5.0 GHz and 16 MB cache, for a list price of $ 488. Srivatsa claimed that the i9-9900K was the first 5.0 GHz high-volume chip, which he said was "violating the laws of physics". (It looks fast, but no, it 's not that fast.

Not only does the high-end have more hearts than last year (eight against six in the i7-8700K), but the frequencies are higher, it has more cache (2 MB per heart) and it takes in loads the AVX-512 instructions. The processors are designed for overclocking and include TIM Solder.

The demonstrations presented here consisted of playing and broadcasting two games simultaneously on two virtual machines on a processor, and running games at faster speeds. Srivatsa said that all major game designers would support this lineup, including Acer Predator, Asus Republic of Gamers, Lenovo Legend, HP Omen and Dell Alienware.

These chips are known as Coffee Lake-S and are also manufactured using the 14+ process. They use a new chipset called Z390, but can also work with existing Z300 cards. They are available for pre-orders today and should be available on October 19th.

Performance

What's interesting for me is the performance numbers. For chips of the 9th generation, Intel generally claims that the game is better performing 10 to 11% (images per second) or multitasking compared to last year's chips (8700K). Three year old PC (running 6700K). But for some things, the performance is much better: Intel claimed a faster video editing of 34% on Adobe Premiere Pro, compared to the chip of last year, and 97% faster compared to a system of three years.

Mesh architecture

For the X series, Intel engineers talked a lot about the mesh architecture of the chip, which connects all cores to memory and IO controllers using rows and columns; It is said that the unified memory architecture results in lower and more predictable latency. For comparison, Intel compared its 18-core Core i9-9980XE to its 32-core Threadripper Ryzen 2990WX. (As always, I take the supplier reference numbers with a grain of salt, and you should too.)

That said, Maya's performance has been improved by 27%, which offers many low-threaded features (suggesting that an 8-core version would actually be faster than the 18-core version). up to 108% better performance in Premiere Pro because they are actually memory-sensitive; and performance up to 13% faster in Unreal because they use all cores. They pointed out that there was not a number that covered everything.

Conclusion

And it is there that everything is complicated for me.

For those of us who buy machines for businesses and businesses, it was pretty clear a few years ago: Core i7 for the high end, Core i5 in the middle and Core i3 for beginners, with a few Workstation parts designed for those who run very specialized applications. (Before that, you could choose according to the clock speed). But now that processors are moving faster, the task has become more difficult.

Today, just about every consumer application – Office, Acrobat, the Web browser and even basic videoconferencing – works perfectly across the entire Core lineup (and the AMD Ryzen lineup, for that matter). But for high-end applications, you need to know if the application actually requires more cores, more bandwidth, higher frequency cores, and so on. And I guess most IT departments simply do not know it.

But it's even more complicated than that. Some functions in some applications use a lot of hearts; others use only one or two threads. (An example: Excel is multi-threaded and evolves well, but Visual Basic for Applications does not.) To emphasize: It has become much more difficult to choose the right processor for your application. This may be suitable for an individual: members of a community can determine which is the best processor for their particular application. But for a business buyer supporting a number of different applications, it is almost certain that you will end up with a product that represents a compromise. And that's the price we pay for such a segmentation.

[ad_2]
Source link