AMD Ryzen 5 vs. Ryzen 7: Which Ryzen CPU provides the most bang for your buck?

Note: This article was first published on 14th May 2017.

Ryzen begins to take shape

AMD has a lot riding on its new Ryzen processors. After practically ceding the high-end CPU market to Intel for years, the company really needs Ryzen to succeed. So far, reviews have been generally positive, and save for slightly lackluster 1080p gaming performance, the consensus has been that the Ryzen 7 chips are quite attractive workstation parts for their prices.

Furthermore, AMD attributed the middling gaming performance to the lack of developer-side optimizations, and it says that we can expect a better showing once game developers have had the time to patch their games to take proper advantage of Ryzen’s architecture and high thread count.

So far, things are looking up, and early patches for Ashes of the Singularity: Escalation, Total War: Warhammer, and Dota 2 are quite promising. If things continue apace, Ryzen could see its most glaring weak spot become less of an issue.

Earlier this month, AMD further fleshed out its Ryzen line-up with the announcement of four mid-range Ryzen 5 processors. And like their Ryzen 7 bedfellows, these beat their equivalent Intel chips handily in terms of core and thread count.

The 6-core/12-thread Ryzen 1600X and 1600 were intended as even more affordable chips for gamers, Twitch streamers, and workstation users, while the 4-core/8-thread Ryzen 1500X and 1400 were focused more narrowly on gaming performance.

That said, with the vast majority of games today still benefiting more from higher core clocks than multiple cores, we’ll be looking closely at gaming performance in this review to see if a chip like the Ryzen 5 1500X (3.5GHz, 16MB L3 cache) could end up being a CPU bottleneck.

Here’s a table summarizing the key specifications of all the Ryzen chips announced so far (contrary to initial reports, XFR is actually enabled on all Ryzen chips, but -X SKUs have more headroom):

One interesting thing to note is that Ryzen 5 actually faces a very different playing field than Ryzen 7. The latter went up against Intel’s Haswell-E and Broadwell-E chips, both of which are based on an over two-year-old microarchitecture. Furthermore, AMD’s absence from the high-performance computing space meant that Intel was able to fatten its margins with impunity, and there are few that really think even the 10-core Intel Core i7-6950X (3.0GHz, 25MB L3 cache) is worth all of its US$1,650 price tag.

It also didn’t hurt that the audience for those chips are traditionally not too averse to shelling out top dollar for the best possible CPU performance, but mainstream consumers eyeing the mid-range Intel Core i5 and Ryzen 5 chips are a lot more sensitive to price.

In this segment of the market, Intel has had to price its chips more competitively, so AMD has a lot less wiggle room to undercut the competition on price while offering similar performance. Ryzen 5 is also going toe-to-toe with Intel’s latest Kaby Lake processors and microarchitecture, and clearly AMD’s strategy is to offer more cores and threads than the competition.

For example, the US$249 Ryzen 5 1600X (3.6GHz, 16MB L3 cache) has 6-cores/12-threads, whereas the US$217 Intel Core i7-7600K (3.8GHz, 6MB L3 cache) has just 4-cores/4-threads. That’s a major advantage AMD has in terms of thread count, although Intel still wins in terms of single-core performance.

In this article, we’ve pit the Ryzen 7 and select Ryzen 5 chips against each other to give you an idea of the performance of the entire stack, and included a performance-per-dollar analysis on exactly how much you’re getting for your money. The Ryzen 7 1800X (3.6GHz, 16MB L3 cache) may be the top dog here, but what if the Ryzen 5 1600X (3.6GHz, 16MB L3 cache) is really all you need?

Test Setup and Benchmarks

The configurations of the test setups we used are listed below. We’ve also included an Intel Core i7-7700K and Intel Core i7-6950X for comparisons. Unfortunately, we did not have a Intel Core i7-7600K on hand to compare against the Ryzen 5 1600X.

AMD Ryzen rig

• AMD Ryzen CPU with Noctua NH-U12S SE-AM4

• Gigabyte Aorus AX370 Gaming 5

• 2 x 8GB Corsair Vengeance LPX DDR4-3000 at 2,667MHz (Auto timings: CAS 16-16-16-36)

• ASUS ROG Strix GeForce GTX 1080 Gaming (GeForce Driver Version 378.66)

• Kingston HyperX 3K 240GB SATA 6Gbps solid state drive (one single NTFS partition)

• Windows 10 Home (64-bit)

Intel Kaby Lake rig

• Intel Core i7-7700K (4.20GHz, 8MB L3 cache) with Cooler Master MasterAir Maker 8

• ASUS ROG Maximus IX Formula

• 2 x 8GB G.Skill Ripjaws V DDR4-3000 at 2,400MHz (Auto timings: CAS 15-15-15-35)

• ASUS ROG Strix GeForce GTX 1080 Gaming (GeForce Driver Version 378.66)

• Kingston HyperX 3K 240GB SATA 6Gbps solid state drive (one single NTFS partition)

• Windows 10 Home (64-bit)

Intel Broadwell-E rig

• Intel Core i7-6950X (3.0GHz, 25MB L3 cache)

• ASUS ROG Strix X99 Gaming

• 4 x 4GB Corsair Vengeance LPX DDR4-2133 (Auto timings: CAS 15-15-15-36)

• ASUS ROG Strix GeForce GTX 1080 Gaming (GeForce Driver Version 378.66)

• Samsung SSD 840 Pro 256GB SATA 6Gbps solid state drive (OS + benchmark + games)

• Windows 10 Pro (64-bit)

Coming from our review of the Ryzen 7 1800X, we’ve included a new game benchmark in Ashes of the Singularity: Escalation as it was just updated with optimizations for Ryzen. Here’s a list of all the benchmarks used:

• SYSmark 2014 ver 1.5

• SPECviewperf 12.1

• Cinebench R15

• Handbrake 1.0.2

• 3DMark (2013)

• Ashes of the Singularity

• Ashes of the Singularity: Escalation

• Deus Ex: Mankind Divided

All the benchmarks were carried out in Windows High Performance mode and with the High Precision Event Timer (HPET) disabled in BIOS.

	AMD Ryzen 7 1800X	AMD Ryzen 7 1700X	AMD Ryzen 7 1700	AMD Ryzen 5 1600X	AMD Ryzen 5 1500X	Intel Core i7-7700K Processor	Intel Core i7-6950X
From S$818	From S$599	From S$499	From S$359	From S$289	From S$543	—
From S$818	—	—	—	From S$289	From S$510	From S$2488
AMD Ryzen 7 1800X	AMD Ryzen 7 1700X	AMD Ryzen 7 1700	AMD Ryzen 5	Ryzen 5	Intel Core i7-7700K	Intel Core i7-6950X
Ryzen 7 1800X	Ryzen 7 1700X	Ryzen 7 1700	Ryzen 5 1600X	1500X	Core i7-7700K	Core i7-6950X
3.6GHz	3.6GHz	3.0GHz	3.6GHz	3.5GHz	4.2GHz	3.0GHz
4.1GHz (with XFR technology)	3.8GHz (3.9GHz with XFR)	3.7GHz	4.0GHz (4.1GHz with XFR)	3.7GHz (3.9GHz with XFR)	4.5GHZ	3.5GHZ
8	8	8	6	4 Core / 8 Threads	4	10
100MHz	100MHz	100MHz	100MHz	100MHz	100	100MHz
4 x 256KB	4MB	4MB	3MB	2MB	4 x 256KB	2.5MB
16MB	16MB	16MB	16MB	16MB	8MB	25MB
Integrated Dual Channel (up to DDR4-2667)	Integrated Dual Channel (up to DDR4-2667)	Integrated Dual Channel (up to DDR4-2667)	Integrated Dual Channel (up to DDR4-2667)	Integrated Dual Channel (up to DDR4-2667)	Integrated Dual Channel (up to DDR4-2133 or DDR4-2400)	Integrated Quad Channel (up to DDR4-2133 or DDR4-2400)
PCIe 3.0	PCIe 3.0	PCIe 3.0	PCIe 3.0	PCIe 3.0	16 lanes of PCIe 3.0	PCIe 3.0
95	95	65	95	65	91	140
SSE 4.1/4.2, AVX 2.0	SSE 4.1/4.2, AVX 2.0	SSE 4.1/4.2, AVX 2.0	SSE 4.1/4.2, AVX 2.0	SSE 4.1/4.2, AVX 2.0	SSE 4.1/4.2, AVX 2.0	SSE 4.1/4.2, AVX 2.0
Yes	Yes	Yes	Yes	Yes	Yes	Yes
Socket AM4	Socket AM4	Socket AM4	Socket AM4	—	LGA1151	LGA2011-3
14nm	14nm	14nm	14nm FinFET	14nm FinFET	14nm	14nm
—	768KB	768KB	576KB	384KB	—	640KB
—	—	—	—	—	DMI 3.0	DMI 2.0
—	—	—	—	—	100MHz	100MHz
—	—	—	—	—	Yes	Yes
—	—	—	—	—	Yes (VT-x)	Yes (VT-x)
—	—	—	—	—	Kaby Lake	Broadwell-E

Performance results

SYSmark 2014 ver 1.5

SYSmark is a business productivity benchmark suite that measures the response times of tasks on a PC using real-world applications in the areas of office productivity, media creation, and data and financial analysis. Task response times from simulated user input are used to generate a performance rating that reflects actual user experience, so the faster a PC responds to application workloads, the higher its score will be. The method of measuring response times can take many forms, such as the time it takes for an application to show a pop-up completion message, or how long it takes a progress dialog to disappear and for a user to regain application control.

The applications in SYSmark benefit to a certain extent from multiple cores (the Ryzen 7 1800X and Ryzen 5 1600X are clocked the same, but the 8-core 1800X is overall quicker), but it looks like clock speed wins out in the end, with the Core i7-7700K high starting base clock helping it inch ahead. On the other hand, the 4-core/8-thread Ryzen 5 1500X had the weakest performance because of a combination of less aggressive speeds and lower core/thread count, coming in at 18 per cent slower than the 1600X.

A look at the score breakdowns also shows that the Ryzen 7 chips excel in data/financial analysis-type tasks, which involves crunching, compressing and decompressing data to make forecasts and projections. Here, they were only behind the 10-core Intel Core i7-6950X.

SPECviewperf 12.1

SPECviewperf is used to measure the 3D graphics performance of systems in professional applications. Each individual workload, called a viewset, represents graphics and content from an actual real-world application. SPECviewperf actually runs a total of eight different viewsets, but we’ve picked the four which have the greatest performance variation across CPUs displayed here.

The 3ds-max viewset comes from traces of the graphics workload generated by 3ds Max 2016, while maya-04 is derived from Autodesk’s Maya 2013 application. The catia-04 viewset involves the numerous rendering modes from the CATIA V6 R2012 application, and includes things like anti-aliasing, depth of field, and ambient occlusion. Finally, the sw-03 viewset comes from SolidWorks 2013 SP1, and involves various rendering modes including environment maps.

All the Ryzen CPUs were quite closely matched here, and it’s clear that these workloads aren’t quite suited to taking advantage of high thread counts. For example, the 4-core/8-thread Ryzen 5 1500X was shown to edge out the 8-core/16-thread Ryzen 7 1700X and 1700, likely by virtue of a slightly higher base clock.

Cinebench R15

Cinebench R15 is a better indicator of the Ryzen’s 7 performance because of its ability to utilize up to 256 threads to evaluate a processor’s performance in a photorealistic 3D rendering. We ran both single-core and multi-threaded benchmarks to evaluate single-threaded performance and multi-threaded scalability here.

As we saw in our Ryzen 7 review, the multi-threaded benchmark is where Ryzen really steps up. The 6-core/12-thread Ryzen 7 1600X beat the Intel Core i7-7700K by a good 28 per cent, and the latter is the flagship of Intel’s mainstream CPUs line-up. That said, the extra two cores and four threads on the 1800X meant that it was still 31 per cent quicker than the 1600X

Interestingly enough, the Ryzen 5 1600X had a narrow advantage over the Ryzen 7 1800X in the single-core benchmark. Both chips have the same clock and boost speeds, but the results suggest that the 1600X was operating at the upper end of its boosted frequencies more often.

Handbrake 1.0.2

Handbrake is a video transcoder that converts videos into a format for use on PCs and portable electronic devices, and is a good indicator of a processor’s video encoding capabilities. YouTube content creators, Twitch streamers, and other video creators will be most interested in this performance metric.

The Ryzen 5 1600X did well here and performed closer to the Ryzen 7 chips with relatively quick encoding times. However, the Ryzen 5 1500X was quite a bit slower here, which is probably why AMD isn’t recommending this particular CPU for streamers as it is with the 1600X.

Streaming is very CPU-intensive, and having more cores and threads help gamers take less of a performance hit while broadcasting.

Power consumption

To test power, we ran the energy-01 viewset in SPECviewperf 12.1 and recorded the peak power consumption.

The most notable difference here would be from the lower 65 watt TDP of the Ryzen 5 chips, which translated into a peak power consumption that was around 30 watts lower than the 95 watt Ryzen 7 parts.

Gaming benchmarks and overclocking

3DMark (2013)

The synthetic 3DMark benchmark tests graphics and computational performance at different resolutions, starting at 1080p and going all the way up to 4K.

To further help tease out differences in CPU performance, we’ve included a separate table with the Physics scores in Fire Strike for the Ryzen CPUs. The physics test favors having more cores, and a good example of that would be the comparison between the Ryzen 7 1800X and Ryzen 5 1600X.

While the 1600X has the higher overall score, the 1800X was around 15 per cent faster on the physics test, showing that the former’s lead was from variations on the GPU front.

Another thing you’ll notice going forward is that the more GPU-intensive (which means the CPU is no longer the limiting factor) the benchmark gets, for example because of higher resolutions or settings, the smaller the differences between AMD and Intel gets. Having said that, there were hardly any significant differences in the Fire Strike Ultra test.

Similarly, because the Time Spy DirectX 12 benchmark is also more CPU intensive, all three Ryzen 7 chips managed to beat the Core i7-7700K here. A low-level API like DirectX 12 works best with weaker single-core performance and it improves multi-threaded scaling, which is probably why the Ryzen 7 chips did better here.

Ashes of the Singularity

The Ryzen 7 1800X failed to impress at 1080p gaming, and the same goes for the other Ryzen processors. In fact, all the Ryzen chips ended up within a few frames of each other in The two Intel CPUs were having the time of the day here, waltzing to large leads over their Ryzen counterparts.

Furthermore, the minuscule differences between performance at 1080p and 4K in DirectX 11 show that the Ryzen CPUs are the limiting factor here. In comparison, the Intel CPUs exhibited far more marked performance variations when scaling between resolutions.

Things improve slightly in DirectX 12, although it’s apparent that the Ryzen CPUs are still holding things back. When switching over to Crazy settings, which places a greater part of the performance burden on the GPU rather than the CPU, you start to see more dramatic differences between the different resolutions.

In addition, we've included the CPU-focused DirectX 12 benchmark, which attempts to more clearly illustrate CPU performance by simulating an infinitely fast GPU.

Ashes of the Singularity: Escalation

Here’s where things start to get interesting though. While Ashes of the Singularity was a study in frustration because of the failure of Ryzen to scale properly, the Escalation expansion professes to make things better with a patch that has been optimized for Ryzen. Starting from version 26118, you should see fairly significant improvements because the game is now better able to utilize Ryzen’s unique hardware.

AMD’s Zen architecture has some fundamental differences from Intel’s x86 implementation, and developers will need time to properly adapt their games to it.

After the patch, the Ryzen 7 1800X boasted a 22 per cent improvement at 1080p (High, 8x AA, DX12), while the Ryzen 5 1600X had a more modest 14 per cent increment. Given that Ryzen’s particular strengths include a high number of threads, the patch may be skewed to take advantage of that, which is why the Ryzen 5 chips had less significant frame rate increases.

More importantly, the update helped narrow the gap with the Intel Core i7-7700K across the board. The Intel chip was initially ahead of the Ryzen 5 1600X by a whopping 37 per cent at 1080p ((High, 8x AA, DX12), but that shrank to a 18 per cent lead in Escalation. The improvement was more dramatic for the 1800X, with the Core i7-7700K’s lead decreasing from 35 per cent to a mere 9 per cent at the same settings.

Finally, it also looks like the patch managed to mitigate some of the CPU bottleneck for Ryzen, and there are now larger differences when switching between resolutions.

Deus ex: Mankind Divided

Deus Ex: Mankind Divided is one of the most demanding games out there right now, so while Ashes of the Singularity taxed the CPU quite a bit, Mankind Divided often puts the system in a position where the GPU is the limiting factor.

The result is remarkably similar performance across the board at 1440p and 2160p resolutions, and basically all of the tested resolutions at Crazy settings.

At 1080p and High presets (DirectX 11) however, the Ryzen processors compared quite favorably against their Intel counterparts. Unfortunately, that changed when moving over to DirectX 12, as the Intel CPUs exhibited improved performance while AMD’s chips swung the other way.

That could be a result of a lack of optimization on the part of the developer, and it remains to be seen whether new patches will roll out for this game (and others) that will address similar issues.

Overclocking

We used AMD’s RyzenMaster utility to overclock the Ryzen chips. The software recently received an update that doesn’t require you to manually enable HPET for it to work properly, which made things a bit more convenient.

This table below summarizes the clock speeds, the voltages we used to achieve them, and the respective performance gains. AMD generally recommends 1.4V as a safe voltage (going higher could affect longevity) that should be sufficient for most overclocks, but for the sake of testing we decided to try to push the chips over that.

Overclocking also disables the two-core XFR boost, so clock speeds will not ramp up beyond the current clock speed.

As it turns out, the Ryzen 7 1700 turned in the most impressive performance improvement at 22 per cent. It’s not difficult to see why – it starts off with relatively low stock speeds, but it is still an 8-core/16-thread part, and overclocking it close to 4.0GHz brings it close to the Ryzen 7 1800X and 1700X.

How value-for-money is it?

It’s all well and good to talk about raw performance, but one metric that’s less clear is the price-to-performance ratio, which really just refers to how much bang you’re getting for your buck.

To suss out how worth your while each of the Ryzen chips really is, we’ve generated an aggregate performance composite index for each processor, comprising a single representative score from every benchmark we ran. The Ryzen processors do very well for tasks that properly leverage its heavily multi-threaded architecture, but are less impressive in gaming benchmarks, so this approach aims to take all these performance aspects into account.

We used the Intel Core i7-7700K as our baseline, which means it had an index score of 1.00. All the other scores were calculated relative to it, so a score of 1.15 for the Core i7-6950X means that it is 15 per cent quicker than the Core i7-7700K. Conversely, a score of 0.90 for the AMD Ryzen 7 1700 indicates overall performance that is 10 per cent slower.

That said, this comparison assumes that you’re more interested in overall performance as opposed to a single aspect like gaming or some other use case. If you’re interested in a particular usage model only, we’d refer you back to our individual benchmarks for a better gauge of performance.

In addition, we’ve included a separate price-performance index where we divided the performance composite index by the respective prices of each processor to get a rough indicator of how much performance you’re getting per dollar. Higher values indicate more performance per dollar spent, and we’ve arranged the graph below from the highest performance composite index to the lowest. We’ve also included a table with the local prices for each of our tested processors for easier reference.

Ultimately, the Ryzen 5 1600X and 1500X had the two highest price-performance indexes thanks to their relatively competitive performance and attractive prices. The Ryzen 5 1600X stands out in particular because of how close its performance is to the Ryzen 7 1800X, which is more than twice its price.

If multi-threaded performance is not a top priority, the Ryzen 5 1600X is looking to be a very attractive offering. It is also shaping up to be an excellent alternative to Intel’s mid-range processors, especially if you’re looking for a decent all-round performer.

As we mentioned earlier, we didn’t have the Core i7-7600K for testing, but that is a quad-core part with no hyperthreading that costs around S$380, more than even the 6-core/12-thread Ryzen 5 1600X. Having said that, if you're going to be dealing with heavily threaded workloads in addition to gaming, it is difficult to say no to the latter.

Finally, while we conducted our benchmarks with AMD's flagship X370 chipset, the Ryzen 5 chips are more likely to be used with cheaper B350-based boards that generally cost under S$200. This brings the overall cost of a potential Ryzen 5 system down, further bolstering the competitiveness of the series in terms of price.

Sometimes you don’t have to be the absolute best to be a winner

In a nutshell, Ryzen 5’s main advantage over Intel’s Core i5 range is the higher number of cores and threads it offers while sharing the same price bracket. That’s not unlike what AMD offers with its Ryzen 7 CPUs, all of which are 8-core/16-thread parts, but AMD has been able to bolster its position in that segment with significantly lower prices as well.

That said, we still stand by what we said in our review of the Ryzen 7 1800X. If your focus is solely on gaming, Intel is still the better option because of its higher instructions-per-cycle and clock speeds. It’ll also take some time before enough developers and games optimize for Ryzen’s approach of lower clock speeds and more cores, so 1080p gaming performance (where the CPU is often limiting) is going to be wanting for a while yet.

However, recent developments suggest that the industry as a whole is moving toward better leverage of a higher number of cores and threads, and there are even rumors that Intel will introduce a 6-core CPU as part of its mainstream line-up for its 8th-generation Coffee Lake processors.

And if you’re going to be doing a lot of CPU-intensive work, the Ryzen CPUs will do wonderfully well. The Ryzen 5 1600X is particularly attractive because it is difficult to turn down 6 cores and 12 threads at that price when the closest Intel offering has just 4 cores and 4 threads.

Furthermore, we should point out that Ryzen isn’t as bad at gaming as the headlines may make it sound. It isn’t as quick as the Intel Core i7-7700K for sure, but that doesn’t mean it is terrible. Maybe “good enough” is enough for most folks, especially considering that Ryzen makes up for it with major advantages in CPU-intensive work.

The Ryzen 7 1700 also deserves particular praise because of the performance that overclocking unlocks. It has 8 cores and 16 threads like the 1800X and is initially a bit handicapped by its relatively low 3.0GHz base clock, but we managed to overclock it to close to 4.0GHz, after which it outperformed the stock Cinebench score of the 1800X.

Arguably, the Ryzen 7 1700X and 1700 are both better buys than the 1800X because they offer (or can offer) very similar performance at considerably lower prices. A similar case could be made for the Ryzen 5 1600X, which has the same clock speeds as the 1800X but fewer cores. The 1500X lags behind more, but is also priced very affordably.

For how close they are to each other, just look at their performance composite indexes (we’ve inserted the table here for easier reference):

In summary here's our recommendation:-

• AMD Ryzen 5 1600X - For excellent value for money and good all-round performance. 90% performance of a Core i7-7700K at 35% lower cost is a very powerful statement.

• AMD Ryzen 7 1700 - If you would like to have Ryzen 7 1800X performance and don't mind overclocking to achieve it.

• AMD Ryzen 7 1700X - If you want Ryzen 7 1800X performance at a much lower price point (without overclocking).

All things considered, maybe Ryzen doesn’t have to be the best to be a winner. The line-up doesn’t have the best multi-threaded performance, nor does it take the crown in gaming. But the 10-core Intel Core i7-6950X costs a monumental sum, and the Core i7-7700K doesn't do as well in heavily-threaded workloads. You could say it's a matter of perspective and trading certain things for others – in Ryzen's case, this simply means you gain in terms of better prices and higher core counts, but lose some (but not too much) gaming performance.

	Click here for rundown of pros/cons
	Click here for rundown of pros/cons
	Click here for rundown of pros/cons
	Click here for rundown of pros/cons
	Click here for rundown of pros/cons