In 2020, Intel is going to release a new generation of mainstream processors which have been heavily speculated as being the first 10nm chips. The Ryzen 9 3950X and Threadripper 2950X are both promising to offer good value for money in 2019 because they will be able to compete with those upcoming models from Intel.
The “ryzen 9 5950x vs threadripper 3960x” is a comparison of the two most powerful CPUs on the market today. Ryzen 9 3950X and Threadripper 2950X are both 8-core, 16-thread processors with impressive performance per dollar ratios.
When it comes to maximum performance with high-end CPUs, every tenth of a percent makes a difference.
In this post, we compare AMD’s Ryzen 9 3950X and Threadripper 2950X processors.
We determine which of them has superior performance and which has a better price-performance ratio.
The AMD Ryzen 3000 has been available since July 2019 and has flipped the rankings – and AMD released the Ryzen 9 3950X with 16 cores in November 2019.
Almost all models are unrestrictedly recommended to anybody looking for multi-threading performance in apps; no Intel CPU at a price point that is even somewhat comparable to the newcomers.
In games, Intel maintains a little advantage with the fastest CPUs, but only the top models remain in the lead, thanks to a 20% rise in Ryzen’s CPU limit.
AMD, on the other hand, decides practically everywhere on the whole package of performance, power consumption, and price.
Threadripper 2950X vs. Ryzen 9 3950X results
AMD Ryzen 9 3950X is ranked first.
- On the market, the fastest prosumer CPU
- Most Socket AM4 motherboards are compatible.
- Power efficiency that is unmatched.
- PCIe Gen 4.0 native
- In ST workloads, it doesn’t always capture coffee lake.
AMD Ryzen 9 3950X – The most powerful CPU
AMD has been developing the first 16-core CPU for a midrange platform since the Ryzen 3000 was initially announced in January 2019.
The 12-core Ryzen 9 3900X was still the top model at the start of the CPU series, but the Ryzen 9 3950X now joins the market with four extra cores and a faster clock speed.
And one thing is certain: no mid-range processor has ever been quicker.
AMD repeatedly displays their chip architecture, which consists of numerous dies for one processor, such that the CPU typically outperforms even more costly Intel versions with more cores while also using less energy, thanks to AMD’s use of the 7-nm process to build the dies.
The Zen-2 microarchitecture, compared to the Zen(+) of the previous two Ryzen generations, features a considerable boost in performance per clock (IPC, Instructions per Cycle).
A reminder: In Zen 2, AMD added a day jump prediction to the front end of the CPU cores, in addition to the Perceptron-based jump prediction; this is slower but produces better results.
There were also much bigger branch buffers, as well as a halved L1 instruction cache in lieu of a doubled micro-op cache.
The backend now has a third address unit (Store-AGU), and the bandwidth of the load/store units and the L1 data cache has been increased.
The four floating-point units now function with 256 bits rather than 128 bits, allowing the pipelines to execute AVX2 instructions in a single clock cycle, which is critical in the HPC market.
At the chip level, the L3 cache has been increased from 8 to 16 megabytes, and the memory controller can now support DDR4-3200 instead of DDR4-2933.
AMD claims a 15% increase in IPC for CPU cores, with advancements in L3 cache, memory, and clock speed boosting performance even further.
AMD was able to boost the frequency of the Ryzen 3000 to Matisse by using 7 nm cores, which is why AMD was able to do so in comparison to the Ryzen 1000/2000.
The Ryzen 9 3950X runs at 4.7 GHz at its top, which is near to the Core i9-5 9900KS’s GHz.
AMD had to pick processors over months to attain such a clock speed.
This binning is required to ensure that not only the maximum frequency is delivered to a core, but that the clock-to-voltage ratio on all cores is as efficient as feasible.
Despite having 16 cores instead of 12, the Ryzen 9 3950X has the same nominal thermal power dissipation as the Ryzen 9 3900X.
Only the finest 8C cores are used.
The Ryzen 9 3950X, like its predecessor, is made up of three chips on the inside, which is why this design is known as chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip design: chip A central I/O die (IOD) and two computation cache dies are used by AMD (CCD).
The IOD has a dual-channel DDR4 memory controller, as well as various I/O features including 24 PCIe Gen4 lanes, four USB 3.2 Gen2 ports, and up to two Sata 6 Gbps connectors.
Four lanes are designated for connecting to the chipset, four are for an NVMe SSD, and four more are available for a second SSD.
The motherboard determines whether an SSD has four or two lanes; in the latter case, the two Sata 6 Gbps connectors described above are also accessible.
Not all boards make use of all connections; in other cases, the emphasis is on the chipset’s ports, or PCIe lanes are branched out for external controllers.
The compute cache dies (CCD) have eight cores and are internally made up of two 4-core CCXes (Core Complexes) with 16 MByte L3 cache apiece.
The eight cores do not perform equally because, like all other CPUs, they are susceptible to minor manufacturing variances.
Chips near the edge of a wafer, for example, have somewhat lower quality, but chips in the centre have greater electrical qualities.
After the wafer has been divided, i.e. chopped into individual chips, all of them are examined and assessed.
This is known in the industry as binning (pre-selection): manufacturers test dozens of dies at once to see which clock and voltage they use, how much power they use, and what temperatures they achieve.
Some cores are better suited to high peak frequencies, whereas others may reach 4 GHz at low Vcore.
Both of these criteria must be satisfied with the Ryzen 9 3950X, since the 16 cores must fit inside a power budget of 105 watts (more accurately, up to 142 watts) while also delivering up to 4.7 GHz.
At the very least, AMD indicates which cores are the most suitable for overclocking: A gold star denotes the best core of a chipset, a gray star indicates the best core of each CCX, and a gray circle indicates the second best in the Ryzen Master tool.
The operating system’s scheduler, on the other hand, utilizes perplexingly different data to load single-threaded jobs onto a core, since a high boost clock through CPPC2, not only the electrical quality, is critical.
Basically, Windows 10 v1903 can accomplish this, but Microsoft claims that the scheduler behavior has improved with Windows 10 v1909, which was launched on Tuesday.
Instead of transferring jobs to slower cores, which may perform the workload more slowly, the 19H2 upgrade for the operating system is designed to prioritize favorite CPU cores more aggressively.
There’s also the fact that threads are retained inside a CCX for latency concerns rather than being shifted to the second cluster or even the second chipset in the case of AMD.
Agesa 1004b is a faster version of Agesa 1004b.
The System Management Unit (SMU), which may be upgraded via Agesa (AMD Generic Encapsulated Software Architecture) as part of the mainboard firmware, controls the CPU cores and their clock.
The Agesa is the foundation for each motherboard’s UEFI and provides the essential microcode for the corresponding chips. AMD specifies the Agesa 1004b as a requirement for complete support of the Ryzen 9 3950X.
The Ryzen 3000 alias Matisse has been operating since Agesa 0072, however this early version with limited compatibility was used for testing.
Our pre-launch sample had issues with the clock and the production of temperature sensor data, among other things, and the Agesa 1001 was the only solution.
AMD offered the performance-optimized Agesa 1002 for the launch, which we utilized to put the Ryzen 9 3900X and Ryzen 7 3700X to the test.
AMD corrected a fault in the Ryzen 3000’s random number generator (RDRAND) as well as a situation where PCIe Gen2 was improperly engaged instead of PCIe Gen4 with the Agesa 1003, notably versions 1003a to 1003abb.
In the meanwhile, several features on boards with 16 instead of 32 Mbyte firmware were missing, but they have already been restored.
AMD altered the boost behavior of the Ryzen 3000 with the Agesa 1003abba.
Since then, the Ryzen 3000 has clocked at a higher frequency in the peak region of 25 to 50 MHz, and it no longer runs at a high frequency under extremely little background load.
The current status is the Agesa 1004b, which also applies to this exam.
Matisse (Ryzen 3000) was created independently from the Ryzen 1000 and Ryzen 2000(G) as a branch of the mainline, therefore it reunites the code base.
As a result, all manufacturers use the same software base for earlier boards, such as those with the B350 chip, however only X570 devices have received upgrades thus far.
The Agesa 1004b is identical to the 1002 and 1003 variations in terms of performance:
We evaluate minor benefits or drawbacks, depending on the benchmark, but no substantial differences.
Only the app-launch sub-test in PCMark10 indicates a bigger improvement on the Ryzen 9 3900X, where AMD’s indicated boost clock improvements and utilization of the preferred CPU cores for the Agesa 1004b app app app app startup seem to have an influence.
The X570 also has improved stability, greater compatibility with PCIe devices, more reliable ACPI power states, and quicker post/boost periods, according to AMD.
These times vary depending on the motherboard; with regular settings, our Asus ROG Crosshair VIII Hero (X570) took 7 seconds.
Last but not least, the Agesa 1004b is still required for the new Eco Mode: 105/95 watt chips may be lowered to 65 watts through Ryzen Master under Windows, while 65 watt variants can be set to 45 watts.
We gave it a go for a few minutes, but first we ran the standard benchmarks at full 105 watt TDP.
18 Intel cores are no longer supported.
The AMD processors are put to the test on a ROG Crosshair VIII Hero (X570), a ROG Crosshair VII Hero (X470), and a ROG Zenith Extreme (X399), while the Intel processors are put to the test on a ROG Maximus X Hero (Z370) and a Prime X299-Deluxe.
The memory clock rate is within the manufacturer’s standards, and the images are delivered by a Geforce RTX 2080 Ti in the Founder’s Edition.
On a WD Black SN750, Windows 10 v1903 with programs and games, and Ubuntu 19.10 on a Samsung 970 Evo Pro.
In the firmware versions of the motherboards and processors, all mitigations against Microarchitectural Data Sampling, Meltdown, and Spectre are operational.
The current R0 stepping is used in our Core i9-9900KS, which affects performance somewhat when compared to the P0 iteration.
Under Windows, the 16-core Ryzen 9 3950X outperforms the 12-core Ryzen 9 3900X by roughly 11% in all tests.
If we exclude the benchmarks that only affect one or a few cores, the disparity widens to 27%.
However, it should be noted that the course was developed mainly for desktop processors with eight cores, not for machines with twice as many.
With the introduction of Threadripper v3, we will see results that scale significantly further with multithreading; in this case, the 3950X’s gap from the 3900X and other CPUs grows in lockstep.
In this course, Intel’s eight-core Core i9-9900KS is outperformed by 50%, while the Ryzen 9 3950X is almost twice as fast in programs like Blender.
The following table compares the Core i9-7980XE to the Core i9-7980XE: The current Core i9-9980XE isn’t any quicker than this 18-core CPU, which was Intel’s top HEDT model two years ago.
The Ryzen 9 3950X outperforms the 7980XE by 14%, and is much quicker while compiling Unreal Engine 4 but somewhat slower when exporting with Adobe Premiere.
The new 18-core Core i9-specifications 10980XE’s are currently under NDA.
The Threadripper 2950X, which also has 16 cores, is outperformed by the Ryzen 9 3950X by 21%.
The fast test under Linux reveals that the Core i9-7980XE struggles against the Ryzen 9 3950X:
The Intel processor is slower while encoding certain 4K-HDR-resolved frames of Tears of Steel with AV1, and the same is true for Blender with the school scene.
The Ryzen is only subjected to the dual-channel instead of quad-channel interface while building Qt 5.12 with GCC.
The findings of the 18-core Core i9-10980XE will be released in a few days.
The Ryzen 9 3950X has no surprises when it comes to power consumption:
It is rated at 105 watts by AMD, although it can handle up to 142 watts if the case temperature is below 62 degrees Celsius, same with previous Ryzen 3000 versions.
Under stress, the 3950X is somewhat more efficient than the 3900X, owing to AMD’s improved chip selection.
Both Ryzen 9s use much less power than the Core i9-7980XE or Threadripper.
The eco mode decreases power usage while maintaining a fast speed.
Conclusion: AMD Ryzen 9 3950X is the best performing CPU on the market.
A total of 16 cores has never been seen in the middle class previously.
In little over two years, AMD has quadrupled the number of CPU cores in the mid-range category, beginning with quad cores and excluding the modules of the bulldozer architecture.
Even in well-scaled applications like Blender or Cinebench, the Ryzen 9 3950X comes near to matching the performance of a 24-core Threadripper 2970WX.
The 16-core AMD CPU, thanks to Zen 2 technology and the high clock speed given by 7nm manufacture, performs well under mixed load and is just as ideal for gaming as any other Ryzen 3000.
Only CPUs like Intel’s Core i9-9900K(S) offer a little advantage — quantifiable rather than apparent – but with enormous multithreading, they have no chance.
Even Intel’s versions with more cores, such as the Core i9-7980XE with 18 cores, compute slower than the Ryzen 9 3950X virtually every time.
In terms of pricing, the AMD processor competes with the Core i9-9940X, which is no longer available, and the Core i9-10940X, which will be available shortly, but neither is a credible opponent for the Ryzen 9 3950X.
The two have an advantage only in very bandwidth-constrained conditions, which are uncommon outside of server applications.
As a result, the Ryzen 9 3950X is the greatest mid-range CPU on the market right now.
The Matisse CPU is an excellent solution for creative folks in particular.
Only those who want more PCIe lanes should choose for Intel’s Core i9 or AMD’s thread rippers, which are more costly not only in terms of CPUs but also in terms of platform.
AMD Threadripper 2950X is ranked second.
- It outperforms, consumes less power, and runs cooler than its Core i9 rivals.
- The 16C/32T arrangement packs a lot of power.
- It’s compatible with Threadripper motherboards that are already on the market.
- Ryzen Master is a simple-to-use software application.
- For optimal performance, quad-channel RAM is required.
AMD Threadripper 2950X is a fantastic CPU.
Two years ago, 16 cores and 32 threads on a desktop PC were unthinkable, until AMD came up with the first generation Ryzen Threadripper.
One year and three days later, AMD Ryzen Threadripper 2000 breaches the new ceiling, which had only just been acknowledged as incredible, and once again sets out to give high performance at a reasonable price.
Because getting 32 cores of the Ryzen Threadripper 2990WX for $1,829 is a difficult task.
Intel’s Xeon processors with 28 cores presently cost at least $8,700 USD.
On paper, Intel has no other options for desktop PCs, with the Core i9-7980XE offering 18 cores for little under 2,000 dollars.
The AMD Threadripper 2950X AMD Threadripper 2950X AMD Ryzen Threadripper 2950X AMD Ryzen Threadripper 2950X AMD Ryzen Threadripper 2950X, which was also tested today and has 16 cores and 32 threads, continues the straight tradition of the original flagship Ryzen Threadripper 1950X for the X399 platform.
Significant clock modifications, along with internal enhancements, in certain instances, promise performance advantages in all areas – all for $100 less than the previous year.
At a glance: AMD Ryzen Threadripper 2000
Because addressing 32 cores in a traditional desktop is difficult, AMD splits the new high-end CPUs into two series.
As a result, AMD’s new flagship AMD Ryzen Threadripper 2990WX processor, which has 32 cores and 64 threads and a combined 80 MByte L2 and L3 cache, is being marketed as a workstation processor (“WX”).
AMD is hoping to meet clients that are interested in learning whether they can leverage 64 threads. The Ryzen Threadripper 2970WX, with 24 cores and 48 threads, is in the shadow of the flagship – the same goes for it.
The X-Series, on the other hand, keeps practically everything the same.
The former 16-core 1950X and 12-core 1920X processors will be 1:1 ported to the Pinnacle Ridge design, offering the same major characteristics as their predecessors.
First and foremost, this implies a quicker cache owing to decreased latencies, as well as somewhat higher clock speeds due to 12-nm production.
The Threadripper 2950X and 2920X are two new entry-level Ryzen Threadripper processors for the socket TR4 platform.
The Threadripper 1900X, the smallest Threadripper, will not be renewed and will be replaced.
An summary of the new and old Threadripper in a table
Three previous models are replaced by four new ones, with the smallest model having no replacement at all.
Apart from the fundamental clock rate, AMD only gives the highest potential clock rate since there are always changing clock rates with XFR, XFR2, and Precision Boost.
They are very reliant on the system, its environment, and the applications since they are based on a broad variety of characteristics.
For example, there are no longer any predefined numbers for load on four cores.
Even with AMD, the highest allowable clock rate is usually only applicable for one core, and the lowest is never used in the ideal scenario.
In comparison to the previous year, the costs have been drastically reduced.
The 12-core CPU, in example, has been drastically lowered and now costs $649 at launch, $150 less than last year.
When you double the number of cores, you get double the amount of money: The price for 24 cores is $1,299 USD.
Only the 32-core CPU is available now.
Not all of the models will be available at the same time.
Only the 32-core CPU will be available from stock as of today.
On August 31, the 16-core model will be released. The 12- and 24-core CPUs will be ready in October without a set date.
Except for memory, Pinnacle Ridge is number four.
The two miniature Threadripper 2000 follow in the footsteps of their forerunners:
Under the soldered heatspreader, there are two active and two dummy dies.
The flagship processor, which features 32 cores as well as a 24-core variant, has four dies enabled and put on the big LGA-4096 package for TR4 sockets.
The CPU seems to be an Epyc Server processor at first inspection, however something is missing:
The memory controllers of two CCXs are linked to the controllers of the other two CCXs through Infinity Fabric rather than being sent directly to the outside.
The classic quad-channel memory interface remains. However, it can now support DDR4-2933, similar as Ryzen 2000.
Production using a 12-nm process allows for faster clock speeds.
The CPU cores are made up of components from the Zen+ family, which are identical to those present in Ryzen 2000.
Higher clock speeds are possible because to the 12 nm manufacturing process.
And, like its predecessor, AMD claims that during binning, the top 5% of chip manufacturers are sorted out and retained for Threadripper.
This guarantees that Threadripper 2000 can clock faster than Ryzen 2000 under certain load circumstances, in addition to having a higher TDP.
Latencies in the cache and RAM have been decreased.
Apart from the process, AMD has been known to focus the majority of Zen+ development on lowering cache and memory latencies.
The changes between Ryzen Threadripper 1000 and Ryzen Threadripper 2000 are somewhat fewer than the differences between Ryzen 1000 and Ryzen 2000 since Threadripper 1000 already has certain latency enhancements compared to Ryzen 1000.
While AMD claims that Ryzen 2000’s L2 cache latency is up to 34% faster than Ryzen 1000’s and the L3 cache latency is up to 16% faster, Threadripper 2000’s advancement over its predecessor is just 9% and 15%, respectively.
Threadripper retains just 2% of Ryzen’s 11 percent lead in DRAM.
There have been no further modifications to Zen+’s architecture.
A redesigned turbo mechanism, on the other hand, provides an extra boost in performance.
For multi-core loads, there’s a new Turbo.
Precision Boost 2 is Threadripper’s new Turbo mode, which takes a number of factors into consideration when determining the CPU’s clock speed.
Since Raven Ridge in mid-February, the second iteration of this technology has been in operation, raising and lowering the frequency in 25 MHz increments in milliseconds.
The limit is defined, among other things, as follows in contemporary graphics cards:
There are power and temperature limits on the CPU.
The clock is slowed by whatever is accomplished first (below the maximum possible and firmly defined turbo).
In the Ryzen 2000 test, Techtestreport has previously detailed all of the information.
XFR 2 with a temperature offset of 27 degrees
Threadripper 2000 also supports XFR 2, and the auxiliary turbo maintains known temperature and power consumption restrictions, but now operates even when all of the CPU’s cores are active.
In most circumstances, the temperature should be too high, but there is no general exclusion for XFR with XFR 2, and the cores are not limited.
AMD doesn’t say how much more clock speed XFR 2 can squeeze out of the CPU.
To make XFR even more effective, AMD included an offset temperature in the X versions of the original Ryzen processors, which fooled the fan controller into thinking the temperature was 20 degrees higher.
The fan controller behaved as if the CPU was really 80 degrees heated if the CPU was truly 60 degrees warm, like the Ryzen 7 1800X, to maintain the CPU for XFR below real 60 degrees.
This number is much higher with the Ryzen Threadripper 2000. Most tools, such as HWiNFO, are aware of this quirk and show both temperatures.
Only DDR4-2933 with a maximum of four modules
AMD inflated the specs for maximum bandwidth a little during the PR presentations and ran the systems on DDR4-3200, although the CPUs are only officially defined for DDR4-2933 — and only when used with single-rank modules.
Only DDR4-1866 is available when the platform is fully outfitted with dual-rank memory.
Threadripper 2990WX and 2950X test results
In the test, AMD Ryzen Threadripper 2990WX and 2950X battle on an Asus ROG Zenith Extreme.
The BIOS has been upgraded to a beta version, with a release date of July 31st.
The test setup is basically identical to that utilized in the large Ryzen 2000 test, which includes Threadripper predecessors, ensuring comparability.
According to AMD, all TR4 socket motherboards should be capable of supporting 32 cores.
This necessitates a BIOS update in each scenario, however BIOS flashback should not pose the same issues as with Ryzen 2000, when new CPUs could only be booted on mainboards with the updated BIOS already loaded.
In ordinary life, 32 cores are (still) generating issues.
Even the X399 platform, with its socket TR4, only supports a quad channel memory interface, therefore the limitation of the two big CPUs to a quad channel interface only helps to indirectly separate Epyc from the X399 platform.
As a result, the I/O has also been moved.
After everything is said and done, two of the four dies have no memory interface and no links to the outside world.
The “Compute Dies” are instead transferred internally to the other two “I/O Dies,” which subsequently transmit them to the outside.
In the worst-case situation, this would undoubtedly result in longer latencies, since requests from the Compute Dies’ cores must first transit via switching centers and the Infinity Fabric On Package (IFOP) into the I/O Dies, where they must then be routed to the RAM, before returning the same way.
In a diagram, WikiChip has clearly shown this.
AIDA64’s memory bandwidth measurement reveals the implementation method’s limits.
The memory bandwidth on the 32 core, for example, is lower in MB/s than on the 2950X with half as many cores, despite the fact that nothing on the platform has changed but the CPU, and both provide a four channel interface to the outside.
The latency measurement demonstrates that the editorial crew did not make any errors here, since the AMD findings are met – and the shortcomings of the 2990WX are clearly obvious.
The DRAM discrepancies may be addressed by the fact that AMD tested using DDR4-3200, whilst Techtestreport utilized the declared limit of DDR4-2933.
The performance reductions in apps that (now and) might still accompany the cutbacks are something AMD hasn’t spoken about much in the press.
Even with 64 threads, many applications must be capable of handling them.
The Windows scheduler is also implicated, according to AMD.
They’re already working with Microsoft on improving Windows to support more than 32 threads on Threadripper 2000.
The memory interface and 64 threads are both problematic with the program.
The fact that the RAM controllers are not to every software’s satisfaction may be observed, for example, in the newest version 1.22 of TrueCrypt’s successor VeraCrypt.
The performance of the 32-core is only half that of the 16-core, and corresponds to the level that would be obtained if the 32-core was solely used for dual-channel memory mode.
On the Ryzen Threadripper 2990WX, the test with just two memory locks yields the same result as the test with four memory locks: roughly 10.5 GB/s.
This software currently despises the compute-die I/O-die configuration.
Threadripper 2950X with 16 cores, on the other hand, always uses the quad channel memory interface and achieves slightly under 21 GB/s, same like its predecessor.
Even using legacy mode in the Ryzen Master Tool and reducing the cores did not enhance the 2990WX’s performance.
Even with the current version 18.05 of 7-zip, the difference between two and four memory banks on the 2990WX is hardly noticeable: With two modules, the performance is 107,335 MIPS, while in quad channel mode, the performance is 111,624 MIPS.
This is barely a three-percentage-point rise. Even with the extra memory bars and cores, the 16-core unit already produces almost 106,000 MIPS — growing through the additional memory bars and cores is ultimately not conceivable.
Other programs, such as Cinebench or POV-ray rendering, or Adobe applications, may scale through the threads, but the performance of the 2990WX is still below that of the 2950X in certain circumstances since the memory/fabric bandwidth is far too low to feed the 32 cores with enough data.
As a result, AMD made three more rendering programs accessible to editorial offices for testing one working day before the event began.
Techtestreport’s course is independent of manufacturer recommendations, and these recommendations show that rendering is now the most common application area in which 32 cores may be employed.
Many workstations are being acquired specifically for this reason.
However, there are certain exceptions.
Aside from Blender’s normal test, which places the 2990WX ahead of the 2950X, the situation is even reversed in another project file:
The 16-core takes an hour to finish, while the 32-core takes an hour and 20 minutes.
It doesn’t matter whether Blender is in final version 2.79 or experimental version 2.80: the project won’t be able to load all threads on AMD’s Threadripper in any instance.
In addition, a test with two projects in Handbrake 1.1.1 shows that the 32-core is performing poorly. It’s close to 16-core, but not quite there yet.
The stress on the cores causes it to fail here as well.
In games, benchmarks
Threadripper 2000, AMD’s game mode announced last year, is also available for gaming.
On a 16-core processor, this reduces the number of cores and threads by half; on a 32-core processor, they are even quartered. AMD hopes that by doing so, it would be able to address game compatibility issues.
With the 16-core CPU, Game Mode was no longer required to allow games to operate together as it was last year.
In this mode, though, one or the other title put in a little more effort:
Threadripper 2950X can catch up to Ryzen 7 2700X in certain games thanks to Game Mode.
This should ideally be the case, given that the clock rates are almost identical.
The “Game Mode” is required for the 2990WX.
However, with the 32-core, Game Mode was required to avoid crashes in Far Cry 5 and severely poor FPS in Total War.
FPSs with less active cores drew notice in other games as well. However, in the games, the rating is eventually removed since it distorts the image too much.
Conclusion: AMD Threadripper 2950X is a fantastic CPU.
Threadripper 2950X: no issues with 16 cores
The Ryzen Threadripper 2950X with 16 cores, on the other hand, has a consistently favorable image.
In comparison to its predecessor, its performance is improving in all areas, and no new tangible issues have been observed on the test track as of summer 2018.
As a result, the processor offers a far more complete overall image.
Of course, 16 cores and 32 threads want to be loaded by the correct applications as well, but they are more diversified as a result of the predecessor’s year-long preparation.
And, since every program that is devoted to multi-threading support does not yet employ all 32 threads, there is still opportunity for advancement in this CPU.
In any case, it should be evident at the end of the day, before purchasing, what the CPU will be used for.
In comparison to the middle class, the HEDT platform is and will stay associated with a large premium; the traditional gamer is considerably better off there.
However, the 16-core’s increased PCIe lanes, when combined with the correct applications, make it very appealing.
The 32-core, on the other hand, still has too many construction sites on the home desktop; it should only be considered as an alternative to Epyc by those who know how to manage it.
Conclusion: Threadripper 2950X vs. Ryzen 9 3950X
Performer of the Year
Second Position
Model
AMD Ryzen Threadripper 2950X
Result of the Test
June 2020 Result of the Test 9.8/10 Excellent
June 2020 Result of the Test 9.6/10 Very Good
Cores
16 CPU cores (3.5 GHz up to 4.7 GHz Max Turbo )
six cores (3.8 GHz up to 4.4 GHz Max Turbo )
Performance of the Workstation (Applications)
Pros
- On the market, the fastest prosumer CPU
- Most Socket AM4 motherboards are compatible.
- Power efficiency that is unmatched.
- PCIe Gen 4.0 native
- It outperforms, consumes less power, and runs cooler than its Core i9 rivals.
- The 16C/32T arrangement packs a lot of power.
- It’s compatible with Threadripper motherboards that are already on the market.
- Ryzen Master is a simple-to-use software application.
Cons
- In ST workloads, it doesn’t always capture coffee lake.
- For optimal performance, quad-channel RAM is required.
Performer of the Year
Test Result
June 2020 Test Result 9.8/10 Excellent
Cores
16 CPU cores (3.5 GHz up to 4.7 GHz Max Turbo )
Performance of the Workstation (Applications)
Pros
- On the market, the fastest prosumer CPU
- Most Socket AM4 motherboards are compatible.
- Power efficiency that is unmatched.
- PCIe Gen 4.0 native
Cons
- In ST workloads, it doesn’t always capture coffee lake.
Second Position
Model
AMD Ryzen Threadripper 2950X
Test Result
June 2020 Test Result 9.6/10 Very Good
Cores
six cores (3.8 GHz up to 4.4 GHz Max Turbo )
Performance of the Workstation (Applications)
Pros
- It outperforms, consumes less power, and runs cooler than its Core i9 rivals.
- The 16C/32T arrangement packs a lot of power.
- It’s compatible with Threadripper motherboards that are already on the market.
- Ryzen Master is a simple-to-use software application.
Cons
- For optimal performance, quad-channel RAM is required.
The AMD Ryzen 9 are the upper-middle-class overlords.
The Ryzen 9, two variants with presently 12 and 16 cores, are AMD’s greatest CPU for contemporary software in the upper mid-range, giving remarkable performance on a four-year-old infrastructure, the AM4 socket.
Intel’s new mainstream flagship Core i9-10900K is likewise very fast, but with just ten cores and 20 threads, it can’t compete with AMD Ryzen across the board.
Overall, the AMD Ryzen 9 3950X will be the best option for most users.
Especially from the standpoint of performance and price-to-performance ratio.
AMD is said to be releasing a Ryzen 4000 series around the end of the year.
We’ll keep you informed!
The “ryzen 9 vs threadripper for gaming” is a question that has been asked many times. In this article, we will compare the Ryzen 9 3950X to the Threadripper 2950X and determine which one is the best value for your money.
Frequently Asked Questions
Which is better for gaming Ryzen 9 or Threadripper?
Is Threadripper 2950X good for gaming?
A: Threadripper 2950X is a high performance LGA-2066 processor. It has 8 cores and 16 threads that run at 3.8GHz with boost clock of 4GHz to help you game well on PC or other devices like laptops, etc.
Is the Ryzen 9 3950X worth it?
A: The Ryzen 9 3950X is a high-end CPU with many features that are designed to make it an attractive option for video game developers and experienced users. Its architecture, specifications, and price point all put it in the top tier of AMDs offerings. If youre looking for something powerful enough to handle any task thrown at it without breaking too much of your bank account, then this might be what you need.
Related Tags
- threadripper 3990x vs ryzen 9 3950x
- ryzen 9 vs threadripper 3
- threadripper vs ryzen 9 5950x
- ryzen 9 3900xt vs threadripper 1950x
- 3950x vs 2950x
