• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Resolved.

Status
Not open for further replies.
Seems people have had better luck than me. I can run 5425 as well but the temps get too high for my use. I also run my fans at quiet so while the PC runs fine and there is no issue, I just can't take the fan noise. I guess will settle at 5.3 Gs.
 
KickAssCop said:
Seems people have had better luck than me. I can run 5425 as well but the temps get too high for my use. I also run my fans at quiet so while the PC runs fine and there is no issue, I just can't take the fan noise. I guess will settle at 5.3 Gs.
Click to expand...
FWIW, there does seem to be a good bit of variation in these CPUs as far as OC potential. I watched a youtube video where a guy could do 5500 at 1.23V. Sadly, mine just won't stabilize at 5500.

My CPU stays under 90C (max temp) and averages 75 to 85 under AIDA64 stress testing. I do have an Arctic Freezer II 420, so I've got some cooling (originally bought it for an i9 10940X that drew less power overclocked than a stock 14700K!).
 
Check some of the little things too like disabling IGP, and lowering SOC voltage. My Gigabyte board just defaults to 1.25v SOC and it does just fine at around 1.15v SOC.
 
9800x3d trickling in this morning on Amazon at MSRP. Using the HotStock app (free tier). Been able to finalize like 2 or 3 to test, then cancelled.

Ship date showing 1/14.
 
GotNoRice said:
It always makes good sense to keep an eye on what voltage your CPU is being fed, but 10x Scalar doesn't guarantee that your CPU will see dangerous voltage, it's just one variable.

I'm currently running 10x Scalar with -60 All-core curve optimizer offset (My BIOS lets me set -60, but still not sure if the actual limit is -50). With PBO +200Mhz, set to motherboard limits, and 102BCLK I'm seeing 5534Mhz maintained even during all-core loads at only 1.155v. That's far far less voltage than should be dangerous. Although there are a lot of different opinions on what the max safe voltage is, most people claim numbers in the 1.3-1.4v range, with the most conservative number I've seen thrown out being 1.25v.
Click to expand...

10x scalar targets 1 year of usable life vs. typical 10 years. It's not the voltage but the electrical field across the gate oxide that typically drives this, along with electromigration concerns. Voltage is only one factor in that. I wouldn't ever target over a 2x scalar, and even then I'd just leave that one alone tbh. While the other factors won't generally affect chip reliability, scalar will...
 
bobzdar said:
10x scalar targets 1 year of usable life vs. typical 10 years.
Click to expand...

What does that even mean? Because it's certainly not "typical" for a CPU to die after 10 years. Do you have references for any of this?

bobzdar said:
It's not the voltage but the electrical field across the gate oxide that typically drives this, along with electromigration concerns. Voltage is only one factor in that.
Click to expand...

What factor other than voltage do you think would cause any of that?

My 9800X3D, currently using 10X scalar, usually sits at 1.1v during actual Gaming, and stays under 1.0v during idle (including most web-browsing and office tasks). There is simply no possible way that I'm putting my CPU in any danger at those voltages. Now if we were talking about CPUs that were being kept at 100% load 24/7 folding or mining or whatever, then maybe that would change the equation.
 
GotNoRice said:
What does that even mean? Because it's certainly not "typical" for a CPU to die after 10 years. Do you have references for any of this?



What factor other than voltage do you think would cause any of that?

My 9800X3D, currently using 10X scalar, usually sits at 1.1v during actual Gaming, and stays under 1.0v during idle (including most web-browsing and office tasks). There is simply no possible way that I'm putting my CPU in any danger at those voltages. Now if we were talking about CPUs that were being kept at 100% load 24/7 folding or mining or whatever, then maybe that would change the equation.
Click to expand...
Apparently it affects not only voltage, but amperage as well.

But I'm no expert, I'm just repeating what I've seen here and here by virtue of searching the internet.

I tried the "AMD recommended 7 to 10" scalar for the 9800X3D, ran some benchmarks (most notably Cinebench R23 multicore) and noticed margin of error differences vs 1X at my 5425 1.24V 6000 settings. So, I set it to back to 1X and called it a day.
 
Scalar 1x is working fine. I'm finding no measurable difference in performance or overclocking ability so no reason not to use 1x if it's getting the job done.

However, I would appreciate a link to some information and definitive explanation of scalar reducing CPU life if anyone has one. I don't believe AMD would be recommending a setting that will time out a chip in 1 to 3 years. Everything I can find about Scalar having any specific direct relation to years of CPU life is all internet forum hearsay.

If anyone has links to solid documentation or commentary from an actual AMD employee or something from someone with a strong enough technical background in CPUs to speak with authority I'm all ears to learn something new.

Also, as a note of interest, the scalar setting is not locked behind the "AMD Overclocking" settings that force an acceptance of a disclaimer. You would think that would be absolutely necessary if scalar 10x could reduce chip life to as little as a year.
 
Advil said:
Also, as a note of interest, the scalar setting is not locked behind the "AMD Overclocking" settings that force an acceptance of a disclaimer. You would think that would be absolutely necessary if scalar 10x could reduce chip life to as little as a year.
Click to expand...
Interestingly enough, it is under Advanced, AMD Overclocking on my Asrock X670E Steel Legend, and it does require accepting a disclaimer.
 
Advil said:
Scalar 1x is working fine. I'm finding no measurable difference in performance or overclocking ability so no reason not to use 1x if it's getting the job done.

However, I would appreciate a link to some information and definitive explanation of scalar reducing CPU life if anyone has one. I don't believe AMD would be recommending a setting that will time out a chip in 1 to 3 years. Everything I can find about Scalar having any specific direct relation to years of CPU life is all internet forum hearsay.

If anyone has links to solid documentation or commentary from an actual AMD employee or something from someone with a strong enough technical background in CPUs to speak with authority I'm all ears to learn something new.

Also, as a note of interest, the scalar setting is not locked behind the "AMD Overclocking" settings that force an acceptance of a disclaimer. You would think that would be absolutely necessary if scalar 10x could reduce chip life to as little as a year.
Click to expand...
It is very hard to A/B test using a scalar and have conclusive evidence on degradation. From what I have seen scalar adds volts when CPU is running high clocks increasing temperatures (for me to 96 C) when unchecked. Even without scalar the CPU runs fine so my suggestion would be to keep at Auto or 1X. Also in my testing the added voltage was from 0.1 to 0.15 volts so from 1.2 volts to 1.3 or 1.35 volts. These are just observation of highs in HWINFO. Not sure what are the sustained volts but this was sort of negating the whole PBO undervolt that I was doing. Without scalar and -20 my CPU maxes out at 1.19 volts.
 
GotNoRice said:
What does that even mean? Because it's certainly not "typical" for a CPU to die after 10 years. Do you have references for any of this?



What factor other than voltage do you think would cause any of that?

My 9800X3D, currently using 10X scalar, usually sits at 1.1v during actual Gaming, and stays under 1.0v during idle (including most web-browsing and office tasks). There is simply no possible way that I'm putting my CPU in any danger at those voltages. Now if we were talking about CPUs that were being kept at 100% load 24/7 folding or mining or whatever, then maybe that would change the equation.
Click to expand...

Heat also plays a factor in GoX breakdown, voltage, heat and current for electromigration.

I have a degree in microelectronic engineeering and worked as a reliability engineer for 3 years, so was tasked with testing chips until they failed then determining what caused the failures. We would increase temp and voltage to simulate a chip being in service for years in a much shorter amount of time. We had Early Life Tests (which was basically extended burn-in) and HAST (highly accelerated stress tests). We would test 1000 chips via ELT and then based on the failure rate, extrapolate the 10 year failure rate for those chips - basically if more than 1 of the 1000 chips failed it was considered a reliability issue. With HAST, it was testing a sampling of chips until they failed at +25% voltage and +25% heat, which usually was in the days-weeks time to validate the ELT testing. There was a ton of math, burn in extrapolation based on product maturity etc. that I've forgotten or glossing over. But the scalar factor is very similar to what we had as an acceleration factor based on voltage and temp.

Per AMD, PBO scalar adjusts the FIT/FITness/FailuresInTime limit of your part by that factor. So 10x means it fails 10x faster - or put another way, 10x more failures in the same amount of time. Most consumer chips are specced for 10 years of use at 4-6 hours per day (vs. 10 years 24/7 for server parts), so a 10x scalar means you go from 10 years expected life without failure to 1. I wouldn't touch it - at one point I was an expert in that field but not any longer. Take it for what it's worth.

I should add - to get to 10 years expected life for all parts, they generally would last MUCH longer than that, but when they started to fail it would be on a log scale. So if the target life was 1 year instead of 10, they would hit that life limit at maybe 1.5-2 years and failure rate would start to skyrocket hockey stick style.
 
Last edited:
1000007593.jpg
 
Had one backordered on Amazon but Newegg just popped. Should get it by Dec 10th. Used free tier HotStock app for alert.

Easy peasy.
 
Epos7 said:
Nice. Where did you get it from? Still waiting on my Amazon order from November 12 to ship.
Click to expand...
I ordered from Amazon on November 14th. I read a thread on Reddit where people were getting a shipping confirmation from Amazon with all different order dates. Some were after my order date. Seems random.
 
Safe and Effective said:
I ordered from Amazon on November 14th. I read a thread on Reddit where people were getting a shipping confirmation from Amazon with all different order dates. Some were after my order date. Seems random.
Click to expand...
Probably regional. I ordered mine from Amazon on the 14th and got it on the 29th.
 
arestavo said:
Probably regional. I ordered mine from Amazon on the 14th and got it on the 29th.
Click to expand...

Weird. I'm on the West Coast and we have warehouses all over the place here. Ordered on the 12th and it hasn't shipped yet.

Edit - looks like they just charged my card today so it'll be on its way soon I expect.
 
Epos7 said:
Weird. I'm on the West Coast and we have warehouses all over the place here. Ordered on the 12th and it hasn't shipped yet.

Edit - looks like they just charged my card today so it'll be on its way soon I expect.
Click to expand...
Even more strange. I'm in Idaho. Prime vs not prime?
 
Seeing as Amazon workers are still on strike I believe their shipping times are bad right now. Everything that I have ordered lately from them is being shipped out via economy UPS.
 
So, I did some more overclocking/tweaking with my 9800X3D w/ X870E Nova Wifi.

Using just PBO +200Mhz offset and a -20 Curve Optimizer I'm running at 5.4Ghz at 1.225v and a 75w power draw when gaming.

When I increased the FSB from 100Mhz to 101.5Mhz I got a CPU speed of 5.5Ghz, but it ran at 1.320v and a 96w power draw when gaming. :eek:

So, I think I'm going to settle with 5.4Ghz. The added CPU speed isn't worth having to run at that high of a voltage for me long term.
 
Darkswordz said:
So, I did some more overclocking/tweaking with my 9800X3D w/ X870E Nova Wifi.

Using just PBO +200Mhz offset and a -20 Curve Optimizer I'm running at 5.4Ghz at 1.225v and a 75w power draw when gaming.

When I increased the FSB from 100Mhz to 101.5Mhz I got a CPU speed of 5.5Ghz, but it ran at 1.320v and a 96w power draw when gaming. :eek:

So, I think I'm going to settle with 5.4Ghz. The added CPU speed isn't worth having to run at that high of a voltage for me long term.
Click to expand...

Have you tried larger curve optimizer offsets yet? These CPUs can take a lot more on average compared to previous generations. I started at -30, before working my way downward. Right now my CPU never uses more than 1.166v under gaming and standard all-core workloads. That's with 102 BCLK, boosting to 5534Mhz. The only thing that ever pushes it above 1.2v is all-core AVX workloads, where I will see 1.221v.

What are you using for power measurement? CPU PPT value?
 
Epos7 said:
Weird. I'm on the West Coast and we have warehouses all over the place here. Ordered on the 12th and it hasn't shipped yet.

Edit - looks like they just charged my card today so it'll be on its way soon I expect.
Click to expand...

Still no shipment. I chatted with Amazon and it sounds like they don't fulfill orders in the order they were received. They have to be available in your nearest fulfillment center in order for your order to ship. So I may have to keep waiting until January.
 
Epos7 said:
Still no shipment. I chatted with Amazon and it sounds like they don't fulfill orders in the order they were received. They have to be available in your nearest fulfillment center in order for your order to ship. So I may have to keep waiting until January.
Click to expand...
I guess that makes sense to save on shipping costs. The bit about it having to be at your nearest fulfillment center is BS though. It only needs to be there if they don't feel like shipping it from elsewhere. I just got a 285k for my Z890 build from Amazon a couple days ago and they shipped it from Florida to Chicago. It arrived on 12/5, more or less in the middle of their quoted 11/29-12/9 delivery window. I'm a little baffled as to how it ended up in Florida. Mine is "Made in US with global components", so probably mostly fabbed in Taiwan and assembled at Intel's Foveros facility in New Mexico.
 
zandor said:
I guess that makes sense to save on shipping costs. The bit about it having to be at your nearest fulfillment center is BS though. It only needs to be there if they don't feel like shipping it from elsewhere. I just got a 285k for my Z890 build from Amazon a couple days ago and they shipped it from Florida to Chicago. It arrived on 12/5, more or less in the middle of their quoted 11/29-12/9 delivery window. I'm a little baffled as to how it ended up in Florida. Mine is "Made in US with global components", so probably mostly fabbed in Taiwan and assembled at Intel's Foveros facility in New Mexico.
Click to expand...

Yeah, the markets that are cheapest to ship to probably get fulfilled first when possible. Makes sense, but I probably won't order similar items from Amazon in the future for this exact reason, so I wonder if the pennies they save are worth it.
 
Epos7 said:
Still no shipment. I chatted with Amazon and it sounds like they don't fulfill orders in the order they were received. They have to be available in your nearest fulfillment center in order for your order to ship. So I may have to keep waiting until January.
Click to expand...
I live in the US. I ordered 11/8 and received my chip on 11/23, originally was supposed to be delivered mid January.
 
techie81 said:
I live in the US. I ordered 11/8 and received my chip on 11/23, originally was supposed to be delivered mid January.
Click to expand...
I don't think most people actually expected it to take until January to arrive. For new out of stock items they tend to set pretty conservative delivery estimates. Lets them ship however is most economical I guess.
 
Likely it won't take until Jan but even if it did... Stick it out. I wouldn't want to drop an order at MSRP at the moment. It probably will go up, and certainly won't go down.
 
Anyone air-cooling a 9800x3d, what max temps are you seeing?

Mine hit 96C playing Marvel Rivals.
 
Yes, air-cooled with a peerless assassin. Playing City Skylines II I get into the 60's, idles ~40. Cinebench 2024 was in the 80's (Tctl/Tdie).

Anyone running one of these guys on Windows 10? How is your experience. Windows 11 is buggy ass hell for me (fresh install).
 
GoodBoy said:
Anyone air-cooling a 9800x3d, what max temps are you seeing?

Mine hit 96C playing Marvel Rivals.
Click to expand...
Just built my system this weekend. Saw ~62 while testing with Crysis 3 on an APX120-67. Using -20 CO.

I'm VERY impressed by how cool this thing is compared to my 9900k
 
GoodBoy said:
Anyone air-cooling a 9800x3d, what max temps are you seeing?

Mine hit 96C playing Marvel Rivals.
Click to expand...

pendragon1 said:
with what?
Click to expand...
Noctua D15S Chromax Black with a 120mm fan added. I am using a graphene pad instead of paste, which I used with the last CPU and that one never got over mid 70's (Intel 10900X)
KickAssCop said:
Try PBO -20 at least if not done already.
Click to expand...
Thanks, I will look at that. Edit: You mean the curve optimizer correct?

I had some incorrect sensors selected on my Chassis fan's in Fan Xpert inside Armoury Crate (mobo not yet supported by Fan Control). Re-did the fan curves a little plus corrected the sensor selections, and now playing the same game I am seeing 84.6C after some time playing. So already a good improvement.

My next thought was to go back to thermal paste, but really like using the graphene pads if I can get temps down to mid 70's, I will stick with it.
 
Last edited:
Status
Not open for further replies.
Back
Top