9900k or 3800x

One more parting thought: I do not like spending money on mechanical spinners, and frankly, I don't need that much space these days. So I have a ton of existing 2TB hard drives and I like to put them in a RAID.

Check out Storage Spaces if you haven't. For reliability and repair purposes as well as speed (sustained and random), it makes little sense to use more than RAID1 these days, and Storage Spaces will run mirrors at full cumulative drive speed all day. So you can build a 'pool' out of pairs of mirrors and be done with it, if you like. Further, there's almost no benefit to using 'hardware' RAID controllers for smaller installations and for 'NAS' functionality. The performance isn't there, and something like Storage Spaces (or a number of *nix alternatives) is / are absolutely portable between operating system installs.

At worst, you can check it out pretty quickly with whatever you have available and see if it's something that might work for you.
 
Check out Storage Spaces if you haven't. For reliability and repair purposes as well as speed (sustained and random), it makes little sense to use more than RAID1 these days, and Storage Spaces will run mirrors at full cumulative drive speed all day. So you can build a 'pool' out of pairs of mirrors and be done with it, if you like. Further, there's almost no benefit to using 'hardware' RAID controllers for smaller installations and for 'NAS' functionality. The performance isn't there, and something like Storage Spaces (or a number of *nix alternatives) is / are absolutely portable between operating system installs.

At worst, you can check it out pretty quickly with whatever you have available and see if it's something that might work for you.

I will. This might be especially useful, since I don't boot from mechanical drives and I could switch between Intel and AMD more easily.
 
This thread I started by asking one simple question has so much info and opinions.(not really a simple question when you think about it)
In the end, being more plug and play is what won for me. I experienced issues in games so I returned the 3900x and went Intel.
No regrets either way. I mean I want AMD to do well. we all win when that happens. I saved $150 going to Intel I9 from the 3900x + Gigabyte Master MB so that was a bonus.
5GHz was too easy on my 9900k
 
Regarding content creation Dan_D

Some creators will go for >8 cores. Typically these are people who do a lot of rendering.
As of right now, even if you do a lot of rendering one of the fast 8 core machines will fit a good number of cases. Puget typically recommend the 9900k, that says something. Now that you can relatively cheaply stuff one with 64 gig of ram (or even 128 if you want to go down that road), those who edit 4K or do high end photo work are super happy.

I mean, it’s 8 cores, 10 years ago 4 cores was the norm, 10 years from now it will probably be 16-32 cores and things will be different again

Regarding streaming, while you will get marginally better output from a cpu encode, Nvidia’s nvenc has improved a lot over the last 2-3 years and you would be hard pressed to see the difference (IMO) with twitch streaming and boy is it fast. I now use this for a lot of my non streaming video encodes. Heck QSV has improved a fair bit too and this comes “free” with a lot of intel processors. This makes more cores of marginal use in this particular use case.
 
Last edited:
This makes more cores of marginal use in this particular use case.

Pretty much. There are reasons to prefer something else, but using transcoding blocks on GPUs (wherever they are) is the most prompt way to get it done.

And if one cares significantly about streaming, they wouldn't be doing the encode on the same box that they're playing on- not because it couldn't 'handle it', but because of the significant I/O overhead that raises the spectre of introducing performance inconsistencies. And they can get better quality using a second box.
 
No it's not - This is the point I'm trying to get across, and have been trying to for about 2-3 pages.

The 8700k is a more mature platform, the bugs have generally been worked out of the system. The 3600 is still effectively in the "teething" stage. There are issues, they are being addressed, but they're not there yet.

It may take Nvidia a month or two to sort out their issues, and that may mean more "random WHEA errors".

AGESA fixes for destiny and Linux may come out soon, or be a few weeks/months out, no one in the public knows.

On that note, I pity the Linux users who migrated on week 1 of release who that bug affected.

I cannot deal with unknowns about when things will be fixed, time is money for me, and time to fix means less money for me. Whether this matters to you is another thing.

You make it sound like Ryzen 3000 is unusable....
 
Agree with the general consensus. For OP's use case, 9900k is best.

Yeah, he could save a little money with a 3800X or 3700X, but IMHO it's not worth it today. Ask me again in a couple of months.

OTOH if you're looking for rendering/productivity, the calculus is different. I just recommended a 3700X to a friend of mine who does boatloads of rendering, and very little gaming, and wanted to build a budget rendering workstation with near-HEDT performance on the cheap (as in so cheap, the 9900k and 3900X were entirely off the table). For that, it's Zen 2 or go home, basically.

But again, for a 70% gaming mixture and you have the money? 9900k all day.
 
And if one cares significantly about streaming, they wouldn't be doing the encode on the same box that they're playing on- not because it couldn't 'handle it', but because of the significant I/O overhead that raises the spectre of introducing performance inconsistencies.
And this appears to be an area where Zen 2 beats Intel. Unfortunately, there aren't yet many streaming tests. And most of them just report the framerates on the streamer's game. But, there are a couple of testers who reported on the actual experience for the stream and the streamer. And 8 core Zen 2 seems to be better than 8 core Intel.
 
And this appears to be an area where Zen 2 beats Intel. Unfortunately, there aren't yet many streaming tests. And most of them just report the framerates on the streamer's game. But, there are a couple of testers who reported on the actual experience for the stream and the streamer. And 8 core Zen 2 seems to be better than 8 core Intel.

Which ones saw that, the ones that I watched were the exact opposite. That includes frame times.
 
And this appears to be an area where Zen 2 beats Intel. Unfortunately, there aren't yet many streaming tests. And most of them just report the framerates on the streamer's game. But, there are a couple of testers who reported on the actual experience for the stream and the streamer. And 8 core Zen 2 seems to be better than 8 core Intel.

If you're trying to do software encoding on the same box that you're playing on- well, the point is that that's pretty stupid, and represents a narrow use case. Either use hardware to keep the I/O from affecting gameplay and deal with the quality limitations or use an external encoder / separate box and do it right.
 
Which ones saw that, the ones that I watched were the exact opposite. That includes frame times.
Another, which shows massive favor to AMD


This one is a tighter race, but once he starts looking at frame times and dropped frames, AMD is the winner. (ah this one doesn't have the 8 core Zen 2. Just the 3900x)
 
I should have been more specific "And 8 core Zen 2 seems to be better than 8 core Intel"



Not really. Even watching your video's I fail to see how you came up with that conclusion.
 
I should have been more specific "And 8 core Zen 2 seems to be better than 8 core Intel"



Not really. Even watching your video's I fail to see how you came up with that conclusion.
More dropped frames and worse frame times during streams, according to these videos.
 
Buy the AMD, be done with it and GET OUT OF HERE. Simple as that.

Never fails......since 2005.

Pentium D 950 or X2 3800+.


This is a good thing. It's good to see these debates again. We haven't for a long, long time. I like it when AMD and Intel go to war. It's good for us.

PS: I still have my X2-3800 at 2.7GHZ. That was an awesome chip! In fact... I think I might just boot it up for fun again.
 
Some people care about resale value of their CPU and some don't. Intel in my experience holds better value. Even the lower end Intel's hold value well. I always sell my old computer parts and last month parted out my wife's computer since she no longer needs it for work. Bought the I7 7700k for $310 new and just sold it for $289 after using it for just over 2 years. At the time I was considering the 1700x. The 1700x are selling for about $160 or less now... So in the end it would have cost me more to go AMD between needing a GPU and bad resale. Intel almost always has a better platform. For whatever reason AMD has historically had very little interest in making chipsets for their own CPU. I was told that in person back in 2000 or 2001 when I took a tour of the Austin HQ.
 
If you're trying to do software encoding on the same box that you're playing on- well, the point is that that's pretty stupid, and represents a narrow use case. Either use hardware to keep the I/O from affecting gameplay and deal with the quality limitations or use an external encoder / separate box and do it right.

Eh, if I did it, I'd just use this same box. But then again, I'm not buying high core count CPUs just for something like this either.
 
Some people care about resale value of their CPU and some don't. Intel in my experience holds better value. Even the lower end Intel's hold value well. I always sell my old computer parts and last month parted out my wife's computer since she no longer needs it for work. Bought the I7 7700k for $310 new and just sold it for $289 after using it for just over 2 years. At the time I was considering the 1700x. The 1700x are selling for about $160 or less now... So in the end it would have cost me more to go AMD between needing a GPU and bad resale. Intel almost always has a better platform. For whatever reason AMD has historically had very little interest in making chipsets for their own CPU. I was told that in person back in 2000 or 2001 when I took a tour of the Austin HQ.

There's a reason for this.

With Intel, you switch chipsets/platforms every two generations or so. AMD holds a platform longer. Now as Dan has pointed out, this has both benefits and drawbacks. Part of AMD's compatibility problems probably stem from the fact that they hold onto a platform longer than they should. I wonder if some of Zen 2's quirks stem from how hard it was to maintain AM4 socket compatibility (AMD themselves said this was difficult). But it has benefits too. I can drop a 3900X in my X370 board and enjoy far more performance than when Zen was originally released for cheap, and minimal effort/no platform cost.

So what happens is, if you want to hold on to your Intel build longer, but still want to swap out CPUs, you're going to try to find the fastest thing on that platform. For instance, the 7700k goes for obscene money on ebay (relative to what it probably should go for) because it was the fastest CPU for its platform. The 9900k is likely to be the same someday.

This, however, incurs a cost for you as well. When it comes time to upgrade, the option to just drop in a new CPU 3 or 4 generations newer isn't there. You have to upgrade the whole platform. So you're not really saving anything when it comes to holistic net cost. Yeah, you sell your existing CPU (if it's one of the best for that platform) for more than you might have expected otherwise, but you're locked into swapping out the whole platform too.

Neither option is bad in this respect. Both have benefits/drawbacks. Choose what makes sense to you. (for OP I continue to recommend a 9900k).
 
There's a reason for this.

With Intel, you switch chipsets/platforms every two generations or so. AMD holds a platform longer. Now as Dan has pointed out, this has both benefits and drawbacks. Part of AMD's compatibility problems probably stem from the fact that they hold onto a platform longer than they should. I wonder if some of Zen 2's quirks stem from how hard it was to maintain AM4 socket compatibility (AMD themselves said this was difficult). But it has benefits too. I can drop a 3900X in my X370 board and enjoy far more performance than when Zen was originally released for cheap, and minimal effort/no platform cost.

So what happens is, if you want to hold on to your Intel build longer, but still want to swap out CPUs, you're going to try to find the fastest thing on that platform. For instance, the 7700k goes for obscene money on ebay (relative to what it probably should go for) because it was the fastest CPU for its platform. The 9900k is likely to be the same someday.

This, however, incurs a cost for you as well. When it comes time to upgrade, the option to just drop in a new CPU 3 or 4 generations newer isn't there. You have to upgrade the whole platform. So you're not really saving anything when it comes to holistic net cost. Yeah, you sell your existing CPU (if it's one of the best for that platform) for more than you might have expected otherwise, but you're locked into swapping out the whole platform too.

Neither option is bad in this respect. Both have benefits/drawbacks. Choose what makes sense to you. (for OP I continue to recommend a 9900k).

What you're saying makes sense and is true but I think there is more to it. Because even the I5's hold value well from what I can see. Lets include the fact that Intel CPU's are rarely discounted even when replaced by next gen and are more sought after (higher demand). Plus most Intel CPUs include an iGPU. So maybe people are buying I5's for virtual machines?
 
What you're saying makes sense and is true but I think there is more to it. Because even the I5's hold value well from what I can see. Lets include the fact that Intel CPU's are rarely discounted even when replaced by next gen and are more sought after (higher demand). Plus most Intel CPUs include an iGPU. So maybe people are buying I5's for virtual machines?

Not really seeing that. Take Kaby Lake. 7700k is going for obscene prices (~$280 - $300). The contemporary 1800X goes for way less (~$170), which is more sensible.

Compare to contemporary Ryzen 1600. It goes for similar prices as the 7600k - maybe a tad less - which is more or less what we would expect given market positioning at the time both CPUs were released. So the 7600k held value about how we should expect, but the 7700k is way out of alignment.
 
There's a reason for this.

With Intel, you switch chipsets/platforms every two generations or so. AMD holds a platform longer. Now as Dan has pointed out, this has both benefits and drawbacks. Part of AMD's compatibility problems probably stem from the fact that they hold onto a platform longer than they should. I wonder if some of Zen 2's quirks stem from how hard it was to maintain AM4 socket compatibility (AMD themselves said this was difficult). But it has benefits too. I can drop a 3900X in my X370 board and enjoy far more performance than when Zen was originally released for cheap, and minimal effort/no platform cost.

So what happens is, if you want to hold on to your Intel build longer, but still want to swap out CPUs, you're going to try to find the fastest thing on that platform. For instance, the 7700k goes for obscene money on ebay (relative to what it probably should go for) because it was the fastest CPU for its platform. The 9900k is likely to be the same someday.

This, however, incurs a cost for you as well. When it comes time to upgrade, the option to just drop in a new CPU 3 or 4 generations newer isn't there. You have to upgrade the whole platform. So you're not really saving anything when it comes to holistic net cost. Yeah, you sell your existing CPU (if it's one of the best for that platform) for more than you might have expected otherwise, but you're locked into swapping out the whole platform too.

Neither option is bad in this respect. Both have benefits/drawbacks. Choose what makes sense to you. (for OP I continue to recommend a 9900k).

Well said. I'd also like to point out that Intel tried greater socket longevity with LGA 775 and it caused a ton of problems. Many of the same problems in fact, that we see with AM3 / AM3+ and AM4. LGA 775 was probably Intel's longest lived socket lasting a total of about 7 years. Compatibility with LGA 775 based systems was equally messy with some motherboards using intermediate chipsets being capable of supporting Pentium 4, Pentium D and Core 2 Duo CPU's. When the quad core CPU's came out, you had motherboards that could support Core 2 Duo CPU's but were less than ideal for Core 2 Quad. The 680i SLI motherboards were known for this. You also had Core 2's based on 65nm and 45nm lithography. That too caused issues as some motherboards couldn't do both. Again, the afore mentioned NVIDIA chipset based boards were known for this issue. The 700 series could do both, but the 600 series couldn't. The familiar problem of having to update your BIOS before installing the new CPU was also present in these days.

Basically every problem we see with AMD's socket longevity was present back in the LGA 775 days. This isn't an issue specific to AMD, but rather one that goes with the territory when you commit to backwards compatibility with computer hardware for too long. Even going one generation back brings many of these problems into the equation. Intel simply doesn't deal with it anymore. The good thing is Intel has kept the thermal solutions the same for a really long time. This has been helpful in ensuring water block compatibility as well as the ability to continually reuse the larger and more expensive air coolers. In fact, longevity of the thermal solutions actually goes pretty much all the way back to LGA 1156, with many of those solutions being capable of working on LGA 775. Intel has at least maintained thermal solution compatibility far longer than AMD has. AMD changed it's thermal solution when moving to AM4. There are some coolers and mounting plates that have bolt / screw patterns for both AM3 and AM4, and sometimes you can modify the former to work with the latter. Of course, socket TR4 is its own beast and given the shape of the CPU, its easy to excuse AMD for that. The last CPU that had a similar shape and size (Pentium Pro) needed no such thermal solutions to operate.
 
Not really seeing that. Take Kaby Lake. 7700k is going for obscene prices (~$280 - $300). The contemporary 1800X goes for way less (~$170), which is more sensible.

Compare to contemporary Ryzen 1600. It goes for similar prices as the 7600k - maybe a tad less - which is more or less what we would expect given market positioning at the time both CPUs were released. So the 7600k held value about how we should expect, but the 7700k is way out of alignment.

Skimming through watchcount looks like Ryzen 1600 go from about $60 to $100 and I5 7600k from about $100 to $189 (WTF?lol) with most 7600k above $145 and most 1600 under $95.
 
Skimming through watchcount looks like Ryzen 1600 go from about $60 to $100 and I5 7600k from about $100 to $189 (WTF?lol) with most 7600k above $145 and most 1600 under $95.

Fair. I withdraw my point re: 7600k prices.

Overall point about platform cost vs. resale remains.
 
I game at 4K / 60Hz too. All I play are shooters for the most part. The key is consistency in frame rates, input lag, and so on. I don't believe your a better player at 144Hz vs. 60Hz. I think there is a slight potential advantage using a faster display, but I think this is often over stated.

Back in the day I consistently beat my friends that were on Intel 386 on LAN parties playing DOOM2 on my 486. When we swap seats I consistently lost ALL matches.
The frames matter. A lot.

Whether there is a limit at which it doesn’t matter, please let me know if you find out.
 
Back in the day I consistently beat my friends that were on Intel 386 on LAN parties playing DOOM2 on my 486. When we swap seats I consistently lost ALL matches.
The frames matter. A lot.

Whether there is a limit at which it doesn’t matter, please let me know if you find out.

Well obviously there is a threshold were your frame rates are too low, even if they are consistent.
 
Well obviously there is a threshold were your frame rates are too low, even if they are consistent.

Don’t get me wrong, I’d love a 4K screen, but my entertainment budget can’t stretch to 2080ti :)

My ultimate screen would be 4 to 5K, but hiDPI. I occasionally use DSR to prettify my 1440p, for now.

Just saying.
 
My 2 cent.
Amd has better I/O, better multithread, security.
At a cost of ever so slight reduction in singlethread.
and higher latency. - This is masked 96% by huge L3 cache.

I find the 9900K difficult to recommend as it's pricey for "little", the platform sucks in comparison with I/O etc, the 9700K is a different story altogether and a much smaller investment as well and I'll gladly recommend it to hardcore gamers with high end gpu's.

I may weight I/O too much but it's really important if anyone with a haswell have tried VR and having usb hdd and M2 drives etc and getting the dreaded not enough USB bandwidth, systems freezing for seconds at a time being choked to death because of I/O.

So questions in the end is, is I/O important and how long do you expect to have it where both reasons = more to amd (keeping longer term).
 
Last edited:
My 2 cent.
Amd has better I/O, better multithread, security.
At a cost of ever so slight reduction in singlethread.
and higher latency. - This is masked 96% by huge L3 cache.

I find the 9900K difficult to recommend as it's pricey for "little", the platform sucks in comparison with I/O etc, the 9700K is a different story altogether and a much smaller investment as well and I'll gladly recommend it to hardcore gamers with high end gpu's.

I may weight I/O too much but it's really important if anyone with a haswell have tried VR and having usb hdd and M2 drives etc and getting the dreaded not enough USB bandwidth, systems freezing for seconds at a time being choked to death because of I/O.

So questions in the end is, is I/O important and how long do you expect to have it where both reasons = more to amd (keeping longer term).


I think your a bit off base on the I/O front. If your seriously comparing a modern X570 system to Haswell, then sure. I'll agree. Comparing Z390 vs. X570 is a slightly different story. Keep in mind that USB bandwidth on those older boards is usually constrained by the fact that much of the time, they are using USB hubs on the motherboard to expand the number of ports resulting in more ports, but less bandwidth to go around. Sure, at the bridge to the CPU from the PCH, AMD has a clear advantage right now on paper. When AM4 and TR4 launched, I thought the PCIe lanes dedicated for storage that bypass the PCH would make a huge impact. They don't. In actual testing there is no difference. Sure, AMD has PCIe Gen 4.0, so if you like benchmarking drives and have ponied up the cash for a PCIe Gen 4.0 drive already, then yes, you will see bigger numbers on the AMD side.

But as many people like to point out, when your gaming or talking about the responsiveness of the system, this means nothing. People can't tell the difference between SATA based 2.5" SSD's and the fastest NVMe drives. I've got one of the Corsair MP600 PCIe Gen 4.0 drives on my test bench and it's not any snappier than my Core i9 9900K test system using a PCIe Gen 3.0 Samsung 970 EVO NVMe drive. X399 allows you to use PCIe / M.2 drives directly off the CPU's PCIe lanes. Intel will too, but you've got to pay for a vROC license key to do certain things. The advantage is clearly AMD's. Or is it? As it turns out, not so much. In actual testing between X299 and X399, Intel's just as fast. DMI 3.0 is a PCIe x4 link running Intel's own protocol with virtually zero overhead. Keep in mind AMD's X470 and X399 platforms also had a 4x link with more overhead than Intel. Not only that, but the PCH's on the AMD side also didn't support PCIe Gen 3.0. They were 2.0 ports only. X570 is all Gen 3.0, but with that comes a much higher price tag.

Despite Z390's downlink to the PCH still only having 4x lanes, you have to really work to actually saturate this link and cause performance issues. Even while benchmarking you won't see this come into play unless you pile on multiple benchmarks for different subsystems at the same time in an unrealistic work load. While I can do it, it's not something you would typically be able to do in a real use case. The only times you could, you should be on HEDT and you'd configure the system to use the PCIe lanes from the CPU for your storage or buy the vROC key for RAID support etc. On the USB side, you have dedicated ports to the CPU for AMD. Advantage AMD right? Well not so much. As it turns out, Intel's USB is still slightly faster than AMD's despite this advantage. Straight up USB counts are about equal, but AMD has them allocated differently. AMD has done slightly better here by giving you more of the newer ports, but few people max out USB ports anyway, so this isn't really that big of a deal either way.

When we turn to SATA and RAID support, this looks like another lock for AMD. AMD does support eight SATA devices vs. six, so that one we have to give to AMD. However, on the RAID front, this is decidedly in Intel's favor. Intel has several advantages here. Intel supports RAID 0, 1, 10 and RAID 5. AMD only supports 0, 1, and 10. AMD only supports 64k and 128k stripe sizes. Intel supports 4, 8, 16, 32, 64, and 128k stripe sizes giving you more control over how the array performs for specific workloads. Intel's option ROM for configuring the arrays is vastly superior and much easier to use than AMD's. Intel's iRST software in Windows is nice and easy to use while AMD's 90's style web page that forces you to login to it is hot garbage. Most of the time, the software doesn't even work. I've tested this on countless boards since X370 came out.

The one thing I'll give AMD on the RAID front is the ability to setup a single AHCI volume in a "Raidable" mode so that you can make it part of an array later without changing controller modes and causing the system to BSOD on startup. StoreMI is also great, but I don't know how many people actually use it. I've never bothered on my Threadripper system.

Now, don't get me wrong. I think X570 is the better platform in terms of connectivity and I/O overall. However, it's also more expensive and the real world difference isn't massive to non-existent.
 
Yeah the PCI-E 4 thing is a bit over hyped. Its certainly a PR win for AMD. However, most people will struggle to come up with a workload, which actually utilizes it. A 2080ti only loses a couple of FPS on PCI-E 2.0 and loses zero on 3.0. I've only ever seen 2.0 be a real GPU bottleneck, for multi-gpu setups.

Similarly, 4.0 allows higher potential for storage drives. But, most current drives don't even max out 3.0. I bet even a 4-way raid with decent drives would struggle to max 3.0 for most reads/writes. And it would cost about the same as one 4.0 drive.
 
Here's something that could help.

I know this thread isn't about 3700x, but here's a result of overclocked 3700x ~@4250Mhz in Deus Ex: Mankind Divided compared to 9900k @ 5 Ghz on ultra at 1440p, using overclocked 1080Ti.

Screenshot (55).png
Screenshot (78).png
 
I know this thread isn't about 3700x, but here's a result of overclocked 3700x ~@4250Mhz in Deus Ex: Mankind Divided compared to 9900k @ 5 Ghz on ultra at 1440p, using overclocked 1080Ti.

57.4 / 43.9 * 100 = ~30% increase, and going from mid 40's to high 50's for the minimum framerate is likely to be perceptable for most.
 
It's like I said, sometimes the differences are academic, and sometimes they aren't. Destiny 2 is a prime example of that. Intel provides a perceptibly better experience than AMD does in that game. It's as simple as that.
 
It's like I said, sometimes the differences are academic, and sometimes they aren't. Destiny 2 is a prime example of that. Intel provides a perceptibly better experience than AMD does in that game. It's as simple as that.


Well I know what you'll be going for in your next pc :)
 
Well I know what you'll be going for in your next pc :)

Not necessarily. The chipset driver for Destiny 2 is more of a work around than an actual fix. As I understand it, a fix in the AGESA code is what's really required to do things right. Using the chipset driver work around is still very problematic. It lets you start the game and play it, but there are caveats to doing so. I get crashes under various circumstances which are easily reproduced. Things like alt-tabbing out of the game will cause it to quit responding and so on. I'm willing to bet that the performance of Ryzen 3000 series processors isn't really as bad as it seems right now. If after a time I do not see a real resolution to this issue, then yes, my decision is pretty much made for me.

And while I was in the technical test for Ghost Recon: Breakpoint, I didn't get a chance to test it on Ryzen. I sunk a ton of hours into Wildlands. I enjoyed Breakpoint enough that it may be another title I play quite a bit going forward. I have no idea how Ryzen performs in that game.
 
Not necessarily. The chipset driver for Destiny 2 is more of a work around than an actual fix. As I understand it, a fix in the AGESA code is what's really required to do things right. Using the chipset driver work around is still very problematic. It lets you start the game and play it, but there are caveats to doing so. I get crashes under various circumstances which are easily reproduced. Things like alt-tabbing out of the game will cause it to quit responding and so on. I'm willing to bet that the performance of Ryzen 3000 series processors isn't really as bad as it seems right now. If after a time I do not see a real resolution to this issue, then yes, my decision is pretty much made for me.

And while I was in the technical test for Ghost Recon: Breakpoint, I didn't get a chance to test it on Ryzen. I sunk a ton of hours into Wildlands. I enjoyed Breakpoint enough that it may be another title I play quite a bit going forward. I have no idea how Ryzen performs in that game.


I think you misread me. I think you'll be going intel for your next pc.
 
I sunk a ton of hours into Wildlands. I enjoyed Breakpoint enough that it may be another title I play quite a bit going forward. I have no idea how Ryzen performs in that game.

It works ok, but not great, I tried it with 3700x. Wildlands uses DirectX 11 engine. 1% lows will suffer. I had to overclock it to get my lows back up to my previous 4790k @4.8GHz. One of the reasons why I went for 9900k in the end.

That’s on 1440p mind.
 
Back
Top