Deciding between building a mainstream PC and a high-end desktop has historically been very clear cut: if budget is a concern, and you're interested in gaming, then typically a user looks to the mainstream. Otherwise, if a user is looking to do more professional high-compute work, then they look at the high-end desktop. Over the course of AMD’s recent run of high-core count Ryzen processors that line has blurred. This year, that line has disappeared. Even in 2016, mainstream CPUs used to top out at four cores: today they now top out at sixteen.

Does anyone need sixteen cores? Yes.

Does everyone need sixteen cores? No.

There are two fundamental drivers for most PC builders: cost and performance. Users who want a gaming machine are going to put their dollars in what gives them the best gaming performance. Users that want to edit video are going to look at content creation focused hardware. For those in the business world, the added incentive of extra performance is being able to offset or amortize those costs with an improved work rate. For the video editor needing a week per video, if they can spend +40% to reduce the render time by half then it can pay off over a short period of time.

As we move through 2019, users are doing more with their systems. Even at the low end, users might have double monitors where they game and watch their favourite streamer at the same time. High end users might reserve certain cores for different tasks, ensuring that there’s always some horsepower for the high-throughput tasks or virtual machines. Even though processors became ‘multi-core’ over a decade ago, we all as users are only recently adjusting how we do things to be more parallel, and the hardware is coming up to match our demands.

To that end, AMD’s Ryzen processors have been timely. The first generation mainstream Ryzen hardware in 2017 was a breath of fresh air in a market that had become sufficiently stale to be unexciting. With the color drained, AMD’s Ryzen enabled up to eight cores on a single CPU, and at the time aimed to throw its weight against Intel’s hardware in the class above. The new architecture didn’t push ahead on day one clock for clock, but it enabled a different paradigm at an obscenely reasonable price point.

Enter round 2, and Zen 2. Earlier this year AMD pushed again, this time putting 12 cores in the market for the same price as 8, or what had been the 4-core price point only three years prior. In three years we had triple the cores for the same price, and these cores also have more raw performance. The frequency wasn’t as high as the competition, but this was offset by that raw clock-for-clock throughput and ultimately where the competition now offered eight cores, AMD offered 12 at a much lower power consumption to boot.

Today is round 2 part 2: taking that same 12-core processor, and adding four more cores (for a 50% increase in price), and not only going after the best consumer processor Intel has to offer, but even the best high-end desktop processor. This is AMD squeezing Intel’s product portfolio like never before. What exactly is mainstream, anyway?

AMD’s new Ryzen 9 3950X has a suggested retail price of $749. For that AMD is advertising sixteen of its latest Zen 2 cores built on TSMC’s 7nm process, running at a 3.5 GHz base frequency and a 4.7 GHz single-core turbo frequency. The TDP of the chip is rated at 105 watts and it has 24 PCIe 4.0 lanes as well as dual memory channels that support up to 128 GB of DDR4-3200.

AMD 'Matisse' Ryzen 3000 Series CPUs
AnandTech Cores
TDP Price
Ryzen 9 3950X 16C 32T 3.5 4.7 8 MB 64 MB 16+4+4 1+2 105W $749
Ryzen 9 3900X 12C 24T 3.8 4.6 6 MB 64 MB 16+4+4 1+2 105W $499
Ryzen 9 3900 12C 24T 3.1 4.3 6 MB 64 MB 16+4+4 1+2 65W OEM
Ryzen 7 3800X 8C 16T 3.9 4.5 4 MB 32 MB 16+4+4 1+1 105W $399
Ryzen 7 3700X 8C 16T 3.6 4.4 4 MB 32 MB 16+4+4 1+1 65W $329
Ryzen 5 3600X 6C 12T 3.8 4.4 3 MB 32 MB 16+4+4 1+1 95W $249
Ryzen 5 3600 6C 12T 3.6 4.2 3 MB 32 MB 16+4+4 1+1 65W $199
Ryzen 5 3500X 6C 6T 3.6 4.1 3 MB 32 MB 16+4+4 1+1 65W OEM

It wasn’t too long ago that this price range used to be the realm of AMD’s high-end desktop Threadripper processors, which started at 8 cores and we up to 32 cores. AMD is now shifting that paradigm as well, with this 16-core chip being at $749, and AMD’s next generation Threadripper 3000 processors starting at 24-cores at $1399. When AMD CEO Dr. Lisa Su was asked earlier this year what would happen given the drive to more cores for the mainstream processors, her response was ‘as Ryzen goes up, Threadripper goes up-up’. This is the realization of that.

It is worth noting that the price is likely to be higher at retail initially, as demand is expected to be high and stock levels haven’t been defined – given the popularity of the 12-core chip, it would seem that users wanting the mainstream platform always want the best.

Going AM4: The Battle with Motherboards

When the AM4 platform was first launched, technically with pre-Zen hardware, it supported four cores. The same platform now goes all the way up to sixteen cores, which is no small task. The flip side of this comes down to motherboard support: some AM4 motherboards were not designed with high-power sixteen core processors in mind. Some motherboards built on the AM4 socket were for the budget market, and will struggle when it comes to this 16-core part.

AMD has attempted to at least segment its AM4 market a little. Only the latest AM4 chipset, the X570 chipset, has official support for the Ryzen 3000-series PCIe 4.0 connections. In order to enable the PCIe 4.0 lanes on the processor as qualified by AMD, users will have to purchase an X570 motherboard, otherwise these lanes will run at half speed (PCIe 3.0) in non-X570 motherboards.

The quality of the motherboard is likely to affect turbo frequencies as well. AMD’s turbo algorithms are influenced in part by the ability of the power delivery to push current through from the power supply. We are seeing X570 motherboards range from $170 all the way up to $999. This isn’t saying that doubling the cost of the motherboard will double the ability to turbo, but as seen with the previous Ryzen 3000 series chips, the motherboard choice (as well as the cooling it uses) will matter.

All the X570 motherboards we’ve tested recently are up to the task of taming the Ryzen 9 3950X. Here’s a list of what we’ve tested:

Users looking at motherboards have to find the right mix of capacity, cost, and features. We did a visual inspection of all 35+ launch models.

Toe-to-Toe: Intel Core i9-9900KS / Core i9-9980XE / Core i9-10980XE

With the mainstream and high-end desktop market now seemingly merging, there are many angles to consider different competitive parts between Intel and AMD. If we compete purely on PCIe lanes, then we might put the Core i9-9900KS (8-cores) up against the 3950X (16-cores), although there is a big price difference ($513 vs. $749). If we compare on pricing, the nearest processor to the 3950X would be either the 9900KS (mainstream) or the Core i9-10940X ($729), however while 3950X has more cores than either, but doesn’t have as many PCIe lanes/memory lanes as the 10940X. If we go for core count, then Intel’s sixteen Core i9-9960X would be the obvious candidate, although this CPU is a lot more expensive (until Intel reduces the price) and is technically an X299 processor, so has more PCIe lanes and memory channels.

Unlocked CPU Pricing
and Select Others
(MSRP Pricing)
Cores AnandTech Cores Intel*
(OEM Pricing)
    $900-$999 18/36 Core i9-10980XE ($979)
Ryzen 9 3950X ($749) 16/32 $700-$799 14/28 Core i9-10940X ($784)
    $600-$699 12/24 Core i9-10920X ($689)
    $500-$599 10/20
Core i9-10900X ($590)
Core i9-9900KS ($513)
Ryzen 9 3900X ($499) 12/24 $400-$499 8/16 Core i9-9900K/F ($488)
Ryzen 7 3800X ($399) 8/16 $350-$399 8/8 Core i7-9700K/F ($374)
Ryzen 7 3700X ($329) 8/16 $300-$349    
    $250-$299 6/6 Core i5-9600K ($262)
Ryzen 5 3600X ($249) 6/12 $200-$249    
Ryzen 5 3600 ($199) 6/12 Below $200 4/4 Core i3-9350K ($173)
*Intel quotes OEM/tray pricing. Retail pricing will sometimes be $20-$50 higher.

There is no easy comparison between any of the processors. AMD is pushing the boundaries of the mainstream dual channel memory processor regime, and Intel doesn't have an equivalent in that space. Intel can match it in the high-end desktop space, but therein lays other issues with PCIe lane counts and memory channel support disparity between the two, as well as Intel’s current retail options being high-priced variants. Intel’s published next generation hardware is set to be launched sometime in November, and with it a number of price cuts, however given the known differences between Intel’s current and Intel’s next generation processor line, the performance gain is not expected to be particularly big.

Going For Power: Is 105W TDP Accurate?
Comments Locked


View All Comments

  • Drazick - Sunday, November 17, 2019 - link

    The DDR Technology is orthogonal.
    I want Quad and the latest memory available.
  • guyr - Friday, December 20, 2019 - link

    Anything is possible, of course. 5 years ago, who would have predicted 16 cores in a consumer-oriented CPU? However, neither Intel nor AMD has made any moves beyond 2 memory channels in the consumer space. The demand is simply not there to justify the increase in complexity and price. In the professional space, more channels are easily justified and the target market doesn't hesitate to pay the higher prices. So, it's all driven by what the market will bear.
  • alufan - Saturday, November 16, 2019 - link

    weird intel launches its chip a couple of weeks ago and it stayed upfront and main story for over a week, AMD launches what is in effect the best CPU ever tested by this site and it lasts a few Days before being pushed aside for another intel article am sure the intention by the reporters is to be fair and unbiased however I can see how the commercial motives of the site are being manipulated looks like intels up to its old tricks again, the thread ripper article lasted even less time but no chips have been tested(or at least released) yet which I guess makes sense
  • penev91 - Sunday, November 17, 2019 - link

    Just ignore everything Intel/AMD related on Anandtech. There's been an obvious bias for years.
  • Atom2 - Saturday, November 16, 2019 - link

    There has never been a situation as big as this one, where the bench software was benchmarked more than the hardware. Comprehensive overview of historic software development? Whatever the reason, it seems that keeping back AVX512 to only select few CPUs, was an unfortunate decision by Intel, which only contributed to the situation. Yes, you know, if you compile your code with compiler from 1998 and ignore all the guidelines how to write fast code ... Voila... For some reason however, nobody tries to run 20 year old CPU code on GPU though.
  • chrkv - Monday, November 18, 2019 - link

    Second page "On the Ryzen High Performance power plan, our sustained single core frequency dropped to 4450 MHz" - I believe just "the High Performance" should be here.
    Page 4 "Despite 5.0 GHz all-core turbo being on the 9900K" - should be "9900KS".
  • Irata - Tuesday, November 19, 2019 - link

    Quick question: Are any of your benchmarks affected by the Mathlab issue (Ryzen are crippled because a poor code path is used due to a vendor ID check for "genuine Intel" )?
  • twotwotwo - Tuesday, November 19, 2019 - link

    Intel's had these consumer-platform-based "entry-level Xeons" (once E3, now E) for a while. Despite some obvious limits, and that there are other low-end server options, enough folks want 'em to seed an ecosystem of rackmount and blade servers from Supermicro, Dell, etc.

    Anyway, the "pro" (ECC/management enabled) variant of Ryzen seems like a great fit for that. 16 cores and 24 PCIe 4 lanes are probably more useful for little servers than for most desktop users. It's also more balanced than the 8/16C EPYCs; it's cool they have 128 lanes and tons of memory channels, but it takes very specific applications to use them all with that few cores (caching?). Ideally the lesser I/O and lower TDPs also help make denser/cheaper boxes, and the consumer-ish clocks pay off for some things.

    The biggest argument against is that the entry-level server market is probably shrinking anyway as users rent tiny slices of huge boxes from cloud providers instead. It also probably doesn't have the best margins. So maybe you could release a competitive product there and still not make all that much off it.
  • halfflat - Thursday, November 21, 2019 - link

    Very curious about the AVX512 vs AVX2 results for 3dPM. It's really unusual to see even a 2x performance increase going from AVX2 to AVX512 on the same architecture, given that running AVX512 instructions will lower the clock.

    The non-AVX versions, I'm presuming, are utilizing SSE2.

    The i9-9900K gets a factor of 2 increase going from SSE2 to AVX2, which is pretty much what one would expect with twice as many fp operations per instruction. But the i9-7960X performance with AVX512 is *ten times* improved over SSE2, when the vector is only four times as wide and the cores will be running at a lower clock speed.

    Is there some particular AVX512-only operation that is determining this huge performance gap? Some further analysis of these results would be very interesting.
  • AIV - Wednesday, November 27, 2019 - link

    Somebody posted that it's caused by 64 bit integer multiplies, which are supported in AVX512, but not in AVX2 and thus fallback to scalar operations.

Log in

Don't have an account? Sign up now