CPUs


Now Shipping: AMD Ryzen Threadripper 1900X & Ryzen Pro CPUs

Now Shipping: AMD Ryzen Threadripper 1900X & Ryzen Pro CPUs

Back when AMD announced their leading-edge Ryzen Threadripper 1950X and 1920X, the company also tucked in an announcement that there would be a third, cheaper Threadripper: the 1900X. Now after about a month’s wait since the first two chips launched, the 1900X is finally hitting the streets.

AMD High-End Ryzen SKUs
  Cores/
Threads
Base/
Turbo
XFR L3 DRAM
1DPC
PCIe TDP SRP
TR 1950X 16/32 3.4/4.0 +200 32 MB 4×2666 60 180W $999
TR 1920X 12/24 3.5/4.0 +200 32 MB 4×2666 60 180W $799
TR 1900X 8/16 3.8/4.0 +200 16 MB 4×2666 60 180W $549
Ryzen 7 1800X 8/16 3.6/4.0 +100 16 MB 2×2666 16 95 W $499
Ryzen 7 1700X 8/16 3.4/3.8 +100 16 MB 2×2666 16 95 W $399

Of AMD’s Threadripper product stack, the 1900X is easily the most unusual of the chips. Whereas the 1950X and 1920X were the full-fledged version and the cheaper-but-still-thread-heavy-alterative respectively, the 1900X doesn’t really fit into either of those buckets. In fact with just 8 cores active, it has more in common with the Ryzen 7 1800X than it does the rest of the Threadripper family.

So why is AMD releasing an 8 core version of Threadripper when one Zeppelin die can accomplish the same thing at $100 less and almost half the TDP? Because while the 1900X may be a few cores short of a full Threadripper, it retains the platform’s immense I/O and memory bandwidth capabilities. In other words, 4 DDR4 memory channels and 60 + 4 PCIe lanes, offering twice the cumulative memory bandwidth and three times as many freely available PCIe lanes as AMD’s standard Ryzen processors. In fact other than cores, the only other thing it loses out on from the 1950X is L3 cache, with 16MB rather than the 32MB of its higher core counterparts. Conversely, its 3.8GHz base clockspeed makes it the highest base clockspeed of all of AMD’s Ryzen CPUs, standard and Threadripper.

It does bear mentioning that Threadripper’s multi-die layout does comes with the same potential caveats as the other SKUs, which is why the 1900X isn’t going to be equal to or better than the 1800X in all scenarios. However with a MSRP of $549, the 1900X is now the cheapest way to get a high-I/O x86 processor. So for tasks that require a more limited number of threads and a lot of I/O and memory bandwidth, the 1900X should fit comfortably in its niche.

As for reviews of the new processor, unfortunately AMD did not sample the 1900X in advance. We’re angling to get one in soon, so please stay tuned for that.

Meanwhile also shipping out the door from AMD today is their Ryzen PRO processors. Announced back at the end of July, these are enterprise IT -focused versions of AMD’s standard Ryzen processors, offering longer support periods along with features such as enhanced security and added manageability features. Accordingly, AMD’s partners have started showing off their latest enterprise-systems, with Dell, HP, and Lenovo all set to begin shipping Ryzen PRO-equipped systems in the following weeks.

AMD Ryzen PRO Specifications
  Cores/Threads Frequency Cache TDP
Base Boost L2 L3
Ryzen 7 PRO 1700X 8/16 3.5 GHz 3.7 GHz 4 MB 16 MB 95 W
Ryzen 7 PRO 1700 3 GHz 3.7 GHz 65 W
Ryzen 5 PRO 1600 6/12 3.2 GHz 3.6 GHz 3 MB
Ryzen 5 PRO 1500 4/8 3.5 GHz 3.7 GHz 2 MB
Ryzen 3 PRO 1300 4/4 3.5 GHz 3.7 GHz 8 MB
Ryzen 3 PRO 1200 4/4 3.1 GHz 3.4 GHz
Best CPUs: August 2017

Best CPUs: August 2017

In our series of Best Guides, here’s the latest update to our recommended CPUs list. All numbers in the text are updated to reflect pricing at the time of writing (29-Aug). Numbers in graphs reflect MSRP.

Best CPUs, August 2017

Sometimes choosing a CPU is hard. So we’ve got you covered. In our CPU Guides, we give you our pick of some of the best processors available, supplying data from our reviews.

CPU Recommendations: August 2017
AMD Segment Intel
Ryzen 7 1800X
Ryzen 7 1700
$431
$300
Peak Gaming / VR $325
$337
Core i7-7740X
Core i7-7700K
Ryzen 5 1600 $210 Smart Gaming / VR $198 Core i5-7500
Ryzen 3 1300X $130 Sub $700 Gaming $116 Core i3-7100
Athlon X4 960
Athlon X4 860K
?
$66
Sub $500 Gaming $85 Pentium G4560
A12-9800 + AM4
A10-7890K + FM2+
A10-7860K + FM2+
$111
$143
$96
Integrated / eSports
Threadripper 1950X $1000 CPU Workstation $1000 Core i9-7900X
Threadripper 1900X
EPYC 7251
$549
$475
Memory Workstation $213 Xeon W-2123
Raven Ridge APUs   One to Watch   Core i9-7980XE
Core i7-8700K
Core i3-8100

The majority of our recommendations aim to hit the performance/price curve just right, with a side nod to power consumption as well.

When FPS is Required: CPUs for Peak Gaming Performance or Peak VR
The AMD Ryzen 7 1800X/1700 or Intel Core i7-7740X/i7-7700K

If ‘peak’ is to mean ‘money no object’, then we typically suggest some of the most expensive processors, regardless of platform cost. Under this heading, the Ryzen 7 1800X at $499 (MSRP) represents the best AMD has to offer on a mainstream platform, with the highest frequency in the AMD Ryzen 7 range, eight cores, and consistent memory performance. There isn’t much headroom for overclocking beyond 4.0 GHz, but as a single or dual GPU solution AMD has you covered. The AM4 motherboard platform is also now mature, with the latest AGESA updates showing some good gains in performance. AMD has stated that they are committed to future AGESA updates to boost performance as well.

Users who still want a peak experience and feel comfortable with overclocking can instead save $170 from the list price by going after the Ryzen 7 1700, which is the same as the 1800X except for a slightly lower clock frequency. This processor can be overclocked to 1800X or near 1800X levels in most environments, allowing users to get another $170 on a graphics card, CPU cooler, memory or storage.

For users going down the Intel route, the new Kaby Lake-X based Core i7-7740X represents the top step for performance out of the box. The only way to get 4.5 GHz for any other processor is through overclocking. Similar to the AMD parts, this CPU supports up to two way SLI and up to 64GB of DRAM, but the downside here is that the X299 motherboards that this CPU requires start at almost double the average price, although they can be upgraded to larger Skylake-X CPUs for users with deeper pockets over the next few years.

The peak experience on Intel can also be had by the Core i7-7700K, a Kaby Lake-S processor that can hold a frequency close to 5.0 GHz with some overclocking, but can also be placed into a Z270 motherboard, reducing the system cost to get the same experience and putting more money aside for other components. As for a pure out-of-the-box gaming experience without overclocking, the Core i7-7740X still wins however.

For users that want to move beyond two GPUs in a truly money-no-object gaming and streaming and video editing environment, then it becomes a choice between AMD’s Ryzen Threadripper processors and Intel’s Skylake-X components. More on those later.

Virtual Reality on a Reasonable Budget: The Smart Gaming Choices
The Intel Core i5-7500 or AMD Ryzen 5 1600

For processors that still need to perform but do not blow a hole in the wallet, both AMD and Intel cater for this market actively. On AMD’s side, we’ve chosen the Ryzen 5 1600. This processor uses the same underlying technology and base as the higher-class Ryzen 7 processors, but removes some peak performance in workloads such as video conversion in order to be offered at a lower price point. It comes with six cores and twelve threads with a good out-of-the-box performance that can be pushed higher with a nice overclock. To save a few pennies, the retail version also comes bundled with a CPU cooler worth using.

(1080p) GTX 1060: Grand Theft Auto V, Average Frame Rate

Intel’s side of the fence is a bit fuzzier. The Core i5-7500 is what we’ve chosen here: a quad core processor without hyperthreading that will give a near-identical gaming experience to the Core i7 parts, especially in older games that rely on single thread performance. If that is all you care about, the Core i5-7500 (or i5-7400 to save a few more $) with a 200-series ecosystem and a beefy GPU will not disappoint. For a user that wants to process video, or do some heavy work, it might be worth cutting a couple of percent on that gaming performance and going for the AMD parts that offer twelve threads compared to Intel’s four. For heavy work with gaming, it’s hard to miss AMD in this price range when comparing the two.

My Wallet, She Grips So Tight: Gaming on a $700 Budget
The AMD Ryzen 3 1300X or Intel Core i3-7100

Our suggestions for processors in $700 PCs get an update here on the AMD side, as since the last guide we have had AMD’s Ryzen 3 processors come to the market. These are quad core parts in the $110-$130 range, with the same features and nearly the same high frequencies as their Ryzen 5 and Ryzen 7 siblings, but aimed at this lower price. Out of the two Ryzen 3 processors, the 1300X gets the gold star by being a higher frequency out of the box, a larger XFR boost, and having plenty of room to overclock as well on the cooler provided. 

(1080p) GTX 1060: Ashes Escalation, Average Frame Rate

The competition from Intel sits at the door with the Core i3-7100. The trade offs here are a higher single thread frequency, but only two cores with four threads rather than a full four cores. For some of the newer games, or for more multi-tasking activities, that could be of concern to users but for a cheap gaming system, the Core i3-7100 should have no issues with eSports-like titles or driving mid-range graphics cards. Some users might prefer going after something like the Pentium G4560, which is only a few MHz lower on frequency but has half the list price, which is why it sits in our next recommendation.

Shoestrings Anonymous: Cheaper Gaming on $500
The Intel Pentium G4560 or AMD Athlon X4 860K or AMD Athlon X4 960

The Pentium G4560 is heralded as the diamond in the rough for sub $100 processors. It comes out of the gate with a sizeable 3.5 GHz frequency, two cores, four threads, and a list price of $64. In previous generations, a chip like this would be almost double the price if not more, which makes it an interesting purchase for users who rely on cheaper components for something like a $500 gaming system. Technically this CPU is an OEM processor, so finding stock might be tricky, or the fact that it usually sits at $80 or so at retail, but even at that price it still becomes the processor of choice.

(1080p) RX 480: Rocket League, Average Frame Rate

For now, AMD’s Ryzen processor family and newest silicon does not drop down to this price bracket. However since our last CPU guide, AMD has officially released their Bristol Ridge processors to market. These were available in OEM pre-built PCs for the last year or so, but now come to market as the best of AMD’s previous generation of products. For users looking at investing in a discrete graphics solution to pair with an AMD Bristol Ridge CPU, the Athlon X4 960 sits at the top of the CPU options here, marking an upgrade over our last Q2 recommendation by moving to AM4 motherboards and support for DDR4, as well as performance improvements. 

AMD stated these CPUs should be on shelves by August 18th, and we’re waiting to get one in ourselves for testing. Ultimately the X4 860K is easier to purchase, but relies on the older FM2+ based motherboards, which is why it still appears in our recommendations. 

Play Now, Play Later: Gaming on Integrated Graphics
The AMD A12-9800, AMD A10-7890K or AMD A10-7860K

Building a smaller or cheaper gaming system, either as a full-time PC or as a precursor to buying a large discrete GPU, AMD’s new Bristol Ridge processors are the best options. These are updates over the previous generation Kaveri processors, and use the latest AM4 platform for PCIe 3.0 support as well as DDR4 and the latest audio/networking controllers. The A12-9800 sits at the top of this processor list, and should be around $120 depending on location. 

Grand Theft Auto V on Integrated Graphics

Users who want to cut a few more dollars off the build, or depending on the availability of the new A12 processors, can still look at the Kaveri A10 parts. The A10-7890K is the best processor in that range, although the A10-7860K is practically as-good but at a lower price point. These parts still use DDR3 memory which might require diving into the pre-owned bin, but paired with some fast memory and they can still tackle a good number of modern games at low-to-medium settings very well.

Not many users take the line of integrated graphics because GPUs like the RX460 can be relatively cheap (when people aren’t mining) and usually outperform what is on offer. Nonetheless, almost all mainstream CPUs come with integrated graphics for several reasons: invest now in a good CPU then later in a GPU, or the amount of space limited in the system is small, or features such as QuickSync or OpenCL are needed for the user workloads.

Juiced to the Max: Megatasking Processors for CPU Workstation Horsepower
The AMD Ryzen Threadripper 1950X or Intel Core i9-7900X

For users who want it all, the best of the best, or actually need lots and lots of compute power, then both AMD and Intel want to sell you some clever sand. Since our last guide, AMD launched its new Ryzen Threadripper processors and X399 motherboard platform, offering up to sixteen of their new Zen cores, up to 32 threads, sixty PCIe lanes to play with, and a high turbo frequency. The Threadripper 1950X is the top processor here, retailing at $999, and in our testing stretched out a sizeable lead on almost all of our hardcore throughput tests. The Threadripper CPUs are based off of AMD’s server products, and the way they are manufactured does mean that a few compute edge cases need additional optimization, but AMD has attacked this market hard with a product very few people were expecting. The 1920X is also available, with 12 cores and hyperthreading for $799, also with 60 PCIe lanes.

The recent set of high-end CPU releases means there is a lot of choice, as well as variation in performance and price. Intel’s current top offering is a 10-core Core i9-7900X, which pushes through Intel’s traditional conservative frequency and power ratings to give a very fast multi-core processor that also bundles in compute features such as AVX-512. The Core i9-7900X is the current top component of Intel’s Skylake-X CPU line, with a full set of 44 PCIe 3.0 lanes and takes advantage of the new X299 chipset for motherboards to take advantage of more controllers. Intel also offers two other CPUs in this range, with reduced core counts and reduced PCIe lane offerings. 

Rendering: Blender 2.78

Both platforms have tradeoffs here. AMD has the raw core count and more PCIe lanes, at the expense of a few watts. Intel instead has AVX-512 and higher single thread performance, but has fewer PCIe lanes (which decrease with cheaper CPUs). Both systems use quad-channel DDR4 memory, and both have different approaches to how each of the cores talk to each other: Intel’s new network Mesh allows for scalability within one large piece of silicon, while AMD’s multi-silicon strategy uses their new Infinity Fabric for scale. We’ve seen both methods have upsides and downsides, along with improvements and regressions. 

Doubling Down on DRAM Density: CPUs for Memory Workstations
AMD Ryzen Threadripper 1900X (Aug 31st) or Intel Xeon W-2123
or AMD EPYC 7251 (if you can find it)

Some workloads just need lots of memory. The more memory in a system, the better – even with DRAM prices currently skyrocketing, some compute tasks just need memory. Both of AMD and Intel’s best consumer platforms can accept up to 128GB of DRAM, suggesting that it comes down to the cheapest processor on each: Intel’s Core i7-7800K is around $350, while AMD’s Threadripper 1900X will be at $549 when it comes out at the end of the month.

For users looking to get serious about memory capacity, the tables turn the other way when we look at the enterprise platforms. Both AMD and Intel have launched new server processors in the last few months, so pricing and availability become strong elements into our recommendations here.

The winner in what we recommend falls at AMD’s EPYC ecosystem, which can run up to 2TB per socket when using 128GB LRDIMMs. Those memory modules run at $4000+ each if you can find them, ensuring the $64000+ cost of the DRAM is more of a consideration than what the processor cost is. The availability of EPYC platforms from the OEMs is still a work in progress however: unless you want to deploy a substantial number of servers, it’s going to take some time (several months) before EPYC is fully deployed to customers other than the big cloud providers. Even the low-end CPUs that support 2 TB per socket, priced around $450, are not ‘freely’ available with AMD and OEMs focusing on the higher margin parts first. 

By comparison, Intel’s partners are shipping new Xeon platforms today. These come in two variations: the Xeon-SP platform based on Xeon Bronze, Xeon Silver, Xeon Gold and Xeon Platinum, and also the Xeon-W platform that reuses the LGA2066 socket from the consumer space. The Xeon-W 2123 processors, with their cheaper motherboards, can support up to 512GB of DDR4 memory, and are destined to find their way into retail packaging. The Xeon-SP processors start support at 768GB of DRAM, although to get special 1.5 TB support versions, these will cost ~$3000 (list price) above the regular versions and come via OEMs only. These double memory parts start in the Xeon Gold range, and the base versions vary around $3000 anyway.

The 2017 August CPU Market Roundup

So far the first half of this year has been mildly insane. Back in 2016, we had the best part of 1.5-2 platform launches and it was a quiet year on the CPU side. So far this year we have had three or four launches, with another few in the pipeline to come. With Kaby Lake, Ryzen 7 and Ryzen 5 out of the way, in comes Kaby Lake-X and Skylake-X to the party, followed by Skylake-SP, Threadripper, EPYC, and potentially more still on the invite list later this year such as Intel’s 8th Generation.

For consumers, it can be a fun time. With new platforms comes an opportunity to upgrade, either through the increase in performance or just because you want the latest and greatest. The idea is that the newest processors are more performant, or lower power, or fit into a particular niche better (and hopefully are the same or lower cost overall). For users who wanted to invest in AMD this year, Ryzen has been a good offering and ThreadRipper is aiming at the multi-core market. For users looking to upgrade that i7-2600K or i5-2500K, Intel is trying hard to tempt them with Kaby Lake processors and even Kaby Lake-X, with the best part of 25-35% IPC and some extra MHz as well. For anyone that wanted a 10-core CPU and thought $1721 was too much for the i7-6950X, Intel has you covered with the Core i9-7900X at $999-$1049 now as well.

Our big CPU reviews for 2017 have covered all the launches so far, and well worth a read,

Intel 2017 Reviews

AMD 2017 Reviews

In the pipe coming in the next few weeks include our AGESA retest on Ryzen, Skylake-X Gaming analysis, hopefully some reviews of the 1900X/7920X that should be going to market this week. and we’ve put in a request for some EPYC and Xeon-W processors as well.

One of the overriding issues so far this year worth mentioning is platform maturity. With new platforms come new challenges, and as far as we understand, extreme deadlines. Motherboard manufacturers, for both AMD and Intel, have had to rush through some of the production of their initial motherboards at launch. When we reviewed Ryzen 7 and Kaby Lake-X, both of those reviews did not have gaming results due to erroneous results on young hardware. At this point we expect both platforms to be running smoothly, but as an indication that this year is about time to market, it’s a big one to note for early adopters (and reviewers that end up wanting to throw products out a window).

Prepare for Processors: Ones to Watch
AMD Raven Ridge, Intel Core i9-7980XE, Intel Core i7-8700K, Intel Core i3-8100

Despite this year being one of the busiest on record for CPU launches, the finishing line is almost in sight. AMD’s focus between now and the end of year, aside from the Threadripper 1900X coming out at the end of August, is going to be on their Raven Ridge laptop components that pair four Zen cores with a Vega GPU. We will see desktop versions of these processors, but not until the end of the year at the earliest.

Intel still has some bandwidth coming up to make life on the desktop interesting, however. The rest of the Skylake-X processors, from 12 to 18 cores, are set to launch in the next couple of months. Intel has stated that the 12-core 7920X should be at the end of August, with the 14 and 16-core parts in October. The full 18-core is likely to hit our scene around the same time as well. These parts will compete against AMD’s 16-core Threadripper 1950X, although AMD will have a price advantage: Intel’s 16-core already has an announced $1700 price tag, compared to $999 for AMD. When we get the parts for review, it will be an interesting showdown.

The other coin in Intel’s launch schedule is going to be the new 8th Generation coming to Desktop. We’ve already reported that the Intel launched the 8th Gen with a quartet of refreshed laptop processors, but these were refreshes of previous generation parts with double the cores. Users are expecting the desktop components to be new designs (‘Coffee Lake’) and as we reported on in the last couple of weeks, Intel is telling partners that the top Core i7-8700K processor will have six cores, meaning that after a decade of Intel’s mainstream platform offering quad cores, it marks a long-awaited step up to hex-cores. Out of the data we obtained and verified last week, the interesting component for me was the Core i3-8100. This will be a quad-core processor, similar to the current Core i5 parts, but should come in around $120. This level of base performance, on paper, should be a nice entry level processor compared to the dual cores that are slowly limiting what a non-gaming PC is now required to do. 

Perhaps the downside to this 8th Gen launch will be the socket and motherboard situation: these new processors are expected to be in the LGA1151 form factor, the same as the current generation, but will require new 300-series motherboards that will not be backward compatible with older processors. That in itself is not a new thing – we are used to having Intel refresh platform capability every two generations, however we typically get a new LGA form factor as well as a jump in capability/reduction in process node. This will be Intel’s third desktop processor on 14nm, breaking that cycle, and no official word has been given to whether the new chipset has new features, leaving some of us to wonder if backward compatibility isn’t possible due to product segmentation. The big fuss around a new process node, as well as a new socket form factor or change in the form factor naming helps ease the transition from one platform to another when they are not backward compatible. Intel should understand this from previous experience. If it is still named LGA1151, then it only adds to the confusion and becomes one less way to differentiate between different product families.

Links to Other Guides

 

 

Hot Chips: IBM’s Next Generation z14 CPU Mainframe Live Blog (5pm PT, 12am UTC)

IBM’s mainframe business is a large element of their business. The z series is part of that business, and a very complex design, citing interesting and different processor design techniques. Here at Hot Chips, IBM is taking us through their latest z14 design, using a 14nm process node, optimized pipelines, pervasive encryption and other features.

Hot Chips: Intel Knights Mill Live Blog (4:45pm PT, 11:45pm UTC)

Another talk from Hot Chips, this time on Intel’s Knights Mill (KNM). The Intel Knights family stems from their Xeon Phi product line, although KNM is a bit different, with machine learning specific changes. It’s not a completely new Xeon Phi design, but Intel wants to go after the machine learning market. Today’s talk will go into some of those changes. (We’re battling some wifi here, so pictures may come later).

Intel Launches 8th Generation Core CPUs, Starting with Kaby Lake Refresh for 15W Mobile

Intel Launches 8th Generation Core CPUs, Starting with Kaby Lake Refresh for 15W Mobile

This year has been enjoyably eventful for processor releases. Intel launched their 7th Generation processors, Kaby Lake, in January. Then we had AMD release their new high-performance microarchitecture in Ryzen, EPYC and Threadripper. Intel then launched their Skylake-SP Xeon Scalable Platform, based on an upgraded 6th Generation core design, and we’re expecting new AMD APUs for mobile later this year.

And adding to that list this morning is once again is Intel. Today the company is launching its new 8th Generation family of processors, starting with four CPUs for the 15W mobile family. The launch of these processors was perhaps spoiled by Intel jumping the gun a few days ago and listing the processors on its own public price list, but also we have started to see laptop and mobile designs being listed at various retailers before the official announcement.

There are two elements that make the launch of these 8th Gen processors different. First is that the 8th Gen is at a high enough level, running basically the same microarchitecture as the 7th Gen – more on this below. But the key element is that, at the same price and power where a user would get a dual core i5-U or i7-U in their laptop, Intel will now be bumping those product lines up to quad-cores with hyperthreading. This gives a 100% gain in cores and 100% gain in threads.

Obviously nothing is for free, so despite Intel stating that they’ve made minor tweaks to the microarchitecture and manufacturing to get better performing silicon, the base frequencies are down slightly. Turbo modes are still high, ensuring a similar user experience in most computing tasks. Memory support is similar – DDR4 and LPDDR3 are supported, but not LPDDR4 – although DDR4 moves up to DDR4-2400 from DDR4-2133.

Specifications of Intel Core i5/i7 U-series CPUs
7th Generation 8th Generation
  Cores Freq +
Turbo
L3 Price   Cores Freq +
Turbo
L3 Price
i7-7660U 2/4 2.5/4.0 GHz 4 MB $415 i7-8650U 4/8 1.9/4.2 GHz 8 MB $409
i7-7560U 2.4/3.8 GHz $415 i7-8550U 1.8/4.0 GHz $409
i5-7360U 2/4 2.3/3.6 GHz 3 MB $304 i5-8350U 4/8 1.7/3.6 GHz 6 MB $297
i5-7260U 2.2/3.4 GHz $304 i5-8250U 1.6/3.4 GHz $297

Another change from 7th Gen to 8th Gen will be in the graphics. Intel is upgrading the nomenclature of the integrated graphics from HD 620 to UHD 620, indicating that the silicon is suited for 4K playback and processing. During our pre-briefing it was categorically stated several times that there was no change between the two, however we have since confirmed that the new chips will come with HDCP 2.2 support as standard for DP1.2a, removing the need for an external LSPCON for this feature. Other than this display controller change however, it appears that these new UHD iGPUs are architecturally the same as their HD predecessors.

Fundamentally these are what Intel calls a ‘4+2’ silicon design, featuring four cores and GT2 integrated graphics, whereas the last generation used 2+2 designs. The 4+2 design was also used in the mainstream desktop processors, suggesting that Intel is using those dies now for their 15W products rather than their 45W+ products. That being said, Intel is likely to have created new masks and revisions for this silicon to account for the lower power window as well as implementing HDCP 2.2 support and other minor fixes.

Now by having quad-core parts in the 15W form factor, performance on the new chips is expected to excel beyond what has been available from the previous generation of Core i5-U and Core i7-U processors. However Intel and its OEMs have a tight balancing act to walk here, as 15W is not a lot of thermal headroom for a two core CPU, let alone a four core one. At the same time we have started to see the 15W U-series parts find their way into smaller and even fanless notebook designs, which are more prone to throttling under sustained workloads, and quad core CPUs in this segment could exacerbate the issue. However, for the larger 13-15-inch designs with active cooling, moving down from a 35W-45W quad core processor down to 15W will likely offer substantially better battery life during intense loading, should OEMs swap out H-series chips for the new U-series chips in their designs.

Intel’s big aim with the new processors is, as always, to tackle the growing market of 3-5+ year old devices still being used today, quoting better performance, a better user experience, longer battery life, and fundamentally new experiences when using newer hardware. Two years ago Intel quoted 300 million units fit into this 3-5+ year window; now that number is 450 million.

Intel provided this shot of a wafer containing these new refresh dies, which by my math gives 22 x 32.7 dies per wafer. Giving some margin for die spacing, this correlates to a 13.6 x 9.1 mm die, at 124 mm2 and 478 full dies per wafer. At a tray cost of $409 per Core i7, and running at ~124mm2 per die, that makes an interesting metric of $3.30 per square millimeter. Intel no longer officially provides die sizes or transistor counts, though a list of $/mm2 would be interesting to compile – for reference some of the high-end Xeons push north of $19/mm2.

Kaby Lake Refresh? 14+? Where’s my Coffee (Lake)?

So despite Intel launching its 7th Generation family in January, today Intel is formally launching the 8th Generation only eight months later. To explain why Intel is breaking the usual 12-18 month cadence for the generation product, it comes down to product positioning.

In the past we are used to a new numbered generation to come with a new core microarchitecture design. But this time Intel is improving a core design, calling it a refresh, and only releasing a few processors for the mobile family. We expect that Intel’s 8th Generation will eventually contain three core designs of product on three different process design nodes: the launch today is Kaby Lake Refresh on 14+, and in the future we will see Coffee Lake on 14++ become part of the 8th Gen, as well as Cannon Lake on 10nm.

Intel’s Core Architecture Cadence (8/20)
Core Generation Microarchitecture Process Node Release Year
2nd Sandy Bridge 32nm 2011
3rd Ivy Bridge 22nm 2012
4th Haswell 22nm 2013
5th Broadwell 14nm 2014
6th Skylake 14nm 2015
7th Kaby Lake 14nm+ 2016
8th Kaby Lake Refresh
Coffee Lake
Cannon Lake
14nm+
14nm++
10nm
2017
2017?
2018?
9th Ice Lake?
10nm+ 2018?
Unknown Cascade Lake (Server) ? ?

Now the Generation name is no longer in direct correlation with underlying core microarchitecture or lithography process. This is going to confuse some users and anger others, although Intel’s official line is along the lines of the fact that lithography process nodes are harder to optimize, smaller nodes benefit in yield from smaller cores and as such their product portfolio has to expand beyond traditional naming in order to provide the appropriate product and the appropriate price point.

In our pre-briefings, Intel only mentioned Coffee Lake in the context of the fact that today’s launch is not Coffee Lake. Because media were expecting this to be Coffee Lake (and expecting it to be a desktop processor launch), the question ‘is this Coffee Lake’ was actually asked several times, and the answer had to be repeated. These four new CPUs are still Kaby Lake CPUs built on the same 14+ technology, with minor updates, and bringing quad cores to 15W.

So when is Coffee Lake on 14++ (or Cannon Lake) coming? Intel only stated that other members of the 8th Generation family (which contains Kaby Lake Refresh, Coffee Lake and Cannon Lake) are coming later this year. Desktop will come in the autumn, and additional products for enterprise, workstation and enthusiast notebooks will also happen. As for today’s 8th Generation U-series announcement, Intel tells us that we should start seeing laptops using the new CPUs hit the market in September.

Update: Along with the product specs for the new mobile SKUs, Intel has also uploaded the new box art for the desktop 8th Gen Core parts to their website. The boxes confirm, among other things, that once these desktop parts will launch they’ll have 6 cores (with HT for the i7) and require 300 series motherboards.

Related Reading