AMD Ryzen 5 1600X 6-Core Benchmarks Leak, Zen Part Number Decoder Detailed
If you thought the Ryzen hype train might start to slow down after weeks of leaks and rumors, think again. The closer we get to Ryzen's release (likely early March, with a possible announcement on February during AMD's Capsaicin event at GDC 2017), the juicier the leaks, it seems. Case in point, the latest bit of Ryzen news involves some interesting benchmarks of the Ryzen 5 1600X 6-core processor.
A quick recap is in order before we get to the numbers. AMD's Ryzen 5 1600X is one of 17 Ryzen SKUs AMD is expected to launch. The entire lineup will consist of 4-core, 6-core, and 8-core processors spread out between Ryzen 3, Ryzen 5, and Ryzen 7 series. There will be eight Ryzen 5 processors evenly split between 4-core and 6-core processors. The Ryzen 5 1600X is the fastest of the 6-core lineup—it has a 3.3GHz base clockspeed and a 3.7GHz boost, along with 3MB of L2 cache.
With the launch being so close, it's a reasonable assumption that several Ryzen processors are out in the wild, hence today's round of leaked benchmarks. These results emerged on a Chinese-language web forum. One of them shows performance metrics from CPU-Z's built-in benchmarking utility. As you can see in the screenshot above, the chip is running with a core voltage of 0.374v. We can also observe that the processor's boost clock kicked in, with the CPU ramping up to 3.56GHz.
The Ryzen 5 1600X scored 1,888 in single-threaded performance and 12,544 in multi-threaded. Those are sort of meaningless numbers by themselves so we fired up the same benchmark and ran it with Intel's Core i7-5960X, an 8-core Haswell-E chip clocked at 3GHz to 3.5GHz; an older Core i7-4970K, a 4-core Devil's Canyon (Haswell) chip clocked at 4GHz to 4.4GHz; a Core i7-4960X, a 6-core Ivy Bridge-E processor clocked at 3.6GHz to 4GHz; and and a Core i7-6900K, an 8-core CPU clocked at 3.2GHz to 3.7GHz. To flesh things out, we also included a validated benchmark run from an Intel Core i7-5930K (6-core Haswell-E, 3.5GHz to 3.7GHz) that appears in CPU-Z's database, as it matches the Ryzen 5 1600X in core count and clockspeed.Here's how things shook out:
We have to be careful not to overemphasize the results here, as this is a single benchmark comparing the results of processors using different testbeds. On top of that, the Ryzen 5 1600X is likely an engineering sample that could have some features disabled, though more importantly it's running on a platform (AM4) that hasn't had time to mature. Final performance will almost certainly be higher than what's observed here.
With all that said, the early results from CPU-Z are promising. They show the Ryzen 5 1600X coming out ahead across the board, except for against the faster clocked Devil's Canyon CPU in single-thread performance (but coming out ahead when tapping into multiple cores and threads). Our benchmark comparison is consistent with what the Ryzen testers had previously observed, which is that Ryzen's IPC (instructions per cycle, or average number of instructions executed for each clock cycle) is higher than Ivy Bridge. It appears to be higher than Haswell and Haswell-E, too.
The testers also claim that Ryzen's SMT (simultaneous multi-threading) is more efficient than Intel's Hyper-Threading. With Ryzen, AMD went with a more traditional SMT design, leaving behind the CMT (clustered multi-threading) approach it used with Bulldozer. Early indications are that the AMD's decision is paying off here.
To the point that Ryzen's final performance will likely be higher than what we're seeing now, the tester's previously noted that an earlier version of Ryzen suffered from a poor memory controller, but that more recent ES chips show significant improvement.
Ryzen also has a healthy instruction set capability: MMX (+), SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4A, X86-64, AMD-V, AES, AVX, AVX2 , FMA3. The inclusion of AVX2 is particularly interesting, as it comes in handy for processing visual data commonly encountered in consumer imaging and visual processing workloads. In that regards, including AVX2 in Ryzen brings some parity between it and with Intel's newer CPU architectures dating back to Haswell.
Beyond the benchmarks, there are reasons that overclockers should be excited about Ryzen. One is that the testers observed during their hands-on time with Ryzen that it it doesn't suffer from the "cold bug" that affects some processors when using extreme cooling, such as LN2. That should make things easier for hardcore overclockers in search of OC'ing and benchmarking records. And for the vast majority, Ryzen ES chips have been able to hit 4.3GHz to 4.5GHz on air cooling.
A quick recap is in order before we get to the numbers. AMD's Ryzen 5 1600X is one of 17 Ryzen SKUs AMD is expected to launch. The entire lineup will consist of 4-core, 6-core, and 8-core processors spread out between Ryzen 3, Ryzen 5, and Ryzen 7 series. There will be eight Ryzen 5 processors evenly split between 4-core and 6-core processors. The Ryzen 5 1600X is the fastest of the 6-core lineup—it has a 3.3GHz base clockspeed and a 3.7GHz boost, along with 3MB of L2 cache.
With the launch being so close, it's a reasonable assumption that several Ryzen processors are out in the wild, hence today's round of leaked benchmarks. These results emerged on a Chinese-language web forum. One of them shows performance metrics from CPU-Z's built-in benchmarking utility. As you can see in the screenshot above, the chip is running with a core voltage of 0.374v. We can also observe that the processor's boost clock kicked in, with the CPU ramping up to 3.56GHz.
The Ryzen 5 1600X scored 1,888 in single-threaded performance and 12,544 in multi-threaded. Those are sort of meaningless numbers by themselves so we fired up the same benchmark and ran it with Intel's Core i7-5960X, an 8-core Haswell-E chip clocked at 3GHz to 3.5GHz; an older Core i7-4970K, a 4-core Devil's Canyon (Haswell) chip clocked at 4GHz to 4.4GHz; a Core i7-4960X, a 6-core Ivy Bridge-E processor clocked at 3.6GHz to 4GHz; and and a Core i7-6900K, an 8-core CPU clocked at 3.2GHz to 3.7GHz. To flesh things out, we also included a validated benchmark run from an Intel Core i7-5930K (6-core Haswell-E, 3.5GHz to 3.7GHz) that appears in CPU-Z's database, as it matches the Ryzen 5 1600X in core count and clockspeed.Here's how things shook out:
| Single Thread |
Multi Thread |
|
| AMD Ryzen 5 1600X |
1,888 | 12,544 |
| Intel Core i7-4790K |
1,994 | 8,326 |
| Intel Core i7-4960X |
1,623 |
10,216 |
| Intel Core i7-5930K |
1,583 |
10,375 |
| Intel Core i7-5960X |
1,554 | 12,693 |
| Intel Core i7-6900K | 1,856 | 13,849 |
With all that said, the early results from CPU-Z are promising. They show the Ryzen 5 1600X coming out ahead across the board, except for against the faster clocked Devil's Canyon CPU in single-thread performance (but coming out ahead when tapping into multiple cores and threads). Our benchmark comparison is consistent with what the Ryzen testers had previously observed, which is that Ryzen's IPC (instructions per cycle, or average number of instructions executed for each clock cycle) is higher than Ivy Bridge. It appears to be higher than Haswell and Haswell-E, too.
The testers also claim that Ryzen's SMT (simultaneous multi-threading) is more efficient than Intel's Hyper-Threading. With Ryzen, AMD went with a more traditional SMT design, leaving behind the CMT (clustered multi-threading) approach it used with Bulldozer. Early indications are that the AMD's decision is paying off here.
To the point that Ryzen's final performance will likely be higher than what we're seeing now, the tester's previously noted that an earlier version of Ryzen suffered from a poor memory controller, but that more recent ES chips show significant improvement.
Ryzen also has a healthy instruction set capability: MMX (+), SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, SSE4A, X86-64, AMD-V, AES, AVX, AVX2 , FMA3. The inclusion of AVX2 is particularly interesting, as it comes in handy for processing visual data commonly encountered in consumer imaging and visual processing workloads. In that regards, including AVX2 in Ryzen brings some parity between it and with Intel's newer CPU architectures dating back to Haswell.
Beyond the benchmarks, there are reasons that overclockers should be excited about Ryzen. One is that the testers observed during their hands-on time with Ryzen that it it doesn't suffer from the "cold bug" that affects some processors when using extreme cooling, such as LN2. That should make things easier for hardcore overclockers in search of OC'ing and benchmarking records. And for the vast majority, Ryzen ES chips have been able to hit 4.3GHz to 4.5GHz on air cooling.
The last thing to cover here is how to decipher AMD's lengthy numbering scheme for Ryzen. This will come in handy with all the leaks coming out. For example, here's a screenshot of a quad-core Ryzen processor from the same forum thread:
The chip's model number is listed as ZD3201BBM4K4_34/32_Y. Using the decoder chart, we see it's a third-generation desktop sample with a 3.2GHz base frequency, 3.4GHz boost frequency, and a 65W TDP. It also lets us know that this is a 4-core chip with 2MB of L2 cache and 8MB of L3 cache. To extract these details, all you have to do is match up the letters and numbers on the decoder chart and use the accompanying legend to translate it all.
The chip's model number is listed as ZD3201BBM4K4_34/32_Y. Using the decoder chart, we see it's a third-generation desktop sample with a 3.2GHz base frequency, 3.4GHz boost frequency, and a 65W TDP. It also lets us know that this is a 4-core chip with 2MB of L2 cache and 8MB of L3 cache. To extract these details, all you have to do is match up the letters and numbers on the decoder chart and use the accompanying legend to translate it all.
