Jump to content

Concerning i9-9900 (non-K) Boost

shahaan
Go to solution Solved by porina,
54 minutes ago, shahaan said:

I tried running with unlimited power but the temps rose up to 98C so I stopped P95 immediately. Through trial and error I have settled on a 140W long boost limit and left the 200W short boost limit at its default. Now with P95 I get 4.3Ghz at 77C which is reasonably acceptable, I believe?

Prime95 is one of the worst case workloads you can put on a CPU so you can get high temperatures with it, more so than other software. That's expected. You can control it with power limit like you have done. For less stressful workloads they're less likely to be limited.

 

54 minutes ago, shahaan said:

Cinebench R15 ... 4.6Ghz at 65C.

Cinebench R20 ... 4.5Ghz at 66C.

 

Both Cinebench tests complete in less than 30-60 seconds. Is this normal or am I doing something wrong?

If it completes drawing the image then it is done and that's normal. The run time will depend on how fast the CPU is. Cinebench R20 is a harder load than R15, so it is not unexpected for clock to be a little lower. That you can hit 4.6 with R15 means turbo is reaching its all core clock.

Hi.

 

My system is stable at a sustained 3.6Ghz boost clock running Prime95 (small FFTs L1/2/3 max. heat, power) while reaching no more than 60-deg.C package temp.

 

I'm using Intel XTU with the following tweaks:

- Turbo Boost Power Max 75W (default 65W)

- Core Voltage Offset -0.070V

- Cache Voltage Offset -0.070V

 

I'll appreciate if someone who has experience with this particular CPU can tell me if this is decent performance from this particular CPU or should I be able to get a higher sustained boost clock? Is there any other XTU setting I should look at tweaking to get a little bit more?

 

NB: I'm not looking for an extreme OC and do not want to push the CPU too far past its power specs, but a 60-deg.C package temp tells me I have room to safely/conservatively push it a little further, maybe??

 

Thanks.

 

System:

i9-9900 (non-K)

ROG Strix Z390-i

16GB DDR4-2666

Silverstone FTZ01 mini-ITX

Noctua NH-L12S

Link to comment
Share on other sites

Link to post
Share on other sites

Does it let you relax the power limit further? Keep going as far as your cooling can keep up with. Increasing the power limit is not considered to be OC by Intel. Most enthusiast builds typically run by default at unlimited power limit.

 

Note Prime95 is an AVX workload and that usually results in lower clocks than non-AVX code. Looking it up, it looks like your CPU is capable of 4.6 GHz all core load if it doesn't trigger other power or current limiters. Try Cinebench R15 for example, but you will probably have to increase power limit.

Gaming system: R7 7800X3D, Asus ROG Strix B650E-F Gaming Wifi, Thermalright Phantom Spirit 120 SE ARGB, Corsair Vengeance 2x 32GB 6000C30, RTX 4070, MSI MPG A850G, Fractal Design North, Samsung 990 Pro 2TB, Acer Predator XB241YU 24" 1440p 144Hz G-Sync + HP LP2475w 24" 1200p 60Hz wide gamut
Productivity system: i9-7980XE, Asus X299 TUF mark 2, Noctua D15, 64GB ram (mixed), RTX 3070, NZXT E850, GameMax Abyss, Samsung 980 Pro 2TB, random 1080p + 720p displays.
Gaming laptop: Lenovo Legion 5, 5800H, RTX 3070, Kingston DDR4 3200C22 2x16GB 2Rx8, Kingston Fury Renegade 1TB + Crucial P1 1TB SSD, 165 Hz IPS 1080p G-Sync Compatible

Link to comment
Share on other sites

Link to post
Share on other sites

33 minutes ago, shahaan said:

Turbo Boost Power Max 75W (default 65W)

Intel quote their TDPs at base clock, if you limit to 65W you won't be boosting. There is no "max power" spec, only max temp.

It's perfectly normal to run unlimited and have 50%+ extra power draw at rated boost clocks.

 

The limit is only there to allow system integrators to compensate for insufficient power supply and cooling solutions.

F@H
Desktop: i9-13900K, ASUS Z790-E, 64GB DDR5-6000 CL36, RTX3080, 2TB MP600 Pro XT, 2TB SX8200Pro, 2x16TB Ironwolf RAID0, Corsair HX1200, Antec Vortex 360 AIO, Thermaltake Versa H25 TG, Samsung 4K curved 49" TV, 23" secondary, Mountain Everest Max

Mobile SFF rig: i9-9900K, Noctua NH-L9i, Asrock Z390 Phantom ITX-AC, 32GB, GTX1070, 2x1TB SX8200Pro RAID0, 2x5TB 2.5" HDD RAID0, Athena 500W Flex (Noctua fan), Custom 4.7l 3D printed case

 

Asus Zenbook UM325UA, Ryzen 7 5700u, 16GB, 1TB, OLED

 

GPD Win 2

Link to comment
Share on other sites

Link to post
Share on other sites

22 hours ago, porina said:

Does it let you relax the power limit further? Keep going as far as your cooling can keep up with. Increasing the power limit is not considered to be OC by Intel. Most enthusiast builds typically run by default at unlimited power limit.

 

Note Prime95 is an AVX workload and that usually results in lower clocks than non-AVX code. Looking it up, it looks like your CPU is capable of 4.6 GHz all core load if it doesn't trigger other power or current limiters. Try Cinebench R15 for example, but you will probably have to increase power limit.

Thanks for this.

 

I tried running with unlimited power but the temps rose up to 98C so I stopped P95 immediately. Through trial and error I have settled on a 140W long boost limit and left the 200W short boost limit at its default. Now with P95 I get 4.3Ghz at 77C which is reasonably acceptable, I believe?

 

Cinebench R15 ... 4.6Ghz at 65C.

Cinebench R20 ... 4.5Ghz at 66C.

 

Both Cinebench tests complete in less than 30-60 seconds. Is this normal or am I doing something wrong?

Link to comment
Share on other sites

Link to post
Share on other sites

98°C is below the max temp and is perfectly fine to run at.

Never understood why people freak out about running their desktop CPUs just below max temp when every powerful laptop (and some not very powerful ones, hello macbook) out there does it all the time.

F@H
Desktop: i9-13900K, ASUS Z790-E, 64GB DDR5-6000 CL36, RTX3080, 2TB MP600 Pro XT, 2TB SX8200Pro, 2x16TB Ironwolf RAID0, Corsair HX1200, Antec Vortex 360 AIO, Thermaltake Versa H25 TG, Samsung 4K curved 49" TV, 23" secondary, Mountain Everest Max

Mobile SFF rig: i9-9900K, Noctua NH-L9i, Asrock Z390 Phantom ITX-AC, 32GB, GTX1070, 2x1TB SX8200Pro RAID0, 2x5TB 2.5" HDD RAID0, Athena 500W Flex (Noctua fan), Custom 4.7l 3D printed case

 

Asus Zenbook UM325UA, Ryzen 7 5700u, 16GB, 1TB, OLED

 

GPD Win 2

Link to comment
Share on other sites

Link to post
Share on other sites

54 minutes ago, shahaan said:

I tried running with unlimited power but the temps rose up to 98C so I stopped P95 immediately. Through trial and error I have settled on a 140W long boost limit and left the 200W short boost limit at its default. Now with P95 I get 4.3Ghz at 77C which is reasonably acceptable, I believe?

Prime95 is one of the worst case workloads you can put on a CPU so you can get high temperatures with it, more so than other software. That's expected. You can control it with power limit like you have done. For less stressful workloads they're less likely to be limited.

 

54 minutes ago, shahaan said:

Cinebench R15 ... 4.6Ghz at 65C.

Cinebench R20 ... 4.5Ghz at 66C.

 

Both Cinebench tests complete in less than 30-60 seconds. Is this normal or am I doing something wrong?

If it completes drawing the image then it is done and that's normal. The run time will depend on how fast the CPU is. Cinebench R20 is a harder load than R15, so it is not unexpected for clock to be a little lower. That you can hit 4.6 with R15 means turbo is reaching its all core clock.

Gaming system: R7 7800X3D, Asus ROG Strix B650E-F Gaming Wifi, Thermalright Phantom Spirit 120 SE ARGB, Corsair Vengeance 2x 32GB 6000C30, RTX 4070, MSI MPG A850G, Fractal Design North, Samsung 990 Pro 2TB, Acer Predator XB241YU 24" 1440p 144Hz G-Sync + HP LP2475w 24" 1200p 60Hz wide gamut
Productivity system: i9-7980XE, Asus X299 TUF mark 2, Noctua D15, 64GB ram (mixed), RTX 3070, NZXT E850, GameMax Abyss, Samsung 980 Pro 2TB, random 1080p + 720p displays.
Gaming laptop: Lenovo Legion 5, 5800H, RTX 3070, Kingston DDR4 3200C22 2x16GB 2Rx8, Kingston Fury Renegade 1TB + Crucial P1 1TB SSD, 165 Hz IPS 1080p G-Sync Compatible

Link to comment
Share on other sites

Link to post
Share on other sites

6 hours ago, Kilrah said:

98°C is below the max temp and is perfectly fine to run at.

Never understood why people freak out about running their desktop CPUs just below max temp when every powerful laptop (and some not very powerful ones, hello macbook) out there does it all the time.

I was under the impression that at T-junction 100C the CPU will auto-shutdown to prevent damage. Apart from being 2C away from that event, the heat sink fan gets loud enough to be disturbing when the temp exceeds the 80C range. Not to mention the hotter it runs, the higher the likelihood of hardware errors and failure.

Link to comment
Share on other sites

Link to post
Share on other sites

8 minutes ago, shahaan said:

I was under the impression that at T-junction 100C the CPU will auto-shutdown to prevent damage. Apart from being 2C away from that event, the heat sink fan gets loud enough to be disturbing when the temp exceeds the 80C range. Not to mention the hotter it runs, the higher the likelihood of hardware errors and failure.

Not sure on the names but there's two temperature thresholds. One is when it enters thermal throttling and it'll downclock to try and lower that. If temps keep increasing then at some higher threshold the system will shut down as a protection mechanism.

Gaming system: R7 7800X3D, Asus ROG Strix B650E-F Gaming Wifi, Thermalright Phantom Spirit 120 SE ARGB, Corsair Vengeance 2x 32GB 6000C30, RTX 4070, MSI MPG A850G, Fractal Design North, Samsung 990 Pro 2TB, Acer Predator XB241YU 24" 1440p 144Hz G-Sync + HP LP2475w 24" 1200p 60Hz wide gamut
Productivity system: i9-7980XE, Asus X299 TUF mark 2, Noctua D15, 64GB ram (mixed), RTX 3070, NZXT E850, GameMax Abyss, Samsung 980 Pro 2TB, random 1080p + 720p displays.
Gaming laptop: Lenovo Legion 5, 5800H, RTX 3070, Kingston DDR4 3200C22 2x16GB 2Rx8, Kingston Fury Renegade 1TB + Crucial P1 1TB SSD, 165 Hz IPS 1080p G-Sync Compatible

Link to comment
Share on other sites

Link to post
Share on other sites

5 hours ago, porina said:

Prime95 is one of the worst case workloads you can put on a CPU so you can get high temperatures with it, more so than other software. That's expected. You can control it with power limit like you have done. For less stressful workloads they're less likely to be limited.

 

If it completes drawing the image then it is done and that's normal. The run time will depend on how fast the CPU is. Cinebench R20 is a harder load than R15, so it is not unexpected for clock to be a little lower. That you can hit 4.6 with R15 means turbo is reaching its all core clock.

Indeed, my daily workload does not remotely approach the beating P95 doles out on the CPU. I was tinkering more out of curiosity and boredom. Thanks for your assistance, I'm getting a steady 4.5Ghz at 77C with power limit set to 140W and the cooler fan running silent. I'm happy with this outcome.

Link to comment
Share on other sites

Link to post
Share on other sites

6 hours ago, shahaan said:

I was under the impression that at T-junction 100C the CPU will auto-shutdown to prevent damage. Apart from being 2C away from that event, the heat sink fan gets loud enough to be disturbing when the temp exceeds the 80C range. Not to mention the hotter it runs, the higher the likelihood of hardware errors and failure.

That's under control of the motherboard, but default behavior is normally to throttle to never let it reach Tjmax. On my new one likely for some stupid "extreme overclocking marketing" reason it wasn't and the first time I load tested it was somehow happy to let the CPU reach 106°C without throttling nor shutting down.

 

Went in the BIOS and found the max temp setting, set it to 98 and now it'll always throttle correctly to never exceed it.

F@H
Desktop: i9-13900K, ASUS Z790-E, 64GB DDR5-6000 CL36, RTX3080, 2TB MP600 Pro XT, 2TB SX8200Pro, 2x16TB Ironwolf RAID0, Corsair HX1200, Antec Vortex 360 AIO, Thermaltake Versa H25 TG, Samsung 4K curved 49" TV, 23" secondary, Mountain Everest Max

Mobile SFF rig: i9-9900K, Noctua NH-L9i, Asrock Z390 Phantom ITX-AC, 32GB, GTX1070, 2x1TB SX8200Pro RAID0, 2x5TB 2.5" HDD RAID0, Athena 500W Flex (Noctua fan), Custom 4.7l 3D printed case

 

Asus Zenbook UM325UA, Ryzen 7 5700u, 16GB, 1TB, OLED

 

GPD Win 2

Link to comment
Share on other sites

Link to post
Share on other sites

19 hours ago, Kilrah said:

That's under control of the motherboard, but default behavior is normally to throttle to never let it reach Tjmax. On my new one likely for some stupid "extreme overclocking marketing" reason it wasn't and the first time I load tested it was somehow happy to let the CPU reach 106°C without throttling nor shutting down.

 

Went in the BIOS and found the max temp setting, set it to 98 and now it'll always throttle correctly to never exceed it.

What CPU is that?

Link to comment
Share on other sites

Link to post
Share on other sites

9900K but shouldn't depend on the CPU, more on the mobo/BIOS settings.

F@H
Desktop: i9-13900K, ASUS Z790-E, 64GB DDR5-6000 CL36, RTX3080, 2TB MP600 Pro XT, 2TB SX8200Pro, 2x16TB Ironwolf RAID0, Corsair HX1200, Antec Vortex 360 AIO, Thermaltake Versa H25 TG, Samsung 4K curved 49" TV, 23" secondary, Mountain Everest Max

Mobile SFF rig: i9-9900K, Noctua NH-L9i, Asrock Z390 Phantom ITX-AC, 32GB, GTX1070, 2x1TB SX8200Pro RAID0, 2x5TB 2.5" HDD RAID0, Athena 500W Flex (Noctua fan), Custom 4.7l 3D printed case

 

Asus Zenbook UM325UA, Ryzen 7 5700u, 16GB, 1TB, OLED

 

GPD Win 2

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×