Jump to content

What to set Infinity Fabric clock to?

So I watched this video:

After watching this I was left confused and couldn't really make my own conclusions. 

Right now I have an R7 3700X and Team Dark 3200mhz 14-14-14-31-97 and I currently have it overclocked to 3400mhz 14-14-14-29-43 @1.39v. 

 

My main question is, in my case, what should I set the infinity fabric clock to? For gaming performance what would the best direction be for this kit? 

Link to comment
Share on other sites

Link to post
Share on other sites

The conclusion of the video was basically to just crank the fabric clock as high as it can suffer. You can do some minimal performance testing before/after to check if performance has changed after that.

¯\_(ツ)_/¯

 

 

Desktop:

Intel Core i7-11700K | Noctua NH-D15S chromax.black | ASUS ROG Strix Z590-E Gaming WiFi  | 32 GB G.SKILL TridentZ 3200 MHz | ASUS TUF Gaming RTX 3080 | 1TB Samsung 980 Pro M.2 PCIe 4.0 SSD | 2TB WD Blue M.2 SATA SSD | Seasonic Focus GX-850 Fractal Design Meshify C Windows 10 Pro

 

Laptop:

HP Omen 15 | AMD Ryzen 7 5800H | 16 GB 3200 MHz | Nvidia RTX 3060 | 1 TB WD Black PCIe 3.0 SSD | 512 GB Micron PCIe 3.0 SSD | Windows 11

Link to comment
Share on other sites

Link to post
Share on other sites

One thing that LTT got wrong: the 1:1 ratio (2:1 option available) is between memory frequency (I/O bus clock) and memory controller frequency (UCLK). Infinity fabric clock is automatically set to be equal to UCLK, though it can also be manually set to something else as you please.

 

16 minutes ago, ripper101 said:

My main question is, in my case, what should I set the infinity fabric clock to? 

1700MHz FCLK, because keeping UCLK = FCLK seems to improve performance compared to having FCLK a touch higher than UCLK (let alone lower). However high FCLK itself can also improve performance. Imo if you can get FCLK to be 266MHz or more greater than UCLK, then keep it decoupled (assuming memory kit can't clock any higher while keeping timings at acceptable range). Otherwise, FCLK should always be equal to UCLK. Never sacrifice UCLK to hit high memory frequency (using 2:1 ratio), nothing good will come out of it.

 

16 minutes ago, ripper101 said:

For gaming performance what would the best direction be for this kit? 

What if you disable Gear Down Mode (which is like 1.5T command rate, compared to standard 1T and 2T) and try raise memory frequency higher at the cost of larger memory timings? You should test how high FCLK can go before it gets unstable before hammering the memory to data rate double that of the FCLK.

CPU: i7-2600K 4751MHz 1.44V (software) --> 1.47V at the back of the socket Motherboard: Asrock Z77 Extreme4 (BCLK: 103.3MHz) CPU Cooler: Noctua NH-D15 RAM: Adata XPG 2x8GB DDR3 (XMP: 2133MHz 10-11-11-30 CR2, custom: 2203MHz 10-11-10-26 CR1 tRFC:230 tREFI:14000) GPU: Asus GTX 1070 Dual (Super Jetstream vbios, +70(2025-2088MHz)/+400(8.8Gbps)) SSD: Samsung 840 Pro 256GB (main boot drive), Transcend SSD370 128GB PSU: Seasonic X-660 80+ Gold Case: Antec P110 Silent, 5 intakes 1 exhaust Monitor: AOC G2460PF 1080p 144Hz (150Hz max w/ DP, 121Hz max w/ HDMI) TN panel Keyboard: Logitech G610 Orion (Cherry MX Blue) with SteelSeries Apex M260 keycaps Mouse: BenQ Zowie FK1

 

Model: HP Omen 17 17-an110ca CPU: i7-8750H (0.125V core & cache, 50mV SA undervolt) GPU: GTX 1060 6GB Mobile (+80/+450, 1650MHz~1750MHz 0.78V~0.85V) RAM: 8+8GB DDR4-2400 18-17-17-39 2T Storage: HP EX920 1TB PCIe x4 M.2 SSD + Crucial MX500 1TB 2.5" SATA SSD, 128GB Toshiba PCIe x2 M.2 SSD (KBG30ZMV128G) gone cooking externally, 1TB Seagate 7200RPM 2.5" HDD (ST1000LM049-2GH172) left outside Monitor: 1080p 126Hz IPS G-sync

 

Desktop benching:

Cinebench R15 Single thread:168 Multi-thread: 833 

SuperPi (v1.5 from Techpowerup, PI value output) 16K: 0.100s 1M: 8.255s 32M: 7m 45.93s

Link to comment
Share on other sites

Link to post
Share on other sites

I agree, 1:1:1 MCLK:UCLK:FCLK is your best bet unless you can get drastically higher MCLK. Are you maxed out at 3400mhz? Timings are definitely important, but maxing clock rates is every bit as good. Results at 1800:1800:3600 are great, 1900:1900:3800 are fantastic, even with some looser timings.

 

IMHO with Ryzen in general, crank the clockspeed with speeds synced, then crank down timings as best you can within reason. I wouldn't go much further on voltage though. Some have survived as high as 1.5v without too much issue, but 1.4v is the top of the long term safe voltage for most DDR4 IMHO. 

Link to comment
Share on other sites

Link to post
Share on other sites

22 minutes ago, Jurrunio said:

One thing that LTT got wrong: the 1:1 ratio (2:1 option available) is between memory frequency (I/O bus clock) and memory controller frequency (UCLK). Infinity fabric clock is automatically set to be equal to UCLK, though it can also be manually set to something else as you please.

 

1700MHz FCLK, because keeping UCLK = FCLK seems to improve performance compared to having FCLK a touch higher than UCLK (let alone lower). However high FCLK itself can also improve performance. Imo if you can get FCLK to be 266MHz or more greater than UCLK, then keep it decoupled (assuming memory kit can't clock any higher while keeping timings at acceptable range). Otherwise, FCLK should always be equal to UCLK. Never sacrifice UCLK to hit high memory frequency (using 2:1 ratio), nothing good will come out of it.

 

What if you disable Gear Down Mode (which is like 1.5T command rate, compared to standard 1T and 2T) and try raise memory frequency higher at the cost of larger memory timings? You should test how high FCLK can go before it gets unstable before hammering the memory to data rate double that of the FCLK.

What would be a good frequency to try and achieve? I've seen others get CL14 kits to 3733 MHz 15-15-15-34. Would this be plenty or should I shoot for CL16 at 4000mhz? If I were to go this route would I just leave the FCLK and UCLK to auto?

Link to comment
Share on other sites

Link to post
Share on other sites

8 minutes ago, ripper101 said:

What would be a good frequency to try and achieve? I've seen others get CL14 kits to 3733 MHz 15-15-15-34. Would this be plenty or should I shoot for CL16 at 4000mhz? If I were to go this route would I just leave the FCLK and UCLK to auto?

As I said, you should test how high FCLK can go first. The best sample I've seen on ambient cooling is 1933MHz, while 1900MHz is far more common of a good result. Worst result I've seen is 1733MHz

CPU: i7-2600K 4751MHz 1.44V (software) --> 1.47V at the back of the socket Motherboard: Asrock Z77 Extreme4 (BCLK: 103.3MHz) CPU Cooler: Noctua NH-D15 RAM: Adata XPG 2x8GB DDR3 (XMP: 2133MHz 10-11-11-30 CR2, custom: 2203MHz 10-11-10-26 CR1 tRFC:230 tREFI:14000) GPU: Asus GTX 1070 Dual (Super Jetstream vbios, +70(2025-2088MHz)/+400(8.8Gbps)) SSD: Samsung 840 Pro 256GB (main boot drive), Transcend SSD370 128GB PSU: Seasonic X-660 80+ Gold Case: Antec P110 Silent, 5 intakes 1 exhaust Monitor: AOC G2460PF 1080p 144Hz (150Hz max w/ DP, 121Hz max w/ HDMI) TN panel Keyboard: Logitech G610 Orion (Cherry MX Blue) with SteelSeries Apex M260 keycaps Mouse: BenQ Zowie FK1

 

Model: HP Omen 17 17-an110ca CPU: i7-8750H (0.125V core & cache, 50mV SA undervolt) GPU: GTX 1060 6GB Mobile (+80/+450, 1650MHz~1750MHz 0.78V~0.85V) RAM: 8+8GB DDR4-2400 18-17-17-39 2T Storage: HP EX920 1TB PCIe x4 M.2 SSD + Crucial MX500 1TB 2.5" SATA SSD, 128GB Toshiba PCIe x2 M.2 SSD (KBG30ZMV128G) gone cooking externally, 1TB Seagate 7200RPM 2.5" HDD (ST1000LM049-2GH172) left outside Monitor: 1080p 126Hz IPS G-sync

 

Desktop benching:

Cinebench R15 Single thread:168 Multi-thread: 833 

SuperPi (v1.5 from Techpowerup, PI value output) 16K: 0.100s 1M: 8.255s 32M: 7m 45.93s

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

×