New Single Thread Performance Champion?

Aside from rough microarchitectural performance claims, we don’t typically talk much about first-party benchmark numbers – they are unverified and we’d rather hold station until the reviews where we can talk about our own testing. There are a couple of numbers that I do want to pull out of AMD’s marketing however, if only to put some markers in the sand compared to numbers we already have.

One of the favorite benchmarks that both Intel and AMD have used over the years is Cinebench, a test scene render for Cinema4D. AMD has been promoting it with Ryzen because it holds a performance lead, whereas Intel prefers to highlight it as an example of a less-than-ideal representative workload. We use it in our test suite as only one of a number of rendering workloads, covering a large user base. From our numbers, Intel and AMD’s best processors have the following results in a single threaded test:

CineBench R20 Single Thread Score
(As Measured at AnandTech)
AnandTech uArch Process 1C Turbo Score
Intel Core i7-1185G7 Tiger Lake Intel 10nm SuperFin 4.8 GHz 595pts
Intel Core i9-10900K Comet Lake Intel 14+++ 5.3 GHz 538pts
AMD Ryzen 9 3950X Zen2 TSMC 7nm 4.7 GHz 536pts
AMD Ryzen 9 3900XT Zen2 TSMC 7nm 4.7 GHz 523pts

(Note that in the single threaded test, the power limits ultimately should not apply because one core should not consume all the power of the chip. For the Tiger Lake processor, because this is a nominal 15 W TDP part with a 50 W turbo, this actually does go above the power limit with one core active, as it scores 554. As a result, the 50 W mode with a 28 W TDP was used and scores 595. This is more akin to a desktop processor anyway.)

AMD’s number for its Ryzen 9 5900X, at 4.8 GHz turbo, is 631.

CineBench R20 Single Thread Score
(As Measured by AMD)
AMD uArch Process 1C Turbo Score
AMD Ryzen 9 5900X Zen3 TSMC 7nm 4.8 GHz 631pts

Generation-on-generation, this would be a 17.8% performance increase. Over Intel’s highest performing desktop Core i9, that would be a lead of 17.3%. Over Intel’s latest generation microarchitecture, Tiger Lake, this is a 6% performance increase at the same frequency.

Whichever way you slice it, whether you consider the lead a 17% lead from desktop-to-desktop, or a 6% lead from best-to-best, one thing is clear that if this result is correct, AMD will have the single threaded performance crown.

The previous time that AMD truly held this title was over a decade ago, and ever since the company has been playing catch-up. The argument over having enough single thread performance, if these results are correct, is one that AMD can fight toe-to-toe on. Not only that, but in AMD’s marketing materials, they believe it is a clear cut winner.

Mainstream Gaming

This sort of metric also extends out to gaming, which has been a contentious issue since Ryzen enabled AMD to play in this market again. AMD’s numbers here focus on 1080p gaming for two reasons: firstly it’s an area where CPUs are more important for performance, but secondly because they cite that 1080p monitors are still the bulk of monitor sales for gamers, especially when looking at high refresh rates.

For 1080p high quality gaming, AMD is making two clear claims.

Firstly, when comparing the Ryzen 9 5900X to the Ryzen 9 3900XT, the new processor is stated to boost average frame rates anywhere from +5% up to +50%, with an average around +26%.

Secondly, when comparing the Ryzen 9 5900X to the Core i9-10900K, the new processor is stated to win by 7%, with individual benchmarks varying from a 3% loss to a 21% win.

This latter one is important, as when the Core i9-10900K was released, there as a strong marketing message that this was the best gaming CPU available on the market. Now AMD believes it can make that claim and take the crown. The important element for AMD to manage however is availability – the availability of the 10900K has been so varied that Intel quietly released the 10850K to manage expectations. If AMD can put enough Ryzen 9 5900X processors into the hands of people that want them, then Intel has a bigger step to climb to regain the title when it will reportedly launch its Rocket Lake processors in Q1 next year.


AMD’s Claims

As always, we will wait until we have the hardware on hand to form our own conclusions. These results are AMD first party tests, so the usual standards of caution apply. AMD has a strong history of making accurate claims, confirmed by our testing in recent years, but we still recommend waiting for our review.

Chipset Support

As always, AMD is keen to highlight that the new Ryzen 5000 processors are supported in currently available motherboards. Due to the long life cycle of the AM4 platform, there has been some segmentation between chipset and processor support, however AMD this time around has made it very clear when it comes to supported chipsets. AMD has split its instructions in two.

For 500 series chipset users, motherboards firmware that supports the new hardware should already be available. Users will need to ensure that a minimum AGESA version of 1.0.8.0 is installed which will ensure that the system will boot. For full support and performance, users should update to AGESA 1.1.0.0 for the best experience.

For 400 series chipset users, AMD is currently hard at work with motherboard partners to update their firmware stacks. AMD expects the first beta BIOSes with Ryzen 5000 support to be made available from January, however part of this will be motherboard manufacturer dependent. Full release BIOSes will be at a later date. When asked if the new 5000-series firmware for 400-series chipsets will remove support for older processors, AMD said it would be on a case-by-case basis depending on how the motherboard vendor wanted to enable support.

For users that purchase motherboards that do not have the required BIOS support for Ryzen 5000, the Processor Loan Boot Kit programme will still be available.

A Word on Sticker Shock

It is hard not to notice that AMD is raising the prices of its hardware for this new generation. The company believes that it is truly in a winning position, and that better hardware in the market deserves to be priced at a level that matches this. For any competitive business owner with a belief of a better product, finding that right balance of demand and price is always critical to how to manage sales and ultimately market share.

Anand once said a very insightful phrase to me. “There are no bad products, only bad prices.”

So with AMD asking for another $50 on these processors, is that a lot to ask?

AMD’s argument is that, for performance per dollar, it still retains a healthy lead over its competition in each market segment. The spokespersons were keen to point out that they still remember where AMD has come from to reach this level of performance, and understand that a metric such as performance per dollar has always been high on the list of requirements from its most passionate users. The difference this time around is that, because AMD is claiming it has the best performance on the market, it can now charge that slight premium but still offer a more compelling product.

The success of AMD’s Ryzen 3000 family is easy to see. It offers a very attractive performance per dollar proposition, and nine of the top ten processors on Amazon’s best sellers list are from AMD. Here’s that list, with the pricing as given in our buyer’s guide last week:

Price Options
[#] is Amazon Best-Seller Position
# AMD Price
[1] Ryzen 5 3600 $205
[2] Ryzen 7 3700XT $295
[3] Ryzen 5 3600X $209
[4] Ryzen 5 2600 $149
[5] Ryzen 9 3900X $430
[6] Ryzen 7 3800XT $340
[7] Ryzen 5 3600XT $230
[8] Ryzen 7 2700X $218
[9] Ryzen 3 3200G $100

This is an impressive list, however there a theme I have noticed. Starting from the top, the first four best-selling processors are below $299. The lowest Ryzen 5000-series processor in this launch comes in at the $299 price bracket. Out of that top nine, only #5 and #6 are above $299. This is a clear indication of where the bulk of the market is, especially as AMD is likely in increasing both market share and revenue.

In discussing the pricing with AMD, I noted that the new Ryzen 5000 processors are not only replacing hardware with a $50 higher MSRP, but also replacing hardware that routinely sells below MSRP. This makes the differences more akin to $90-$150. This of course changes some of the dynamic when we start discussing performance per dollar.

AMD’s response to this commentary was the one I would have given if I were in their position. The Ryzen 5000 series is a new product, and the claim of market leading performance means that the early adopters and AMD enthusiasts that want the best on day one will be able to get the hardware they desire. During the initial phase in almost all launches, users looking for the best bang-for-buck build will always look to purchasing the previous generation, which is almost always offered at a good discount as stock transfers to the latest product. AMD believes it has set the pricing of the new Zen 3 processors where it remains competitive, but still balances the message that AMD claims it has the best, most efficient processors.

What Now?

For users looking for processors under that $299 mark, or more offerings at the 65 W TDP, we expect AMD to look at expanding the Ryzen 5000 product list over the next six months. No exact word from AMD was given about what might come in the future. I even asked about Threadripper with Zen 3, but as expected there was no official comment.

As for the processors announced today, November 5th is the date to put in your diary. On this date, reviews will go live, details about the microarchitecture will be revealed, and processors will be available. We’re already pre-testing a LOT of hardware for our review. Let us know what comparisons matter most to you.

 
AMD Ryzen 5000 and Zen 3 on Nov 5th
Comments Locked

510 Comments

View All Comments

  • Oxford Guy - Monday, October 12, 2020 - link

    As for “what end?”, an article here already answered that and said testing data will be forethcoming.

    DDR4 is old and AMD could have made Zen3’s inability to support both DDR4 and DDR5 less of a drawback by giving the chip the ability to support quad channel RAM. Perhaps there will be a third TR refresh that will do that.
  • Halfprice - Monday, October 12, 2020 - link

    "Threadripper doesn't have quad channel RAM". Uhhh, yes it does. All Threadrippers ever made have at least four memory channels. Threadripper Pro which is OEM only has eight memory channels like Epyc.
  • Spunjji - Tuesday, October 13, 2020 - link

    "Threadripper doesn’t have quad channel RAM"
    Wrong.

    Zen 3 Threadripper hasn't been announced yet, but you can guarantee when it does it will use Quad Channel because, once again, Zen 2 Threadripper already does.

    My "complaint" (not a complaint) about a new socket isn't "weak" because it has nothing to do with Threadripper. We were talking about mainstream desktop until you decided to start moving the goalposts around. 🙄
  • awev - Saturday, October 10, 2020 - link

    Looks like the 5000 series has a price increase of 6.7% to 20% compared to their 3000 series counterparts, for a 19% performance increase in the IPC. So for the 5950 the 6.7% increase in price makes the 19% performance increase a better value than the 20% price increase for the 5600.
  • Halfprice - Tuesday, October 13, 2020 - link

    It stands to reason given AMD's situation. Everything they make gets sold and it's hard to get a hold of many of their products. They are production capacity constrained and are launching the best ever processors to whom and at what price should that production capacity go?
  • liquid_c - Saturday, October 10, 2020 - link

    This article seems more like a PR note on AMD’s behalf rather than a normal launch (make that announcement) one.
    Sad part is, those frequencies that AMD claims will (barely) be obtainable on one core and one core only (especially if the recent events, as shown by Tom’s, are to be taken into account) while Intel’s offerings can actually go on 5ghz on all cores (even more, on their higher end offerings, when proper cooling is present). So this means that any game that uses more than 1 core will heavily favour Intel.
    I’m mighty curious about the 1080p gaming tests. Those are most relevant when it comes to CPUs.
  • Qasar - Saturday, October 10, 2020 - link

    liquid_c , sad part is, no one knows that for sure, unless i missed something and you can provide a link, i dont recall any one talking about overclocking, or how high each core gets to, or max all core.
    "So this means that any game that uses more than 1 core will heavily favour Intel. " we will find out for sure on nov 5th
    TBH, who cares about 5ghz, its just a number, and only intel needs to go that high.
  • Spunjji - Sunday, October 11, 2020 - link

    Launches are PR announcements. Or did you forget that Intel just launched Tiger Lake with no products and a PR platform sent to reviewers that, it turns out, really doesn't represent performance in shipping products?
  • Spunjji - Sunday, October 11, 2020 - link

    You also failed to demonstrate how you suppose that "any game that uses more than 1 core will heavily favour Intel" but that's okay, we can safely presume you pulled that out of your ass.
  • shady28 - Sunday, October 11, 2020 - link

    Well for Tiger Lake, what you're saying is false. One google and you'll find plenty of reviews of a pre-prod laptop that was provided within a week or so of launch, speculated at the time to be an MSI laptop. In the past couple of weeks there are a number of reviews of shipping product showing up in multiple places.

Log in

Don't have an account? Sign up now