With the Ryzen 5000 series, it's fair to say that AMD has finally, and fully, eclipsed Intel's performance dominance in desktop PCs. AMD’s flagship $799 Ryzen 9 5950X has landed in our labs, boasting 16 cores and 32 threads bristling with the potent new Zen 3 microarchitecture. AMD’s new halo part expands Ryzen 9’s dominating lead in productivity applications and beats Intel’s competing processors in every other metric, including 1080p gaming performance, by surprising margins. Our 5950X sample even breaks the 5 GHz barrier at stock settings (at least sporadically), outstripping its spec and making it an easy choice for our list of Best CPUs.
But the Ryzen 9 5950X is just the tip of the Zen 3 spear. We also have the more amenable $549 Ryzen 9 5900X that comes with 12 cores and 24 threads. Aside from its bruising performance in applications, it’s even faster than the 5950X in gaming, even beating out Intel's overclocked flagships at 1080p, too.
Much of Ryzen’s early success stemmed from industry-leading core counts and plenty of freebies for enthusiasts, like bundled coolers and unrestricted overclockability paired with broad compatibility. Still, AMD was long relegated to the role of a value alternative.
AMD’s clockwork execution on new Zen architectures has slowly whittled away Intel’s performance superiority with each new launch, though, leaving Intel an ever-shrinking cross-section of advantages. To counter, Intel added more cores and features of its own, but AMD’s relentless innovation left Intel clinging to the life raft of its single-threaded performance advantage.
AMD narrowed the gap when it transitioned to the denser 7nm process and Zen 2 architecture for the Ryzen 3000 chips, which largely reduced Intel’s gaming advantage to the imperceptible level – particularly in the mid-range of the market. With sales surging, AMD has begun to capitalize by repositioning itself as a premium brand. The first signs of that shift began with the company’s recent Ryzen XT lineup, which found the company largely discarding some of the freebies we’ve become accustomed to and tacking on a higher price tag to its almost imperceptibly-faster chips.
|Zen 3 Ryzen 5000 Series Processors||RCP (MSRP)||Cores/Threads||Base/Boost Freq.||TDP||L3 Cache|
|Ryzen 9 5950X||$799||16 / 32||3.4 / 4.9 GHz||105W||64MB (2x32)|
|Ryzen 9 5900X||$549||12 / 24||3.7 / 4.8 GHz||105W||64MB (2x32)|
|Ryzen 7 5800X||$449||8 / 16||3.8 / 4.7 GHz||105W||32MB (1x32)|
|Ryzen 5 5600X||$299||6 / 12||3.7 / 4.6 GHz||65W||32MB (1x32)|
Ryzen 5000 changes the game entirely, though. The chips come with the same refined 7nm process found in the Ryzen XT processors, but AMD paired the node with a ground-up redesign of the Zen core microarchitecture. AMD says the new Zen 3 microarchitecture provides a 19% average increase in instruction per cycle (IPC) throughput, erasing the last vestiges of Intel’s performance advantages while delivering a new level of power efficiency.
According to our tests, the Ryzen 5000 processors deliver, beating Intel in nearly all metrics that matter, including performance, power consumption, and thermals, and largely remove Intel’s performance lead after overclocking. And yes, that includes in 1080p gaming. AMD is also leveraging its position as the only CPU maker that also makes discrete GPUs by rolling out its new Smart Memory Access feature. This new tech boosts gaming performance by enhancing data transfer performance between the CPU and GPU, but it only works if you have a Radeon RX 6000 graphics card, Ryzen 5000 processor, and a 500-series motherboard. We won’t know the full implications of this new tech until the Radeon RX 6000 “Big Navi” launch later this month, but it looks promising.
Now that Ryzen 5000 firmly establishes AMD as the performance leader, the company has hiked up prices by $50 across its entire lineup and left a noticeable gap in its product stack – you'll have to take a steep $150 step up the pricing ladder to get above the entry-level six-core twelve-thread Ryzen 5 5600X. AMD's premium pricing could be an Achilles heel, but it's hard to determine the final pricing story given that AMD's suggested selling prices almost never manifest at retail.
Meanwhile, Intel is left without a response until the first quarter of 2021 when its Rocket Lake chips blast off, bringing a new back-ported Cypress Cove architecture that grants a “double-digit” IPC increase paired with Intel's never-ending 14nm process.
Until then, this is how the high-performance chip market stacks up. To put AMD’s gaming performance claims to the test, we’ve switched over to an Nvidia GeForce RTX 3090 for game testing and also put the company’s new silicon through the paces in a wide range of expanded tests, including several SPEC and Adobe suites.
Ryzen 9 5950X and Ryzen 9 5900X Specifications and Pricing
The Ryzen 5000 series processors come as four models that span from six cores and twelve threads up to 16 cores and 32 threads. AMD increased its Precision Boost clock rates across the board, with a peak of 4.9 GHz for the Ryzen 9 5950X. However, AMD’s unique boosting algorithms can stretch beyond the advertised speeds if you pair the chips with a quality cooler and a motherboard with robust power circuitry. In fact, our Ryzen 9 5950X sample peaked at 5 GHz at stock settings, albeit sporadically, and reached 5.125 GHz when we engaged the auto-overclocking Precision Boost Overdrive feature, which we'll cover below.
AMD increased the boost clock speeds, but it also reduced base frequencies compared to the previous-gen processors. AMD says that if you top the chip with an adequate cooler, it will rarely (if ever) drop to the base frequency, which we confirmed with our testing further below.
|Zen 3 Ryzen 5000 Series Processors||RCP (MSRP)||Cores/Threads||Base/Boost Freq.||TDP||L3 Cache|
|Ryzen 9 5950X||$799||16 / 32||3.4 / 4.9||105W||64MB (2x32)|
|Core i9-10980XE||$815 (retail)||18 / 36||3.0 / 4.8||165W||24.75MB|
|Ryzen 9 3950X||$749||16 / 32||3.5 / 4.7||105W||64MB (4x16)|
|Ryzen 9 5900X||$549||12 / 24||3.7 / 4.8||105W||64MB (2x32)|
|Core i9-10900K / F||$488 - $472||10 / 20||3.7 / 5.3||125W||20MB|
|Ryzen 9 3900XT||$499||12 / 24||3.9 / 4.7||105W||64MB (4x16)|
|Ryzen 7 5800X||$449||8 / 16||3.8 / 4.7||105W||32MB (2x16)|
|Core i9-10850K||$453||10 / 20||3.6 / 5.2||95W||20MB|
|Core i7-10700K / F||$374 - $349||8 / 16||3.8 / 5.1||125W||16MB|
|Ryzen 7 3800XT||$399||8 / 16||3.9 / 4.7||105W||32MB (2x16)|
|Ryzen 5 5600X||$299||6 / 12||3.7 / 4.6||65W||32MB (1x32)|
|Core i5-10600K / F||$262 - $237||6 / 12||4.1 / 4.8||125W||12MB|
|Ryzen 5 3600XT||$249||6 / 12||3.8 / 4.5||95W||32MB (1x32)|
The $799 16-core 32-thread Ryzen 9 5950X comes with a 3.4 GHz base frequency, a 300 MHz reduction compared to the 3950X, and a 4.9 GHz Precision Boost frequency. Intel doesn't really have an answer for the 5950X; the Comet Lake series tops out at ten cores for $488. You can find the 18-core 36-thread Core i9-10980XE for $815 at several retailers, though it comes with all of the normal drawbacks of a high end desktop chip, like the need for a pricey motherboard and quad-channel memory kit. Meanwhile, the Ryzen 9 5950X drops into mainstream motherboards with ease.
The 12-core 24-thread $549 Ryzen 9 5900X comes with a $50 markup over the previous-gen 3900XT. The chips' base frequency declines 200 MHz compared to the 3900XT, but boosts reach 4.8 GHz (a 100 MHz increase). Intel's 10-core 20-thread Core i9-10900K slots in for $60 less than the 5900X ($77 less if you choose to go with the graphics-less F-series model).
If all you care about is gaming, Intel's $453 Core i9-10850K also falls into this bracket. The 10850K offers essentially the same performance as the pricier 10900K in gaming, but is $96 less than the 5900X.
The $449 Ryzen 7 5800X comes with eight cores and 16 threads, just like its previous-gen Ryzen 7 3800XT counterpart, but again comes with a $50 markup. The chip sees a 100 MHz lower base clock than the 3800XT but has the same 4.7 GHz boost. Given the price point, the Core i9-10850K also competes here with similar pricing to the 5800X, while the Core i7-10700K is ~$100 less.
Finally, the 6-core 12-thread $299 Ryzen 5 5600X's base clocks come in at 100 MHz less than the previous-gen 3600XT, while boosts are 100 MHz higher at 4.6 GHz. AMD's 6C/12T Ryzen 5 3600XT had a 95W TDP, but AMD dialed that back to 65W with the 5600X, showing that Zen 3's improved IPC affords lots of advantages.
AMD does have a glaring hole in its product stack: You'll have to shell out an extra $150 to step up from the 6C/12T Ryzen 5 5600X to the 8C/16T Ryzen 7 5800X, which is a steep jump. Based upon product naming alone, it appears there is a missing Ryzen 7 5700X in the stack, but it remains to be seen if AMD will actually bring such a product to market.
As before, AMD only guarantees its boost frequencies on a single core, and all-core boosts will vary based on the cooling solution, power delivery, and motherboard BIOS. You’ll need your own cooler for any Ryzen 5000 chip that exceeds a 65W TDP: The Ryzen 5 5600X is the only Ryzen 5000 chip that comes with a bundled cooler. AMD said it decided to skip bundled coolers in higher-TDP models largely because it believes most enthusiasts looking for high-performance CPUs use custom cooling. AMD recommends a 280mm (or greater) AIO liquid cooler (or equivalent air cooling) for the Ryzen 9 and 7 CPUs if you want to reach the advertised speeds, significantly adding to the overall platform costs.
The Ryzen chips continue to expose 20 lanes of PCIe 4.0 to the user and stick with DDR4-3200 memory as the base spec. However, if the silicon lottery shines upon you, we found that the chips offer much better memory overclocking due to improved fabric overclocking capabilities, which we'll cover below.
These chips drop into existing AM4 motherboards with 500-series chipsets, like X570, B550, and A520 models. You'll need an AGESA 126.96.36.199 (or newer) BIOS to boot a Zen 3 processor. Still, while the early BIOS revisions ensure the processors will work on the most basic level, you'll have to update to an AGESA 188.8.131.52 (or better) BIOS for the best performance. AMD will also add support for 400-series motherboards starting in Q1, 2021, but that comes with a few restrictions.
AMD Zen 3 Ryzen 5000 Series Microarchitecture — The Quick Take
This article focuses on our standard performance, overclocking, and power consumption tests, but we'll update this section with more details on the Zen 3 microarchitecture along with cache, memory, and fabric latency tests. For now, here's a quick look at the highlight reel of Zen 3's improvements.
AMD's 19% IPC increase is the big headline feature of the Zen 3 microarchitecture, but it is especially impressive considering that it leveraged its existing Ryzen SoC to accomplish the feat. In fact, the base SoC is identical to the Ryzen 3000 processors: Zen 3 uses the same 12nm I/O Die (IOD) paired with either one or two 8-core chiplets (CCD) in an MCM (Multi-Chip Module) configuration. The IOD still contains the same memory controllers, PCIe, and other interfaces that connect the SoC to the outside world. Just like with the Matisse chips, the IOD measures ~125mm^2 and has 2.09 billion transistors.
The chiplets have been redesigned, however, and now measure ~80.7mm^2 and have 4.15 billion transistors. That's slightly larger than Zen 2's CCDs with ~74mm^2 of silicon and 3.9 billion transistors.
Just like the previous-gen Ryzen CPUs, processors with six or eight cores come with one compute chiplet, while CPUs with 12 or 16 cores come with two chiplets. AMD also used all of the existing SoC wire routing and packaging to maintain compatibility with the AM4 socket. As shown in the album above, the Infinity Fabric connections between the IOD and compute chiplets (CCD) remain the same and communicate at 32 Bytes-per-cycle for reads and 16B/cycle for writes. Even though these connections remain at the same speed as the previous-gen chips, the redesigned compute chiplets significantly reduce the amount of traffic that flows over the interface.
In the Zen 2 architecture (left), each Zen compute chiplet (CCD) contained two four-core clusters (CCXes) with access to an isolated 16MB slice of L3 cache. So, while the entire chiplet contained 32MB of cache, not all cores had direct access to all of the cache in the chiplet.
To access an adjacent slice of L3 cache, a core had to communicate with the other quad-core cluster by traversing the Infinity Fabric to the I/O die. The I/O die then routed the communication to the cache in the second quad-core cluster, even though it was contained within the same chiplet. To complete the transfer, the data had to travel back over the fabric to the I/O die, and then back into the adjacent quad-core cluster.
Each chiplet now contains one large unified 32MB slice of L3 cache, and all eight cores within the chiplet have full access to the shared cache. This improves not only core-to-cache latency but also core-to-core latency within the chiplet. AMD has also updated the CCD's core-to-core and cache interconnect to a ring topology.
While all eight cores can access the L3 cache within a single compute chiplet, in a dual-chiplet Zen 3 chip, there will be times that the cores will have to communicate with the other chiplet and its L3 cache. In those cases, the communications will still have to traverse the Infinity Fabric via signals routed through the I/O die.
Still, because an entire layer of external communication between the two four-core clusters inside each chiplet has been removed (as seen in the center of the chart above), the Infinity Fabric will naturally have far less traffic. This results in less contention on the fabric, thus simplifying scheduling and routing, and it could also increase the amount of available bandwidth for this type of traffic. All of these factors will result in faster transfers (i.e., lower latency) communication between the two eight-core chiplets, and it possibly removes some of the overhead on the I/O die, too.
These enhancements are important because games rely heavily on the memory subsystem, both on-die cache and main memory (DDR4). A larger pool of cache resources keeps more data closer to the cores, thus requiring fewer high-latency accesses to the main memory. Additionally, lower cache latency can reduce the amount of time a core communicates with the L3 cache. This new design will tremendously benefit latency-sensitive applications, like games — particularly if they have a dominant thread that accesses cache heavily (which is common).
However, the larger L3 cache does come with an increase in L3 latency to the tune of seven additional cycles.
Here's AMD's high-level bullet point list of improvements to the Zen 3 microarchitecture:
- Front-end enhancements:
- Major Design Goal: Faster fetching, especially for branchy and large-footprint code
- L1 branch target buffer (BTB) doubled to 1024 entries for better prediction latency
- Improved branch predictor bandwidth
- Faster recovery from misprediction
- "No Bubble" prediction to make back-to-back predictions faster and better handle branchy code
- Faster sequencing of op-cache fetches
- Finer granularity in switching of op-cache pipes
- Execution Engines:
- Major Design Goal: Reduce latency and enlarge to extract higher instruction-level parallelism (ILP)
- New dedicated branch and st-pickers for integer, now at 10 issues per cycle (+3 vs. Zen 2)
- Larger integer window at +32 vs Zen 2
- Reduced latency for select float and integer operations
- Floating point has increased bandwidth by +2 for a total of 6-wide dispatch and issue
- Floating point FMAC is now one cycle faster
- Major Design Goal: Larger structures and better prefetching — enhance execution engine bandwidth
- Overall higher bandwidth to feed larger/faster execution resources
- Higher load and store bandwidth vs. Zen 2 by +1
- More flexibility in load/store operations
- Improved memory dependence detection
- +4 table walkers in the Translation Look-aside Buffer (TLB)
Notably, AMD also added support for memory protection keys, added AVX2 support for VAES/VPCLMULQD instructions, and made a just-in-time update to the Zen 3 microarchitecture to provide in-silicon mitigation for the Spectre vulnerability.
Naturally, performance and power efficiency will improve as a function of architectural improvements. The reduced traffic on the Infinity Fabric also contributes (it always requires more energy to move data than to process it). Which brings us to IPC.
AMD Zen 3 Ryzen 5000 IPC
AMD chalks its 19% IPC improvement, which is the largest the company has seen in the post-Zen era, up to a number of Zen 3's architectural improvements. The company calculated its IPC improvements from 25 different workloads, including gaming, which seems a curious addition due to possible graphics-imposed bottlenecks, and some multi-threaded workloads. AMD's results show that the IPC improvements vary based on workloads, with up to a 39% improvement on the high end of the spectrum and a 9% improvement on the lower end.
We tested a limited subset of single-threaded workloads to see the clock-for-clock improvements, locking all chips to a static 3.8 GHz all-core clock with the memory dialed into the officially supported transfer rate (AMD used DDR4-3600 for its tests, which is technically an overclocked configuration).
AMD's generational march forward is clear as we move from the left to the right of each chart. Overall, AMD's gen-on-gen IPC increases are exceptional, and Zen 3's IPC obviously beats Intel's Comet Lake chips with ease.
AMD Ryzen 9 5950X Boost Frequency Testing
Before we jump right into the game tests (we have the particulars of our test configurations at the end of the article), we want to highlight the vastly improved clock speeds of the Ryzen 5000 processors we've tested — with the caveat that the silicon lottery applies, so your mileage may vary.
It appears that AMD has under-spec'd its clock rates to avoid another controversy. As per our normal routine, we put AMD's boost clocks to the test in both single- and multi-threaded workloads (methodology here). The lightly-threaded test regimen is designed to extract the highest boost clock rates possible as we step through ten iterations of the LAME encoder, then single-threaded POV-Ray and Cinebench runs, PCMark 10, and GeekBench. To keep the charts 'clean,' we only plot the maximum frequency recorded on any one core during the test.
We ran this series of tests at stock settings and charted the results in the first slide in the above album. Here we can see that the chip peaked at 5.05 GHz, and unlike with the Ryzen 3000 series processors, the unused cores (plotted in black) dropped to a much lower 2.2 GHz frequency (previous-gen chips tended to bottom out at 3.8 GHz). This is a vast improvement: AMD added the ability for individual idle cores to drop into sleep states quickly, which then reduces power consumption and heat generation. This ultimately allows the active cores to boost to higher frequencies, and for longer durations.
Things got interesting when we kicked on the auto-overclocking Precision Boost Overdrive feature with the 'advanced motherboard' setting. In the second slide, we can see the chip boosted to a peak of 5.125 GHz a few times, but reached 5.0 GHz on a more consistent basis. That's incredibly impressive given the rated spec of a 4.9 GHz boost clock.
The third slide plots our custom multi-threaded stress test that consists of multiple iterations of HandBrake, POV-Ray, Cinebench, v-ray, y-cruncher, and blender renders. This is basically throwing the heaviest real-world workloads we have in our arsenal at the chip to see if we can push any active cores below the 5950X's 3.4 GHz base clock. As you can see, the lowest clock frequency we measured on fully active cores weighed in at 3.6 GHz, which is an encouraging sign. Temperatures were also acceptable with the Corsair H115i cooler, peaking at 78C for short durations, albeit with the fans cranking away at high speed. Power consumption peaked right at the 142W PPT limit for brief periods.
AMD has issued guidance on expected frequencies and temperatures for the Ryzen 5000 processors, which we'll include right after this series of identical tests on the Ryzen 9 5900X.
AMD Ryzen 9 5900X Boost Frequency Testing
Here we can see that the Ryzen 9 5900X peaks at 4.95 GHz (frequently) at stock settings, but we didn't receive as much uplift when we engaged the PBO feature. In fact, we didn't measure as many peaks during the overclocked test run, but the chip still easily exceeds its rated 4.8 GHz boost clock.
Naturally, lesser coolers at more mundane settings will peak at higher temperatures. To help align expectations, AMD issued the above guidelines for expected temperatures for various kinds of coolers and the expected voltage ranges for various workloads. We have plenty of our own power testing after the gaming and application benchmarks.
Test Setup and Ryzen 5000 Overclocking
We've included our test system breakdowns at the end of the article. For this round of testing, we updated to Windows 10 Pro version 2004 (build 19041.450), the most recent build available at the time of testing, and all of the benchmarks you see below were generated within the last ten days. We also updated all of our drivers, firmwares, and testing programs to the newest available versions and transitioned from the Nvidia GeForce RTX 2080 Ti to the Gigabyte GeForce RTX 3090 Eagle to reduce graphics-imposed bottlenecks.
We didn't have time to fully explore all-core overclocking with the Ryzen 9 5950X and 5900X, but we weren't able to dial in overclocks that exceeded the all-core boost frequency. As such, we stuck with AMD's Precision Boost Overdrive (PBO), which boosts performance in multi-core workloads while maintaining the high single-core boost clocks.
We had great results with memory overclocking with the Ryzen 9 5900X — we dialed in a 2000 MHz fabric and DDR4-4000 at a 1:1:1 fclk/uclk/mclk ratio, beating the best results we've reached with the previous-gen Matisse processors due to the general limit of a 1900 MHz fabric with the previous-gen chips. Just dial up the CCD and IOD voltage to 1.15V (not higher than 1.2V), and you should be good to go to increase the fabric clock to 2000 MHz.
We weren't quite as lucky with the Ryzen 9 5950X, though, and settled for DDR4-3600 for our overclocked configurations below. We only ceded to the DDR4-3600 due to time constraints, though — I'm sure we can get DDR4-4000 stable with a bit more tuning.
Ryzen 9 5950X and Ryzen 9 5900X Gaming Performance — The TLDR
Here you can see the geometric mean of our gaming tests at 1080p and 1440p, with each resolution split into its own chart (overclocked results are shaded in grey).
For those accustomed to seeing Intel lead the gaming charts, these cumulative measurements might be a shock to your system: AMD's stock Ryzen 9 5950X and Ryzen 9 5900X lead Intel's heavily-overclocked Core i9-10900K and Core i7-10700K in our 1080p gaming suite, at least in terms of average frame rates (Intel's overclocked chips hold a slight lead in 99th-percentile measurements). We also see solid uplift with the Ryzen 9 5900X from overclocking (PBO and memory).
To put things in perspective, take a glance at the delta in 1080p gaming between the previous-gen Ryzen 9 3900XT, which basically runs overclocked right out of the box, compared to the Ryzen 9 5950X. That's a massive generational leap.
Flipping over to the 1440p chart brightens things up a bit for Intel, but only slightly — the overclocked Core i9-10900K returns to its normal spot at the top of the chart, and it still enjoys better 99th percentile frame rates after overclocking. However, AMD still beats Intel in both average and 99th-percentiles at stock settings, cementing the company's commanding lead.
3D Mark, VRMark, Stockfish Chess Engine
We run these synthetic gaming tests as part of our main application test script. We use an RTX 2080 Ti for these tests to facilitate faster testing, but we use the RTX 3090 for all other gaming benchmarks (we don't include these tests in the geometric mean listed above).
As we've come to expect, AMD's core-heavy processors dominate in threaded synthetic tests, like the Stockfish chess engine and 3DMark's DX11 and DX12 CPU tests, but perhaps Ryzen 5000 is most impressive in VRMark. This benchmark leans heavily on per-core performance (a mixture of IPC and frequency), and as you can see from the previous-gen Ryzen processors, AMD has traditionally trailed in this benchmark. Zen 3 rectifies that issue.
AMD says that the Ryzen 5000 processors offer leading performance in a large number of titles, but there will likely still be a period of time before we see targeted game updates to expose the best of Ryzen 5000, just like we saw with previous-gen Zen chips. Here Intel takes a clear lead in the 1440p benchmarks, with the Ryzen 9 5950X and 5900X trailing, albeit not by large margins. AMD's latest chips are much more competitive in the 1080p series of tests.
Far Cry 5
Far Cry 5 finds the overclocked Ryzen 9 5900X trading blows with the tuned Core i9-10900K running at 5.1 GHz at 1080p, which is quite the feat in itself. At stock settings, both Ryzen 5000 chips run neck-and-neck with the 10900K. We see similar trends with the 1440p benchmarks, but the Core i7-10700K reminds us that it has plenty of chops after overclocking, too.
Hitman 2 doesn't seem to scale well from 1080p to 1440p, at least not at the heightened fidelity settings we use for the benchmark, so we stuck with the 1080p test for this title because the same trends carry over to 1440p. Here we see the Ryzen 5000 processors take big strides, while the Ryzen 7 1800X reminds us just how far AMD has come in three short years.
Microsoft Flight Simulator
We're just as excited as anyone else about Microsoft's long long-overdue release of Flight Simulator, and we're sure that serious flight sim fans will want to crank up the resolution on this title. Here we can see that Intel holds a relatively slim lead after overclocking, but the stock Ryzen 5950X beats the 10900K while the 5900X pulls off a tie.
Project CARS 3
Project CARS 3 scales well with additional host compute, and the title obviously responds well to the Zen 3 architecture. Here we see the 5950X take a step back when we engage the auto-overclocking PBO feature, but that isn't entirely uncommon with AMD's auto-overclocking software. In either case, both Ryzen 5000 processors take a healthy lead over the stock 10900K and 10700K.
Red Dead Redemption 2
A glance at the bottom of these charts shows the clear progression of AMD's architectures as it iterated on the Zen design, but in most of the titles we tested, the Ryzen 5000 series represents AMD's biggest generational leap by far.
Shadow of the Tomb Raider
AMD's chips take a sizeable lead at 1080p, but Intel's overclocked chips deliver better 99th-percentile measurements. Flipping over to 1440p, Intel's 10900K reaches the top of the chart, but it took quite a bit of voltage for it to surpass the stock Ryzen 5000 chips.
The Division 2
The Division 2 ends up looking mostly GPU limited, even with the RTX 3090 — at least on the fastest processors. Intel remains competitive when overclocked, but at stock the Ryzen 5000 chips lead by a decent margin at 1080p. Not that anyone is likely to notice the difference between 180 and 195 fps in practice.
Ryzen 9 5950X and Ryzen 9 5900X Application Benchmarks
The geometric mean of both the most lightly- and heavily-threaded tests in our application suite speak volumes. We're quite accustomed to seeing AMD's chips lead in the multi-threaded rankings while trailing, sometimes by sizeable margins, in the single-threaded performance ranking. That isn't the case anymore, as Zen 3 easily leads both rankings.
Some lightly-threaded workloads, like Cinebench and LAME, respond so well to the Zen 3 architecture that it grants AMD a massive lead in the overall rankings. But as you'll see, while the Zen 3 chips excel tremendously at some applications, they also don't suffer from terrible performance in others as we see with Threadripper.
Cinebench has long been AMD's favorite benchmark for a simple reason; the Zen microarchitecture has always performed extremely well in the threaded benchmark. However, AMD has steadily improved its performance in the single-threaded benchmark as well, slowly working its way up the chart.
Apparently those days of small jumps are over, as the Ryzen 9 5950X notches a 20% lead over the 3900XT. That's an incredible gen-on-gen improvement in single-threaded performance. The 5950X also takes a 13% lead over the previous-gen 3950X in the multi-threaded test, which is equally impressive given that both chips have the same 142W power limitation.
Intel's chips take the lead in the single-threaded POV-Ray test, but the remainder of these tests favor the Zen 3 processors by significant margins.
Our encoding tests include benchmarks that respond best to single-threaded performance, like the quintessential examples LAME and FLAC, but the SVT-AV1 and SVT-HEVC tests represent a newer class of threaded encoders. Regardless of the type of encoder, though, AMD's Zen 3 chips impress.
A glance at the bottom of these charts is like a trip down memory lane — that's the traditional position for AMD's chips in web browser benchmarks. These benchmarks are almost exclusively lightly-threaded, so Intel has long held the top of the charts. This series of benchmarks makes a powerful statement about Zen 3's improved single-threaded performance.
Office and Productivity
If you're looking to build a screaming-fast computer, you're probably not doing it to run office applications like Word at breakneck speeds. However, these types of applications are ubiquitous the world over, so snappy performance is important for daily tasks. This is another area that AMD has long offered middling performance, but Zen 3 climbs the ranks in impressive fashion.
Compilation, Compression, AVX Performance
The LLVM compilation benchmark stresses the cores heavily, and here we see that the Ryzen 9 5950X doesn't offer much uplift over the previous-gen Ryzen 9 3950X. The same can be said about the Ryzen 9 5900X compared to its previous-gen counterpart, the Ryzen 9 3900X/T. These muted gains imply that the bottleneck lies elsewhere.
The threaded y-cruncher benchmark again shows limited scaling for the 5950X over the 3950X, and given the memory-heavy nature of this workload, we theorize this boils down to the same limitation on both chips — a dual-channel memory controller that restricts feeding the 16 hungry cores.
Workstation-Class Workload Test Notes
Some of these applications also make an appearance in our standard test suite, but those test configurations and benchmarks are focused on a typical desktop-class environment. In contrast, the following tests are configured to stress the systems with workstation-class workloads, which is a particular strength for the Ryzen 9 processors given their hefty core counts.
We loaded down our test platforms with 64GB of DDR4 memory spread across four modules to accommodate the expanded memory capacity required for several of these workstation-focused tasks. We also outfitted the test systems with PCIe 4.0 SSDs to factor in the platform-level advantage of AMD's support for the faster interface.
Puget Systems Adobe Benchmarks
Puget Systems is a boutique vendor that caters to professional users with custom-designed systems targeted at specific workloads. The company has developed a series of acclaimed benchmarks for Adobe software, which you can find here.
Adobe After Effects CC Render Node Benchmark
The After Effects render node benchmark leverages the in-built aerender application that splits the render engine across multiple threads to maximize CPU and GPU performance. This test is memory-intensive, so RAM capacity and throughput are important and can be a limiting factor.
The Ryzen 5000 chips take a step forward in this benchmark over the previous-gen counterparts, but Intel's processors are impressive in light of their lesser core counts. Much of this could boil down to having more available memory bandwidth per core.
Adobe Premier Pro CC Benchmark
This benchmark measures live playback and export performance with several codecs at 4K and 8K resolutions. It also incorporates 'Heavy GPU' and 'Heavy CPU' effects that stress the system beyond a typical workload. Storage throughput also heavily impacts the score. As such, it isn't surprising to see the Ryzen 5950X and 5900X outstrip the Intel processors in the overall score, but again, the 5950X's gain over the 3950X is relatively slim.
Adobe Photoshop CC Benchmark
The Photoshop benchmark measures performance in a diverse range of tasks, measuring the amount of time taken to complete general tasks and apply filters. This test leans heavily on GPU acceleration, and the Ryzen 5000 processors offer stellar performance in the GPU subtest. Here they outstrip their previous-gen counterparts by large margins and leave Intel's chips by the wayside, too.
SPECworkstation 3 Benchmarks
The SPECworkstation 3 benchmark suite is designed to measure workstation performance in professional applications. The full suite consists of more than 30 applications split among seven categories, but we've winnowed down the list to tests that largely focus specifically on CPU performance. We haven't submitted these benchmarks to the SPEC organization, so be aware these are not official benchmarks. We've upgraded to the new 3.0.4 revision that supports spanning the tests across processor groups and sockets, unlocking the utmost parallelism.
Product Development and Energy, NAMD
The earth’s subsurface structure can be determined via seismic processing. One of the four basic steps in this process is the Kirchhoff Migration, which is used to generate an image based on the available data using mathematical operations. The Ryzen 5000 series processors take the lead in this benchmark over the Intel comparables, but the 24-core Threadripper 3960X unsurprisingly takes a commanding lead due to its hefty core counts.
Flipping over to the Calculix workload tells a different story, though. This test is based on the finite element method for three-dimensional structural computations, and it typically responds well to higher core counts. However, as we've seen often with the Threadripper processors, they can offer overwhelming performance in some workloads, but suffer in others. Here we can see the Ryzen 5000 chips take an easy lead over the rest of the test pool, and it's exciting to think of how the Zen 3 architecture will perform in Threadripper processors.
NAMD is a parallel molecular dynamics code designed to scale well with additional compute resources and is one of the premier benchmarks used to quantify performance with simulation code. The Ryzen 9 5950X puts up a strong showing in this test, but again, we see relatively muted performance scaling over the previous-gen 3950X. However, we have to keep this in context: both chips have to adhere to the same 142W power limit, so the additional performance is impressive in its own right.
Other workloads, like the Fast Fourier Transforms, tell a much more impressive story for the Ryzen 5000 processors, though.
SPECworkstation 3's Rodinia LifeSciences benchmark steps through four tests that include medical imaging, particle movements in a 3D space, a thermal simulation, and image-enhancing programs. These workloads respond well to increased core counts, and as you might've guessed, that bodes well for Ryzen 5000. We even see the Zen 3 chips pull off a few wins against the Threadripper 3960X here, too.
Financial and General
The Python benchmark conducts a series of math operations, including numpy and scipy math libraries, with Python 3.6. This test also includes multithreaded matrix tests that obviously benefit from more cores. These benchmarks have long been a sore spot for AMD's processors, but the Ryzen 5000 chips rectify that issue.
The SPECviewperf 2020 benchmarks are hot off the press from the SPEC committee, so we decided to give it a spin with the Nvidia GeForce RTX 3090 to see how well the Ryzen 5000 processors can push along a GPU in professional rendering applications, which has long been a weakness of previous-gen Ryzen processors.
The Intel processors took the lead in many of the workloads, but it is important to note that AMD has shrunk the performance deltas tremendously.
Ryzen 9 5950X and 5900X Power Consumption, Thermals
Notably, AMD's decision to stick with the AM4 socket still constrains its maximum power consumption to 142W, which means that the company could not increase power consumption for the new flagship models. However, Zen 3's IPC gains allow the Ryzen 5000 chips to stay within the same TDP thermal and electrical ranges as the Ryzen 3000 series CPUs while delivering more performance.
But there's a bit of nuance to the power story, though. As we can see in the AIDA power measurements, both the 5900X and 5950X draw slightly more power under load than their previous-gen counterparts. However, flipping to the 'renders per day per watt' charts shows that the chips are considerably more power-efficient than the Ryzen 3000 processors, meaning they deliver considerably more performance per watt.
Intel's chips are rather inefficient in comparison, which is a natural byproduct of using the older and less-dense 14nm node. Intel has also turned the dial up on the voltage/frequency curve to remain competitive, which also throws efficiency out the window in exchange for higher performance.
The net-net is that the Ryzen 5000 processors will draw far less power per unit of work than any of Intel's 14nm chips, thus resulting in a cooler and quieter system.
Here, we take a slightly different look at power consumption by calculating the cumulative amount of energy required to perform an x264 and x265 HandBrake workload, respectively. We plot this 'task energy' value in Kilojoules on the left side of the chart.
These workloads are comprised of a fixed amount of work, so we can plot the task energy against the time required to finish the job (bottom axis), thus generating a really useful power chart. Bear in mind that faster compute times, and lower task energy requirements, are ideal.
This measure really separates the wheat from the chaff, and the best results fall to the lower left-hand corner of the chart. The Intel chips populate the less-desirable upper right-hand side. Although the Core i9-10980XE makes a valiant attempt to get down to Ryzen territory, it still can't match the previous-gen Ryzen 3000 processors in terms of efficiency. Meanwhile, the Ryzen 5000 series leverages the Zen 3 architecture to great effect and falls further inside the performance-per-watt sweet spot, marking a new level of efficiency for a modern desktop chip.
Taking the Silicon Throne
To the world of enthusiasts that have long been pining for a huge gen-on-gen upgrade, AMD’s Ryzen 9 5950X and Ryzen 9 5900X deliver an almost unbelievable amount of performance improvement over not only AMD’s previous-gen Ryzen processors, but also over Intel’s Comet Lake flagships. The fact that the Ryzen 9 chips regularly break the 5GHz barrier, even at stock settings, is simply icing on the cake.
AMD’s clever re-use of the proven Ryzen SoC design and I/O Die, not to mention the now-mature 7nm TSMC process, allowed the company to focus its resources on delivering a massively redesigned core architecture that takes a big step forward on IPC throughput, which in turn yields higher performance and more power efficiency.
AMD’s decision to unify the L3 cache pays big dividends in applications that profit from low-latency memory access, with gaming being the perfect example. Meanwhile, more nuanced improvements to the branch predictor and front end expose faster performance across the board, yielding big gains in both single- and multi-threaded workloads.
As we can see in our cumulative gaming and application measurements above, AMD has finally scored a clean sweep in 1080p gaming along with performance in single- and multi-threaded applications. That single-threaded ranking above isn't 'just' Cinebench, either — it's a cumulative measure of several different types of workloads. Perhaps most telling, the stock Ryzen 9 processors beat Intel's highly-overclocked flagships not only in gaming, but also in single-threaded performance.
We certainly couldn't have imagined this possibility when AMD launched the Ryzen series a mere three years ago. We test the fastest chips on the planet, and these types of massive generational performance increases are amazing, even to us.
AMD's path from the bottom of the performance charts to the top was a hard-fought win, but while the company now holds the performance crown, it has left Intel a sliver of room to operate as the budget alternative.
The Ryzen 5000 series processors land at significantly higher recommended price points than the previous-gen models, and you'll have to bring your own cooler. The price for entry on the low end is also higher than we're accustomed to, not to mention that you'll have to drop an extra $150 to move up from the six-core Ryzen 5 5600X to the eight-core Ryzen 7 3800X. At least 500-series motherboards are plentiful, and we now have B550 motherboards for budget platforms.
Zen 3’s gaming performance is nothing short of spectacular. However, as we've noted with previous AMD CPU reviews, many of those gains won’t be noticeable to users with lesser graphics cards. The tables have turned, and now Intel CPUs are the ones that are "basically just as fast as AMD" with anything short of the RTX 3080. On the other hand, AMD's upcoming Radeon RX 6800 XT could offer additional gaming benefits over Intel, thanks to Smart Memory Access.
Unfortunately, AMD’s suggested retail pricing rarely has any relation to reality at the checkout lane, so it’s hard to project where pricing will land in a few months. This much is certain, though: AMD will have no problem selling its pricey new silicon to enthusiasts looking for every last bit of performance on a modern platform.
For now, there’s no reason to recommend an Intel Comet Lake processor on the high end unless you need integrated graphics, so we’ll have to wait until Intel slashes pricing to reflect the reality that it is now the budget alternative. We have yet to test the Ryzen 7 and 5 models, and there could be at least some competition in the mid-range — but there is certainly no competition at the top.
Intel does have Rocket Lake waiting in the wings, but those chips won't land until next year and will top out at a mere eight cores. We don't foresee enough of a performance increase from Intel's new architecture etched onto the 14nm process to really tip the scales against AMD's core-heavy models, meaning AMD could reside at the top of the desktop PC game for at least a year, if not longer.
|Intel Socket 1200 (Z490)||Core i7-10700K, Core i9-10900K, 10850K|
|Gigabyte Aorus Z490 Master|
|2x 8GB Trident Z Royal DDR4-3600 - Stock: DDR4-2933, OC: DDR4-4000|
|AMD Socket AM4 (X570)||AMD Ryzen 9 5950X, 5900X, 3950X, 3900XT, 3900X, 2700X|
|MSI MEG X570 Godlike|
|2x 8GB Trident Z Royal DDR4-3600 - Stock: DDR4-3200, OC: DDR4-4000, DDR4-3600|
|Intel Socket 2066 (X299)||Core i9-10980XE|
|MSI Creator X299|
|4x 8GB Trident Z Royal DDR4-3600 - Stock: DDR4-2933, OC: DDR4-3600|
|AMD Socket SP3 (TR4)||Threadripper 3960X|
|MSI MEG X399 Creation|
|2x 8GB Trident Z Royal DDR4-3600 - Stock: DDR4-3200, OC: DDR4-3600|
|All Systems||Gigabyte GeForce RTX 3090 Eagle - Gaming and ProViz applications|
|Nvidia GeForce RTX 2080 Ti FE - Application tests|
|2TB Intel DC4510 SSD|
|EVGA Supernova 1600 T2, 1600W|
|Windows 10 Pro version 2004 (build 19041.450)|
|Workstation Tests - 4x 16GB Corsair Dominator - Corsair Force MP600|
|Cooling||Corsair H115i, Custom loop|
MORE: Best CPUs
MORE: All CPUs Content