AMD Ryzen 5 1500X and 1600X CPU Review

Coming in swiftly behind its big brother we have Ryzen 5. This is AMD‘s offering for the more budget-minded PC user. Today we’ll be looking at the hex-core 1600X and the quad-core 1500X. These CPUs both have the XFR technology we’ve seen from the Ryzen 7 1800X and 1700X allowing for boost speeds over their typical max. Without further adieu, on to the main show.

Specifications and Features

Looking at the specifications table below, the 1600X is a hex-core with SMT for a total of twelve threads and the 1500X is a quad-core with SMT for eight threads. This total core/thread count comes from the use of two CPU Complexes (CCX), more on this arrangement in a second. The base clock comes in at 3.6 GHz and will boost two cores (four threads) to 4.0 GHz for the 1600X and 3.5 GHz boosting to 3.7 GHz on the 1500X. The inclusion of XFR (Xtended Frequency Range) technology allows another 100 MHz (200 MHz for the 1500X) over both the base and boost clocks, when temperature allows. TDP of these two CPUs comes in at 95 W for the 1600X and 65 W for the 1500X. The cooling medium between the die and IHS is solder, instead of thermal paste as Intel has used on their Mainstream CPUs.

The question on everyone’s lips regarding Ryzen 5 has been “how will AMD handle reduced core counts?” and I can tell you they will be balanced between CCX’s. The Ryzen 5 1600X has two CCX’s with three cores enabled on each. The Ryzen 5 1500X has two CCX’s with two cores enabled on each. AMD is keeping it balanced and using their selective core disabling functionality to drop the core count.

Memory on this CPU/platform supports a total of 128 GB with the base specification of DDR4-2400 in a dual channel configuration. It does not support ECC memory.

Regarding PCI Express (PCIe) support, Ryzen offers a total of 24 lanes out of the CPU allowing good flexibility for multiple cards, PCIe-based NVMe SSDs, and other PCIe-based devices. Sixteen of the lanes are dedicated to graphics, four are dedicated to the native M.2 PCIe NVMe slot, and the last four connect to the chipset. Different chipsets will provide their own additional PCIe lanes for even more device connectivity.

Windows 10 is the officially supported platform for Ryzen. That said, there will be drivers available for use with Windows 7 and 8.1, but know there is no official support for these older operating systems.

SpecificationsRyzen 5 1600XRyzen 5 1500X
# of Cores64
# of Threads128
Base Clock Speed3.6 GHz3.5 GHz
Boost Clock Speed4.0 GHz3.7 GHz
Instruction Set64-bit64-bit
Instruction Set ExtensionsSSE 4.1/4.2/4a, AVX2, SHASSE 4.1/4.2/4a, AVX2, SHA
Lithography14 nm FinFET14 nm FinFET
Transistor Count4.8 billion4.8 billion
TDP95 W65 W
Thermal Solution SpecSolderedSoldered
Integrated GraphicsN/AN/A
L1 Cache128 KB I-Cache (64 KB per CCX)
128 KB D-Cache (64 KB per CCX)
128 KB I-Cache (64 KB per CCX)
128 KB D-Cache (64 KB per CCX)
L2 Cache3 MB (512 KB per core)2 MB (512 KB per core)
L3 Cache16 MB (8 MB per CCX)16 MB (8 MB per CCX)
Memory Specifications
Max Memory Size128 GB128 GB
Memory TypesDDR4-2400DDR4-2400
# of Memory Channels22
ECC Memory SupportNoNo
Expansion Options
PCI Express Revision3.03.0
PCI Express Configurations1×16+1×4+1×4, 2×8+1×4+1×41×16+1×4+1×4, 2×8+1×4+1×4
Max # of PCI Express Lanes24 Lanes24 Lanes

The table below is a list of the Ryzen lineup. Every CPU on this list is overclockable, assuming you buy a motherboard with a chipset capable of doing so. Only SKUs with an X on the end have the new XFR (eXtended Frequency Range) technology, note. According to AMD, the X SKU processors are binned and manufactured to be better overclockers.

AMD Ryzen CPU ModelCores/
Threads
Base ClockBoost ClockL3 CacheCooler IncludedXFRTDP
Ryzen 7 1800X8/163.6 GHz4.0 GHz16 MBNoYes95 W-SR3+
Ryzen 7 1700X8/163.4 GHz3.8 GHz16 MBNoYes95 W-SR3+
Ryzen 7 17008/163.0 GHz3.7 GHz16 MBWraith SpireNo65 W
Ryzen 5 1600X6/123.6 GHz4.0 GHz16 MBNoYes95 W
Ryzen 5 16006/123.2 GHz3.6 GHz16 MBWraith SpireNo65 W
Ryzen 5 1500X4/83.5 GHz3.7 GHz16 MBWraith SpireYes65 W
Ryzen 5 15004/83.2 GHz3.4 GHz16 MBWraith StealthNo65 W
CPU Clock Speed Breakdown

To make it absolutely clear what the clock speed is in all loading/temperature scenarios, please see the table below.

AMD Ryzen 5 1600XHigh Temp Speed
(No XFR)
Low Temp Speed
(XFR Active)
All Cores Loaded3.6 GHz3.7 GHz
Two Cores (Four Threads) Loaded4.0 GHz4.1 GHz
One Core (Two Threads) Loaded4.0 GHz4.1 GHz
AMD Ryzen 5 1500XHigh Temp Speed
(No XFR)
Low Temp Speed
(XFR Active)
All Cores Loaded3.5 GHz3.6 GHz
Two Cores (Four Threads) Loaded3.7 GHz3.9 GHz
One Core (Two Threads) Loaded3.7 GHz3.9 GHz

Product Tour

Below are some images from AMD of the product packaging for the new Ryzen 5 CPUs. The 1600X has no cooler with it, the 1500X comes with Wraith Spire (pictured later).

Retail Boxes – Front

Retail Boxes – Rear

Taking a look at the Wraith coolers, we see one more than was in the retail box! AMD also supplied a Wraith Max for testing purposes, which I’ll get to in a later article. On the left is the Wraith Spire which was included with the 1500X.

Wraith Spire and Max – Front

Wraith Spire and Max – Rear

Next up are pictures of the two Ryzen 5 samples we have, front and back. I see no discernible differences between the CPUs other than the laser markings on the IHS.

AMD Ryzen 5 CPUs – Front

AMD Ryzen 5 CPUs – Rear

Benchmarks

The data we have gathered will give us a great idea of its performance both at stock (no turbo), and matching clockspeeds to see IPC performance differences between them all. I have included Kaby Lake results with the i7-7700K, the hex core HyperThreaded Haswell-E i7-5820K, and the big-boy Ryzen 7 1800X.

i7-7700K
i7-5820KRyzen 7 1800X
MotherboardGIGABYTE Z270X-Gaming 8ASRock X99 OC FormulaASUS Crosshair VI Hero
MemoryCorsair Vengeance LPX 2×8 GB DDR4-3000 15-17-17-35Kingston Hyper X 4×4 GB DDR4-3000 15-15-15-35G.SKILL Trident Z 2×8 GB DDR4-3866 18-19-19-39
HDDOCZ Trion 150 480 GBSamsung 950 Pro 512 GBOCZ Trion 150 480 GB
Power SupplyEVGA SuperNova G2 850 WEVGA SuperNova G2 750 WEVGA SuperNova G2 850 W
Video CardEVGA GTX 980 Ti FTW GAMINGGIGABYTE GTX 980 Ti Xtreme GamingEVGA GTX 980 Ti FTW GAMING
CoolingCoolerMaster Glacer 240LCustom Loop with EK LTX CPU Block and 5.120 RadiatorCustom Loop with Alphacool XP3 and 3.120 Radiator
OSWindows 10 x64Windows 10 x64Windows 10 x64

And the test system:

Test Setup
CPUAMD Ryzen 5 1500X/1600X
CPU CoolerCustom Loop with Alphacool XP3 and 3.120 Radiator
MotherboardASUS Crosshair VI Hero
RAMG.SKILL Trident Z 2×8 GB DDR4-3866 18-19-19-39
Graphics CardEVGA GTX 980 Ti FTW GAMING
Hard DriveOCZ Trion 150 480 GB
Power SupplyEVGA SuperNova G2 850 W
Operating SystemWindows 10 x64
BenchmarksSee Below
Equipment
Digital Multimeter

Up first is the MSI B350 Tomahawk motherboard sent by AMD in their care package, seems to be quite a nice budget motherboard for Ryzen. This board boasts a six-phase VRM, isolated audio section, and plenty of connectivity features. There are display outputs on the Tomahawk, allowing a Ryzen-based APU to be used later down the road. Pair this with something from the Ryzen 5 lineup, a decent GPU, and you’ll have a fantastic gaming build without breaking the bank.

This slideshow requires JavaScript.

The Geil “Hardcore Gaming Memory” sent by AMD is from their RGB lineup. Beware tower air coolers though, these sticks are incredibly tall. These require external power for the RGB LEDs, keeping them from drawing extra power over the DIMM slot.

This slideshow requires JavaScript.

Benchmarks Used

All benchmarks were run with the motherboard being set to optimized defaults (outside of some memory settings which had to be configured manually). When “stock” is mentioned along with the clockspeed, it does not reflect the boost clocks, only the base clocks. I tested this way as it seems motherboards are different in how they work out of the box. This takes out any differences in how AMD/Intel utilize their turbo features and how the motherboards handle turbo, so this is more of a “run what you brung” type of testing for stock speeds. Memory speeds were set at DDR4-3000 15-15-15-35 for all testing, regardless of the kit specifications. The only exception to this is the AMD system running at DDR4-2933 16-15-15-35, this is due to how the memory dividers and timings are handled.

After the testing, we then shifted to comparing the AMD and Intel systems all at the same clockspeeds (4 GHz). This testing will flesh out the difference in Instructions Per Clock (IPC) between the samples. This also applies to the gaming tests.

Please note, I re-ran all the 1800X tests for the most consistent results possible. This way all the Ryzen results you see here are on the Crosshair VI with the current BIOS, same cooling, and same Windows 10 updates. Typically I would not have done this, but with a brand new architecture the improvements/tweaks/fixes (both in BIOS and Windows) are rapid-fire.

CPU Tests
  • AIDA64 Engineer CPU, FPU, and Memory Tests
  • Cinebench R11.5 and R15
  • x265 1080p Benchmark (HWBOT)
  • POVRay
  • SuperPi 1M/32M
  • WPrime 32M/1024M
  • 7Zip

All CPU tests were run at their default settings unless otherwise noted.

Gaming Tests

All game tests were run at 1920×1080 and 2560×1440. Please see our testing procedures for details on in-game settings. Due to availability of some of the older CPU’s we will be using a 980Ti for testing the games.

  • 3DMark Fire Strike Extreme
  • Crysis 3
  • Dirt: Rally
  • Ashes of the Singularity
  • Rise of the Tomb Raider

AIDA64 Tests

Up first, the AIDA64 CPU tests. These tests are at the AMD/Intel base clock speeds listed. We can see the 1500X and 1600X making a great showing against the 7700K and 5820K here, coming in trading blows through Queen, PhotoWorxx, and Zlib. The Hash score is far and beyond the i7’s due to the utilization of the SHA extensions. AES tests utilize some benefits of the Zen architecture, which is why Intel trails there.

AIDA64 CPU – Stock
AIDA64 CPU – Raw Data
CPUQueenPhotoWorxxZlibAESHash
1500X @ 3.5 GHz4610818836328.33112210568
1600X @ 3.6 GHz65529209645014678416312
1800X @ 3.6 GHz8364020600663.86398621749
i7-7700K @ 4.2 GHz5121523117378.5191414817
i7 5820K @ 3.3 GHz5939530319434.1229745179

Next we see the FPU tests. In VP8 most of the testing was close to the same due to it only utilizing up to 6 threads, this made the clock speed differences really show up. In Julia and Mandel the Intel CPU’s shine due to some microarchitecture constraints of Zen, much how AMD shone above in AES and Hash. SinJulia is based on the classic x87 instruction set, but we can see how cores scale for this test here.

AIDA64 FPU – Stock
AIDA64 FPU – Raw Data
CPUVP8JuliaMandelSinJulia
1500X @ 3.5 GHz66911774992395995
1600X @ 3.6 GHz727326624142699252
1800X @ 3.6 GHz7949365151902512337
i7-7700K @ 4.2 GHz798035687191975060
i7 5820K @ 3.3 GHz667439957214756448

The memory tests pretty much speak for themselves. We can see the benefit of quad-channel on the 5820K, but otherwise it’s a pretty flat set of results. Memory latency is a bit high on Ryzen currently, but BIOS tweaks are working on that. Let’s be honest here, the vast majority of users will never notice the extra 20-30 ns since most usage cases are not memory constrained.

AIDA64 Mem – Stock
AIDA64 Mem – Raw Data
CPUReadWriteCopyLatency
1500X @ 3.5 GHz43487430803783983.9
1600X @ 3.6 GHz44057428823816784.4
1800X @ 3.6 GHz43833431093789284.6
i7-7700K @ 4.2 GHz42147444163768949.8
i7 5820K @ 3.3 GHz51596469395855359.9

Real World Tests

In the next set of testing the 1600X really stood out to me. In all tests except x265 it beat out both Intel offerings, only the 7700K won over this test in regards to the 1600X. The 1500X was mostly sitting 15-20% behind the 7700K here, which is huge considering the targeted competition for it is an i5.

Cinebench R11.5/R15, POVRay, x265 (HWBot), 7Zip – Stock
Cinebench R11.5/R15, POVRay, x265 (HWBot), 7Zip – Raw Data
CPUR11.5R15POVRayx2657Zip
1500X @ 3.5 GHz8.727961654.5320.9221812
1600X @ 3.6 GHz13.3912182496.0230.6131642
1800X @ 3.6 GHz17.6816003299.7739.7539713
i7-7700K @ 4.2 GHz10.079181960.5433.2525772
i7 5820K @ 3.3 GHz11.010122082.8722.4230617

Pi-Based Tests

In SuperPi the clock speeds and Intel’s IPC advantage really showed up due to it being a single threaded test. In wPrime though the 1600X came up and played with both Intel offerings seen today.

SuperPi 1M/32M, wPrime 32M/1024M – Stock
SuperPi and wPrime Benchmarks – Raw Data
CPUSuperPi 1MSuperPi 32MwPrime 32MwPrime 1024M
1500X @ 3.5 GHz11.829609.3166.671191.333
1600X @ 3.6 GHz11.501591.6005.032125.124
1800X @ 3.6 GHz11.548592.3504.374100.900
i7-7700K @ 4.2 GHz8.796463.4955.201153.589
i7 5820K @ 3.3 GHz10.883541.9534.763142.087

Game Results

Just a reminder, all tests from this point forward have all CPU’s running at 4GHz instead of their stock speeds. And as we expected, almost all game results were within a margin of error of each other. The only really notable difference here is in Ashes DX12, which has been optimized for Ryzen now and can also utilize a lot of cores.

1080p Gaming Results – Head to Head
1440p Gaming Results – Head to Head

Most of the scoring differences here in 3DMark Fire Strike Extreme came from the scaling of cores in the Physics test. Recall from my review of the 1800X how SMT is more efficient than HT, we see this benefit here where the 1500X has a higher Physics score than the 7700K. Note also the 1600X had a higher Physics score than the 5820K!

3DMark Fire Strike Extreme – Head to Head

Head to Head Results

In our head to head results, we ran all of the systems at 4 GHz. This shows the differences in IPC and cores directly. Overall, we see good scaling when upping the CPU speed. I was really impressed with how the 1600X performed beside the 5820K while clock speeds are equal, nice to see it’s still trading blows.

AIDA64 CPU – 4 GHz
AIDA64 FPU – 4 GHz
AIDA64 Mem – 4 GHz
Cinebench R11.5/R15, POVRay, x265 (HWBot), 7Zip – 4 GHz
SuperPi 1M/32M, wPrime 32M/1024M – 4 GHz

Overclocking

Ryzen 5 1600X

Up first for pushing the overclock further is the 1600X. It reached a max speed, on all cores and threads, of 4125 MHz. This was while running DDR4-3200 16-15-15. Below is a screenshot running Cinebench R15 and SuperPi 1M at the elevated speed. There were nice gains over the previous results.

To go further here, I also ran 3DMark Fire Strike Extreme again at these elevated settings. The GPU settings remained the same as gaming testing above. Most of the gain here was in the physics score. It went from 18313 to 18978 here!

Ryzen 5 1500X

Next I turned the screws on the 1500X for a bit. It reached a max speed, on all cores and threads, of 4050 MHz. This was while running DDR4-3200 16-15-15. Below is a screenshot running Cinebench R15 and SuperPi 1M at the elevated speed. Again on this CPU the change in scores was impressive.

To go further here, I also ran 3DMark Fire Strike Extreme again at these elevated settings. The GPU settings remained the same as gaming testing above. Again, most of the gain here was in the physics score. It went from 13338 to 13592 here!

Information from AMD Regarding Overclocking

As a general guideline: a CPU voltage of up to 1.35 V is acceptable for driving everyday overclocks of the AMD Ryzen processor. Core voltages up to 1.45 V are also sustainable, but our models suggest that processor longevity may be affected. Regardless of your voltage, make sure you’re using capable cooling to keep temperatures as low as possible.

While there are never guarantees with overclocking, the majority of users should find that an eight core, sixteen thread, AMD Ryzen processor will achieve 4.2 GHz at a core voltage of 1.45 V. Advanced and accomplished overclockers trying to push record frequencies may find more headroom by disabling cores and/or disabling SMT on motherboards that offer these options in the BIOS.

Overclocking specifics

CPU clockspeed is configured with MULTI*Ref_Clk. AMD Ryzen™ CPUs have 0.25X multipliers.

  1. Ref_clk is 100MHz.
  2. While the ref_clk value is adjustable, system stability may be compromised when deviating from this value.
  3. Users are encouraged to use the unlocked multipliers.

AMD Ryzen™ processors do not use pre-programmed VID tables.

  1. Therefore, there is no fixed Vcore when the CPU runs in its out-of-box condition.
  2. Default Vcore will vary depending on workload and will range from 1.2-1.3625 V.
  3. Overclocking an AMD Ryzen™ processor will snap the voltage to 1.3625 V, but this value can be changed.

Voltages to look for:

  1. CPU Vcore: Look for CPU VID value (type in value)
  2. CPU SoC voltage: Look for VDDCR_SOC value (type in value). Default is 0.99 V. Adjusting this to 1.1 V may help stabilize memory overclocks.
  3. Memory voltage: Look for MEM_VDDIO (type in value) and MEM_VTT (set MEM_VTT to ½ of MEM_VDDIO). VDDIO is voltage (“memory voltage”) supplied to the DRAM ICs, and MEM_VTT powers termination logic inside the DRAM ICs. These values are OFFSETS and will read as zero. Boost Memory VDDIO to stabilize memory overclocking. (e.g. MEM_VDDIO set to +0.025 will bring 1.5 V DRAM to 1.525 V.)

AMD Ryzen Master v1.0.1

AMD recently released v1.0.1 of Ryzen Master, here are the major changes to the software:

  1. Ryzen Master now reports junction temperature, rather than tCTL. See the “temperature reporting” section of this blog for more context on tCTL.
  2. The installer no longer enables or requires HPET when Ryzen Master is installed with a system running an AGESA 1.0.0.4-based BIOS. See the “let’s talk BIOS updates” section of this blog for more context on AGESA 1.0.0.4.

Power Consumption and Temperatures

Here we see a comparison of the power consumption of the overall system for the 1500X, 1600X, and 1800X all running both their base speed and locked at 4 GHz. This was all done with the same GPU settings, motherboard, BIOS, cooling, and number of fans so any differences here are purely due to core count and clock speeds.

Power Graph

As above with the power graph, this spread of temperatures represents the 1500X, 1600X, and 1800X all running both their base speed and locked at 4 GHz. This was all done with the same GPU settings, motherboard, BIOS, cooling, and number of fans so any differences here are purely due to core count and clock speeds.

Note: These temperatures were taken with Ryzen Master prior to the release of v1.0.1.

Temperature Graph

Conclusion

Well, the performance is pleasantly surprising. The 1500X, in most cases, performed on par with the i7-7700K. Even though you can overclock the i7 further, the MSRP of the Ryzen 5 1500X is only $189. That’s a monstrous difference in price/performance. Enough to take a gaming build from a GTX 1060 to GTX 1070, etc. You’d notice the GPU much more than a slight drop in CPU performance. Then the Ryzen 5 1600X is performing well above the i7-7700K and even outperforming the i7-5820K in anything multi-threaded while holding equal in gaming performance. Again, MSRP plays a huge role here with the 1600X coming in at $249, the price of an i5-7600K currently.

Temperature is easily controlled, especially on the 1500X, and system power consumption is surprisingly low. Once we see even more BIOS and OS support coming out for Ryzen performance, stability, and compatibility will continue to get better.

Ryzen 7 performed incredibly, but Ryzen 5 is the people’s processor. Six core i7 performance for an i5 price.

Click here to find out what this means.
Click here to find out what this means.

-Austin (ATMINSIDE)

Discussion

Leave a Reply

Your email address will not be published.


*


  1. Nice write-up ATM! I'd say the real winner is the 1500X. That price/performance at stock setting is most impressive.
    If a person is gonna run stock, the 1500X wins here. You could buy a much bigger GFX card with those savings!
    Yes, very well done review. Makes it easy to compare apples with apples and oranges with oranges.
    As of now, the Wraith Spire is included only with the R7 1700. Strikes me as odd they only bundle it with that one CPU. Is the Max available yet?
    Okay, I was looking at a NewEgg promotion for the 1600X and it said it didn't come with the fan or cooler. So it comes with the others.
    Blaylock
    Excellent review as all ways Mr. INSIDE. The 1600x for dismalDIRGE's build is ready for pickup from MC along with an ASUS Prime X370 Pro. :D

    Sounds like a nice finish to that build!
    ShrimpBrime
    Nicely done. One of the best write ups for AMD RyZen around!

    Means a lot, thanks Shrimp!
    trents
    Yes, very well done review. Makes it easy to compare apples with apples and oranges with oranges.
    As of now, the Wraith Spire is included only with the R7 1700. Strikes me as odd they only bundle it with that one CPU. Is the Max available yet?

    Haven't heard anything about availability of the Max.
    IIRC AMD said in the Max was only for OEMs and not for sale for normal folks. I think I recall reading that when they introduced the new wraith coolers. I remember some not being happy on Reddit about it.
    trents
    Okay, I was looking at a NewEgg promotion for the 1600X and it said it didn't come with the fan or cooler. So it comes with the others.

    Correct. 1600 X does not come with a heat sink. However most overclocking enthusiasts would probably replace it with something bigger or liquid cooling.
    Very well written. I like the breakdown of instruction set advantages in the test results. The nuts and bolts part didn't lose me, which is saying something! :thup: