AMD Ryzen R7 1800X vs. 1700, 1700X Specs
|Ryzen R7 1700||Ryzen R7 1700X||Ryzen R7 1800X||Ryzen R5 1600X||Ryzen R5 1500X|
|Base / Boost||3.0GHz / 3.7GHz||3.4GHz / 3.8GHz||3.6GHz / 4.0GHz||3.6GHz / 4.0GHz||3.5GHz / 3.7GHz|
|Stock Cooler||Wraith Stealth (65W)
|Wraith Spire (95W)
|Wraith MAX (125W)
|Memory Support||2ch/2rank: 1866-2133
|Release Date||March 2||3/2/17||3/2/17||2Q17||2Q17|
The Ryzen R7 1700 is an 8C/16T CPU with the exact same CCX layout as the 1800X and 1700X. In terms of architecture and cache, you’re looking at the same thing; for the unfamiliar, you can read (P1) or watch our original reviews for a straight-shooter recap of the architecture. Nothing has changed in this regard.
The R7 1700, then, is different primarily insofar as its clock-rate. The TDP is lower, sure, but it’s all tied to the frequency and voltage of the chip. Lowered overall voltage also impacts thermals, but we’ve still got thermal, frequency, and voltage headroom to apply potentially significant overclocks.
AMD’s R7 1700 ships with a stock frequency of 3.0GHz, with a boost of 3.7GHz. The CPU we received did not include the Wraith cooler, though we’ve purchased our own 1700 and are awaiting its arrival so that we can independently test the cooling solution. On the topic of cooling, the 1700 requires a much less powerful cooling option than the 1800X, largely thanks to lower voltage and frequency on the 1700. TDP of the 1700 is 65W, with the 1800X and 1700X both at 95W, but with higher native frequencies.
Memory support is the same on paper, just with the caveat that we had greater memory clocking limitations on the 1700 than the 1800X. We were only able to hit 2666MHz out of box, even using the same kit as in the 1800X review, which indicates some sort of IMC limitation on the R7 1700. In our later overclocking tests, we were able to jump back to 2933MHz without issue, further validating that this is likely a stability and clock issue at the lower CPU frequencies.
We focused on overclocking and SMT enabled testing for the 1700 for now, but have disabled SMT and included it on the charts. This generally shows improvement, and is something we’d recommend for gaming-only users until the relevant Windows updates push).
We’re still testing on the ASUS Crosshair motherboard, still with the latest stable EFI release – version 5704. We’ve begun testing on Gigabyte boards and have largely seen no difference in performance (in February BIOS revisions), but will publish those numbers in a separate content piece. We’ve heard there are a few more updates worth testing first.
CPU Test Methodology
Our usual testing methodology is defined further below, per typical copy/paste of methods between reviews, but we have a few explicit items to address with Ryzen.
During testing, GamersNexus discovered that Windows would occasionally engage core parking or other power saving features that could impact framerate in some games, specifically with Ryzen. This was not always detectable, but some games reacted more than others (our biggest observable difference was ~4-5%). As such, all tests were conducted in performance mode. We further discovered that SMT toggling improves performance in nearly all gaming use cases, and so tested with SMT both enabled and disabled. When asked about this, AMD supplied that it is likely on the ISV optimization side.
We’ve been using the newest (and correct) version of AMD’s chipset drivers since we began testing, and we’ve been on ASUS’ version 5704 (latest) BIOS for the Crosshair Hero VI. ASUS was good to provide us this update immediately when we reached out, though AMD officially distributed an older EFI version to reviewers; this older version is detrimental performance in some ways, as it was not the production-ready EFI. In speaking with other editors, MSI boards went through similar EFI updates that drastically changed performance, in some cases. A few toggles are broken in various EFIs we've looked at here, which can potentially skew results if the tester does not validate that the toggles work as advertised (performance modes, for instance, can toggle threads in some extreme cases). It is relatively easy to avoid unintentional corruption of test data so long as the tester rigorously validates thread count, clock-rate, and voltages prior to each major test sequence.
Fortunately, we had already retrieved the correct EFI from ASUS directly, and so did not have to test multiple EFI versions – though we did have to retest the platform and CPU multiple times for other validation. EFI version matters a lot in these very early days of Ryzen. This is an entirely new architecture – not some updated on a long-known design, like KBL was – and vendors were not prepared. Early EFI distributions were based on older uCode. Results will vary between test environments in greater ways than usual.
This part applies to all motherboards, CPUs, and platforms: For testing, we disable any performance bias options made available through EFI. ASUS often offers Cinebench or 3DMark performance bias configurations (boosting performance in specific applications). We disable these for fairness across platforms. All stock CPU options are left enabled for benchmarking, like Boost on Intel and AMD.
A clean image is used for all testing. We have Windows images on multiple SSDs, each labeled for the appropriate CPU architecture. We pull an SSD with the correct label from the shelf (e.g. Skylake, Broadwell-E, Ryzen) and use that for testing. Note also that Windows HPET can affect performance.
Because of how Ryzen’s boosting reacts to temperatures, we put the R7 1800X under the same liquid cooling we’ve used for all our major CPU reviews: the Kraken X62. This eliminates unpredictable thermal throttles during test.
Game Test Methodology
NVIDIA 376.33 drivers were used for all benchmarking. Game settings were manually controlled for the DUT. All games were run at presets defined in their respective charts. All other game settings are defined in respective game benchmarks, which we publish separately from GPU and CPU reviews. Our test courses, in the event manual testing is executed, are also uploaded within that content. This allows others to replicate our results by studying our bench courses.
Windows 10-64 was used for testing.
Each game was tested for 30 seconds in an identical scenario, then repeated three times for parity.
Average FPS, 1% low, and 0.1% low times are measured. We do not measure maximum or minimum FPS results as we consider these numbers to be pure outliers. Instead, we take an average of the lowest 1% of results (1% low) to show real-world, noticeable dips; we then take an average of the lowest 0.1% of results for severe spikes.
Core Components (Unchanging)
- NZXT 1200W Hale90v2
- For DDR4 platforms: Corsair Vengeance LPX 32GB 3200MHz*
- For Ryzen DDR4: Corsair Vengeance LPX 3000MHz clocked to 2933MHz (See Page 2)
- Premiere & Blender tests do not exceed 8GB DRAM. Capacity is a non-issue for our testing, so long as it is >16GB
- For DDR3 platforms: HyperX Savage 32GB 2400MHz
- Intel 730 480GB SSD
- Open Air Test Bench
- Cooler #1 (Air): Be Quiet! Dark Rock 3
- Cooler #2 (Cheap liquid): Asetek 570LC w/ Gentle Typhoon fan
- Cooler #3 (High-end): Kraken X62
Note: fan and pump settings are configured on a per-test basis.
Used for R7 1800X, R7 1700X, R7 1700.
- Gigabyte Aorus Gaming 7 (primary)
- MSI Gaming Pro Carbon (secondary - for thermal validation)
- i7-7700K (x2) samples from motherboard vendors
Both used for the 7700K.
- MSI Gaming M7
- i7-6700K retail
- Gigabyte Z97X G1 WIFI-BK
- MSI GD65 Z77
Dx12 games are benchmarked using PresentMon onPresent, with further data analysis from GN-made tools.
Note: We'd like to add the i5, i3, and FX CPUs, but this was enough for now. We'll add those as we expand into coverage of Zen or i5 Kaby Lake products.
Thermal Test Methodology
Thermal measurement on Ryzen is not necessarily trivial, as most software is incorrect or inaccurate in these early days. See this page for more information.
Power testing is simply done at the wall. We do not presently tap into the rails, and openly identify this as our weakest point in current test methodology. This is something we will eventually work toward revamping. For today, we use wall meters to determine a power delta in A/B tests.
Temperature Over Time – R7 1700
Starting first with temperature over time, we see temperatures drastically reduced versus the R7 1800X, with both CPUs using our Kraken X62 at max fan and pump RPMs. The R7 1700 sustains a completely flat clock-rate under load, where the 1800X had some fluctuation during torture testing. This is largely due to heat, and seeing as the 1700 constantly sits below 50C with this cooler – generally closer to 45C – it’s easy to see why clock fluctuations aren’t happening on the R7 1700 SKU CPU. Far easier to keep this thing thermally controlled.
We are still waiting for our stock cooler to arrive from a purchase and will validate its performance at that time. For now, we’re just using the X62 for a better comparison to our previous results. Those are here, by the way:
HWMonitor TMPIN0 seems to match ASUS’ AI Suite, more or less. ASUS’ AI Suite takes the AMD TSI temperature bus, mixes it with a thermistor, and performs a calculation to bring the number into reality. Raw numbers from some software solutions are not necessarily accurate; in earlier versions of AIDA64, the temperature numbers were off by a factor of ~20-30C, depending on SKU and AIDA version. The best way to validate temperature performance is to use multiple tools, then speak with the vendors of those tools to determine how the temperature is drawn (as we’ve done). HWMonitor seems best right now, though we cross-reference to AI Suite and thermocouples mounted to the IHS (this is not a die reading, mind you).
AMD R7 1700 Auto Voltage Behavior
|1700 Auto vCore 2666MHz DRAM||Blender (nT)||Cinebench (1T)||POV-Ray (1T)||Idle|
|Total System Draw (Watts)||133||80||81||58|
|Core Frequency (GHz)||3.2||3.2-3.75||3.75|
|Core Voltage (v)||1.068||1.286-1.308||1.286|
|1800X Auto vCore, 2933MHz RAM||Blender (nT)||Cinebench (1T)||POV-Ray (1T)||Idle|
|Total System Draw (Watts)||185||92||94||60|
|Core Frequency (GHz)||3.7||3.7 - 4.1||4.1|
|Core Voltage (v)||1.221 - 1.243||1.395 - 1.461||1.417|
Looking now at total system power draw and voltage stats, we observed the R7 1700 operating at 133W peak draw when under stock settings and with a Blender workload. By way of comparison, the R7 1800X (also stock) looks at 185W total system power draw in the same workload. That said, note that our 1800X was capable of pushing 2933MHz memory at stock settings, where we were limited to 2666 on the R7 1700 until a later overclock. This influences results, and we are making that clear. Our tests have never strongly focused on power (at the moment), so we’re doing cursory looks / studies on performance in specific use cases. The 1700 pushes around 3.2GHz when left to auto in Blender, with a vCore of 1.068 volts.
Cinebench 1T, we see power draw at around 80W, compared to total system draw of 92W on the 1800X. This posts the 1700 CPU speed bouncing between 3200 and 3750MHz. This is because, again, Boost and XFR on the AMD CPUs only kicks in when running limited thread workloads. It does not enable for workloads leveraging all threads. Voltage is around 1.286 to 1.308 vCore for this test.
POV-Ray puts us at 3750MHz clock-rate for a single-threaded run, with voltage stable at 1.286. This is lower than the 1800X, contributing to lower temperatures overall.
All of these power tests are run in balanced mode, with all the gaming and production tests executed in performance mode. We have used performance mode since day-one for the performance tests, given some core parking Windows issues with this disabled. This drives power usage up in idle states and does force a higher constant clock (unnecessarily), so keep that in mind when considering your own use cases at home.
Blender Benchmark R7 1700 vs. R7 1800X, i7-7700K
Our Blender benchmark uses an in-house render scene that renders at 400 samples per pixel with a 4K resolution, using 16x16 tile sizes for optimized CPU rendering. This scene is used across all CPUs we test, and represents a real-world use case as created by GN’s animator.
As shown here, the R7 1700 stock, SMT enabled, completes the scene in about 33 minutes. This is between the 7700K ($350) overclocked and the R7 1800X ($500) or 6900K (~$1000) stock CPUs. That’s about where we’d expect the 1700 ($330) to fall, given the 1800X’s performance, but what’s interesting is the overclocking aspect: Assuming you’re willing to put in the five minutes to OC and run a little higher temperature, the 1700 is able to complete the scene in the same time as the overclocked 1800X, as both land at around 3.9 to 4.0GHz in our production tests. We could sustain a 4.0GHz clock in gaming, but production workloads stressed stability and forced us to drop to 3.9GHz.
If you’re willing to overclock and aren’t concerned about IT approval, the 1700 beats the stock 1800X, and ties the overclocked variant (within reasonable render variance). Note that our OC on the 1700 bumps the memory to 2933, since we could only support 2666 stable when no OC was applied.
This more or less invalidates the 1800X in this particular test, if considering it strictly for CPU-accelerated Blender tasks. Although a GPU will still get you further faster in almost all Blender workloads, CPU acceleration is certainly competitive here.
Just like we said in the 1800X review, AMD deserves praise for its production positioning versus Intel’s same-cost (or double-cost) alternatives. Also just like we said in the 1800X review, we’d recommend the 1800X over the 6900K for CPU-targeted production – and again, that’s nearly a direct quote from the last content – but we’d now recommend the 1700 over the 1800X. For our audience – we’re assuming either pure gaming or gaming + some YT or art creation – it’s an even better value for CPU-accelerated production, if you do any. That said, the R7 1700 does need an overclock to kick it into gear. That overclock is trivial, and the cooling requirement isn’t all that high, just know that it’s well worth it with this particular chip. AMD competes with its own product (1800X) when overclocking the 1700. Otherwise, for the most part, Blender and Premiere are happening on the GPU – though other workloads like compiles, should you work with them, would benefit from CPU acceleration.
Adobe Premiere Benchmark – R7 1700 vs. R7 1800X, 7700K
Our Adobe Premiere benchmark is conducted by rendering our EVGA ICX review, representing a real-world workload filled with color corrections, warp stabilizes, and transforms.
In this benchmark, the 6900K leads when overclocked, with the 1800X leading the 6900K marginally when both are stock. Now, as stated last time, again, the marginal lead is made less marginal when considering the half-cost part.
If we look at the R7 1700, we’re performing ahead of the 7700K stock by about 30 minutes, which is a considerable gain when considering near price equivalence (slightly favoring AMD). This is finally starting to look like the Ryzen processor that should have championed the launch, despite initial shipment and brandishing of the 1800X.
Overclocked, the 1700 again outperforms the 1800X, or equals the overclocked 1800X with equal memory speeds. Given the $170 price reduction from the 1800X, we see no reason for our audience or mixed workloads users (on these applications) to buy the 1800X. The 1700 is far and away a better deal. Now, again, Premiere is CUDA-accelerated, so you’d generally want to do these workloads on a GPU. That said, the price positioning is much stronger versus adjacent Intel competition, and finally starts to build a case for Ryzen in our reviews. Users who do CPU-accelerated workloads should of course check benchmarks for the relevant software, but could generally postulate that the 1700 would outperform an i7-7700K stock in software accelerated tasks. But again, check reviews for the actual software you use – it doesn’t all behave the same way with regard to multithreading or frequency responsiveness.
Cinebench – R7 1700 vs. R7 1800X, 7700K, 7600K, etc.
Cinebench served as AMD’s most heavily touted metric for performance this generation, representing a synthetic version of our rendering tests above. With Cinebench – effectively an nt or 1T render simulation – we see the R7 1700 position itself between the 7700K OC and the 6900K (stock, HT1). That’s impressive, though the above real-world render benchmarks should serve as more useful representations of reality. This does generally align with what we saw up there, though.
With regard to 1T testing, the 1700 is markedly behind the 7700K stock or overclocked, and is also behind the 7600K, 6600K, and overclocked 2500K.
The CPU isn’t too far behind an 1800X, and easily catches up with an overclock.
3DMark Firestrike – R7 1700 vs. R7 1800X, 7700K, 7600K, Overclocking
FireStrike certainly isn’t perfectly repeatable, but it’s close enough each time (and is synthetic, which helps). There tends to be some score variance from one pass to the next.
We’re looking at a placement of the 1700 overclocked nearly chart-topping, putting to shame its $170 and $770 more expensive Top 3. We’re sorting by physics score here, note, and that’s the CPU bound task.
For physics FPS, the numbers look like this:
3DMark TimeSpy – R7 1700 vs. R7 1800X, 7700K, 7600K
TimeSpy (Dx12) shows a more combative showing from Intel’s 6900K than FireStrike, but generally does put the R7 1700 in a favorable position. Here’s the FPS:
Performance here is impressive when overclocked, outpacing the stock 6900K at nearly one-third the price. That's synthetic, of course, but bodes well for price:performance.
Let’s move on to games. This is where AMD’s R7 1800X fell short in our last review, ultimately performing about on par with last-gen i5 CPUs while being priced at 2x the cost (or, if you look at frametimes, the 7700K would also be a good comparison). Given the 1700’s significantly lower price, the stack-up should be more interesting. We still don’t recommend an 1800X for gaming, just like we don’t recommend a 6900K for pure gaming. Both are an awful deal for gaming machines, and we are ultimately GamersNexus.
Note that AMD’s own marketing materials from the Polaris launch indicate ~95% of PC gamers, by their marketed numbers, are on 1080p or below for resolution. This image was distributed to media when encouraging that the RX series cards be tested on not just 1440p, but 1080p. Aside from the methodological reasons for testing CPUs at 1080p – like actually looking at comparative CPU performance, not artificially bottlenecked performance – there are also market reasons to test at 1080p. Primarily, as AMD suggests, that most folks are still on 1080p displays. Just like last time, we’ll run a few 1440p tests at the end – though you’ll see that those make AMD’s new CPUs “equal” insofar as exhibiting the same performance given an external bottleneck.
Battlefield 1 – R7 1700 vs. 7700K, 7600K, 1800X
We’re starting with Battlefield.
In Battlefield 1, the AMD R7 1700 SMT1 operates at around 129FPS AVG, or 85FPS 1% lows with 68FPS 0.1% lows. This is sandwiched between the i5-4690K stock and i5-3570K stock, all three of which have more or less identical lows with regard to user perception. Now again, that’s not great price-to-performance, but it is a whole lot easier to justify the R7 1700 as in a gaming machine than the significantly more expensive 1800X. We’ll get to that in the conclusion.
Overclocking the R7 1700 to 4.0GHz and with an equalized 2933MHz memory speed, the CPU is now outperforming the stock 1800X, or performing equally to the 1800X overclocked. The 1800X overclocked to 3.9GHz was able to sustain a more stable clock-rate, with the 1700 fluctuating more frequently, and so places higher in its low values at 109 and 94 versus 89 and 75 on the 1700. This higher sustained clock-rate is primarily a benefit of the additional power throughput, higher voltage, and higher baseline of the 1800X (read: ultimately, more clock stability under a thermally controlled scenario). If you’re OK with higher frametimes on the low-end, the 1700 overclocked is clearly significantly better value, as the average framerate is imperceptibly different. The lows in this particular game are also largely imperceptibly different, though do start to grow distant in the OC versions. Note: When we talk about imperceptible differences, we are referring to a few FPS -- not 10-30FPS (or whatever it may be). Although those differences may be arguably imperceptible, there are two things to consider: (1) 144Hz users care, and (2) more powerful GPUs or multi-GPU configurations will out those differences as resolution increases.
And so the 1700 is a whole lot better than the 1800X in price-to-performance for gaming, but still on-par with Intel i5 CPUs from last generation. Comparing against the i5-6600K, the 1700 stock performs measurably worse across all three metrics -- perception isn’t different, as we said last time, but you generally don’t pay more for something that’s the same or worse. That said, the 1700 includes a stock cooler, so that has some value, and the motherboards should theoretically be cheaper once the AM4 market saturates (but we wouldn’t recommend overclocking on just any motherboard – some of the VRM designs run hot). That brings the price closer to an i5 – not quite there, but close – while offering some significant gains in production. We absolutely would not recommend an i5 CPU for production tasks that you software accelerate, if at all avoidable, and the 1700 helps to fill that mixed workload gap where the 1800X fell short for gamers at its price.
Just like the 1800X, disabling SMT does seem to boost overall performance at best, or result in no significant difference at worst.
Let’s move on to Total Warhammer.
Total War: Warhammer – R7 1700 vs. 1800X, 7700K, 7600K
As a reminder, Total War: Warhammer has some issues with measurement of 0.1% low values; the game is more variable in its output to begin with, but it’s also a lot more sensitive to clock stability on CPUs given the CPU load. We are sticking to AVG + 1% low for now.
That stated, the results place the R7 1700 stock at about the i3-7350K performance range, sandwiched between an i5-2500K overclocked, or i3-6300 stock, and the i3-7350K stock. It’s measurably slower than the 1800X, given the lower base and boost frequencies, and generally not that competitive against Intel in terms of price-to-performance for a pure gaming build. Of course, it’s also got that production benefit that’s so much easier to justify for streaming and content creating gamers at its price-point, considering the lowered price over the 1800X. We still don’t recommend the 7350K at its price, just like we didn’t when we reviewed it, but the point stands that the stock 1700 isn’t all that impressive in terms of raw FPS throughput. It does have better 1% low values – which are a bit more measurably consistent in this game than 0.1% lows – when compared to its neighboring Intel CPUs. The 1700 tends to sit around 88FPS 1% lows when stock, versus the 7350K’s 72FPS 1% lows. Again, as we keep reiterating, that’s not necessarily visible at this point in the game – but does show where advantages lie in each title for each architecture. The thread advantage in Total War, although evidently limited in just how much you get out of it, does have a trend of aiding heavily threaded CPUs (6900K, Ryzen R7).
The overclocked variant is a lot more interesting, particularly because – again – this is a trivial overclock on our CPU. Your mileage may vary, but ours hits 4.0GHz easily in prolonged TW: Warhammer and other burn-in benchmarking. The performance now is at 136FPS AVG, with high sustained 1% low values. The CPU plants itself just under an overclocked 7350K at 5.0GHz, and outmatches the 1800X stock CPU. The extra 100MHz on the clock also helps overcome the overclocked 1800X, since ours was limited to about 3.9GHz.
As for SMT, disabling the feature results in boosted performance even over the overclock. Combining the two will be a fun test for the future.
This again pushes the 1800X further out of our recommendations for gaming builds, or even mixed workload (read: video editing / 3D animation + gaming) builds. The 1700 usurps most potential the 1800X had for our audience when allowing for an overclock.
Watch Dogs 2 – R7 1700 Benchmarks
Watch Dogs 2 is heavily multithreaded, but does push more GPU load than Total War: Warhammer does.
Without an overclock, the stock R7 1700 performs ahead of an i5-6600K in average framerate, matching the 1800X stock in its measurably but imperceptibly worse frame latencies. This performance puts the $330 R7 1700 under the $240 i5-7600K when both are stock, momentarily ignoring other cost factors in the build since we don’t yet know how good that stock cooler is. Overclocking pushes the 1700 past the i5-7600K, and close to the 7700K when hyperthreading is off. Hyperthreading on, we see a big performance advantage for Intel’s architecture, the $340 i7-7700K now performing at 113FPS AVG to AMD’s 86FPS AVG overclocked, or 80FPS AVG without overclock.
Regarding the 1800X, we’re able to effectively match the CPU when overclocking the 1700, and we outperform the stock 1800X.
Ashes of the Singularity (Dx12)
We didn’t have time to add SMT0 to these benchmarks, though based on previous performance, this is one of the use cases that isn’t influenced as much.
Regardless, Ashes plants the R7 1700 (stock) adjacent to the stock, SMT0 1800X with its 2933MHz memory. We’re looking at comparable frametimes and averages across the board, until moving up to the 1800X stock to see a few FPS bump. Overclocking the 1700 to 4.0GHz, further than we could get our 1800X, results in superior performance. Raw framerate makes this look like a shallow victory, but price puts things into perspective.
Both the 1700 and 1800X are handily defeated by an i7-7700K, and perform similarly (with Ryzen overclocked) to an i7-4790K. More on the arguments here below.
Metro: Last Light – R7 1700 Benchmarks
Metro: Last Light positions the R7 1700 between the i5-4690K and i5-6600K, though with significantly better 0.1% low metrics. MLL has historically proven rough on quad-thread CPUs in the absolute low-end of the frametime performance ratings, and that proves true here. AMD’s similarly performing 1700 does hold stronger in lows than nearby i5s, but still underperforms significantly when compared to similarly priced i7 CPUs (including the i7-4790K).
Regarding the 1800X, it’s the same story: Overclocking brings us ahead of the 1800X, though just under the 1800X with SMT off.
GTA – R7 1700 Benchmarks
GTA V, as we previously demonstrated, poses a methodological challenge for benchmarking <8T CPUs that are capable of hitting 187FPS. This is due to engine limitaitons, as the game caps at 187FPS AVG (or ~187.5FPS), and so processors which can exceed 187FPS are not actually logging their complete frame throughput. This truncates results, impacting higher-end CPUs by making them look as if their performance is lower. The same reason is why we’ve eliminated quad core CPUs from this benchmark, as the behavior results in extreme stuttering. Learn more here.
The R7 1700 (stock, SMT1) places right around the i7-2600K, including performance in 1% and 0.1% low values. The 1700 exceeds the 2600K’s average framerate by roughly 6-7%. This positions the stock 1700 below the 1800X’s stock/SMT1 ~125FPS AVG, though overclocking the R7 1700 brings us up above the 1800X stock. The 1800X OC and 1700 OC are effectively equivalent, both below an i7-4790K.
Quick 1440p Benchmarks
Just like last time, our 1440p benchmarks are more limited as they were added for the 1800X testing:
Using AMD’s argument of representing CPUs at higher resolutions and thus introducing GPU bottlenecks, we see the 1800X is effectively irrelevant as a gaming option when GPU limited, since the two produce results which are perceptibly equivalent. Following the path of imposing external bottlenecks, the 1700 is equal to an 1800X in gaming when bumping against a choke point, with the i7-7700K outperforming both by about 10FPS AVG and in low values. The 6900K, as it has been, remains bad value for anyone who only or primarily games (particularly high refresh gamers). The same is true for the 1800X.
With Battlefield 1 at 1440p, we see again that the stock 1700 and stock 1800X are equal in performance, thanks to the imposition of that GPU limitation. Intel is also more or less equal, perceptibly, though still slightly ahead.
Conclusion: For Our Audience, Ignore the 1800X, Look to 1700
The AMD R7 1700 is priced in a way that makes it worthy of consideration at the cost, but with some of the same caveats as the R7 1800X. In the case of the R7 1700, we see mixed workload use cases shine better at the $330 price point, where the CPU deftly outperforms the 7700K in Premiere and Blender rendering tasks when CPU accelerated. If you’re pushing renders to the CPU and doing some gaming, it’s not a bad buy – it just depends on how much production you do versus how much gaming, or how much you care about hitting high refresh rates in games. For folks favoring high refresh rate (e.g. 120-144Hz), the 7700K is the best option – overclocking recommended. Hands down. If you’re only gaming, but don’t necessarily care about high refresh, we’d still point you toward a more cost effective i5 (or even i7, regardless of -K tag). For a mix of gaming + video editing or animation, the 1700 is actually defensible where the 1800X’s price disallowed the same defense. Again, we’re considering our core audience, here. If you’re not part of that, perhaps consider other reviews.
This further reinforces our stance on the 1800X: You are far better off buying an R7 1700 for $330, applying a 5-minute overclock and a half-decent cooler, and netting a chip that outperforms a stock 1800X, or performs equally to an overclocked version. There is no reason to purchase an 1800X if you are OK with the idea of applying an overclock. Now, we’d normally assume that most folks aren’t overclocking – see: “just want the best” people who buy a 7700K – but it’s a few minutes of work and grants performance that minimally equals the R7 1800X.
The one point of hesitation here is on the future binning of these CPUs. We’re not sure if the early 1700 models will OC higher, or if they’re modified 1800X parts that retain the same headroom. Our sample overclocks to 3.9-4.0GHz, depending on workload, and can push memory to 2933 in those overclocked cases. We know some other folks were able to achieve similar overclocks on their review sample R7 1700 CPUs.
We’re still waiting on our purchased 1700 to come in so that we can review the stock cooler, but overall, thermals are significantly reduced over the 1800X. The performance at the price point, when considering a light overclock, is a far better argument in terms of price than the 1800X. Even for production workloads, the overclocked 1700 produces equal performance to the overclocked 1800X.
So again, we strongly advise against the 1800X as a CPU for a gaming machine. If you’re doing zero production, you’re not doing any content creation, then you’re still generally getting a better deal with an i5 or i7 CPU. For folks who are combining content creation (similar to what we do) with gaming, or may be considering streaming, the R7 1700 is actually a viable chip – and far more so than the 1800X.
AMD is its own best competition at the price point, when considering those use cases.
We still have to look at the 1700X, but based on initial results, the R7 1700 looks like the hero of AMD’s initial lineup. It is far easier to argue this CPU, just know that you’re still limited in terms of gaming performance at the cost. Rendering workloads are far boosted over equivalently priced Intel CPUs. We can stand behind the R7 1700 under the right usage conditions – just figure out if you fit into those conditions.
The story may change for Ryzen at some point, but we don’t review based on promises. We review based on what’s available on the market now. We’ll revisit the CPUs as updates roll-out.
Editorial, Test Lead: Steve Burke
Researcher, Tester: Patrick Lathan
Video Producer: Andrew Coleman
Additional Video: Keegan Galalick