RAM-Upgrade: Pavilion TP01-2000i (1K4J0AV)

[marmic]

Hey peeps,

I've a question regarding RAM upgrade.

I've a Pavilion TP01-2000i that comes with HP Stark 8860 Mainboard and a single DIMM Samsung 16GB (DDR4-3200) - P/N: M378A2G43AB3-CWE (DDR4-3200 CL22-22-22-52-1).

Due to the single channel usage, it's obvious that i'll gain a performance boost if I switch to dual channel.

Now I'm struggling with the mainboards RAM clock.

The System is based on a "HP Stark 8860" Mainboard (Intel H570 chipset) and the techspecs say there's a limit at 2933 MHz. (click for TechSpecs)

If I now install a 3200MHz dual channel kit, will they run at 2933MHz clock or will they just run with there native SPD / JEDEC clock ?
for example:
1. Corsair Vengeance LPX 2x16 GB (DDR4-3200 CL16) will they run at 2933MHz clock or will they only run with native SPD/ JEDEC clock (2133MHz)? (click for TechSpecs)

2. Kingston Fury Beast 2x16 GB (DDR4-3200 CL16) will they run at 2933MHz clock or will they only run with native SPD/ JEDEC clock (2400MHz)? (click for TechSpecs )

Both DIMM kits work at 1,2V with the "Plug'n Play clock", but would need 1,35V at 3000 or 3200 MHz

The BIOS is extremly restrictive, so it isn't possible to set xmp active or do any other (RAM) configuration.

I hope my question is not to confusing 😉
Thanks in advance

marmic

[banhien]

@marmic 

Its specs

         https://support.hp.com/hk-en/document/c06908156

It has 

8 GB DDR4-2933 SDRAM (1 x 8 GB)

Its motherboard

           https://support.hp.com/hk-en/document/c06638240

which says

• Dual channel memory architecture (One DIMM per channel)
• Two DDR4 UDIMM (288-pin) sockets
• Supports up to PC4-23400 (DDR4-2933)
• Supports 4 GB and 8 GB DDR4 UDIMMs
• Supports up to 4 GB on 32-bit systems

  note:

  32-bit systems cannot address a full 4.0 GB of memory.

• Supports up to 32 GB (unbuffered) with two 16 GB DIMMs on 64-bit systems

  note:

  Maximum memory shown reflects the capability of the hardware and can be limited further in the operating system.

You may be able to install 3200MHz RAM but it will run at 2933MHz

Regards.

[marmic]

Hi @banhien 

---

---

Thank you but I think you got the wrong build.

I won't picking beans here, just setting it right for others with the same system/ question 😉

---

Its specs

         https://support.hp.com/hk-en/document/c06908156

---

My System is not "TP01-0000i (8BG41AV)" but TP01-2000i (1K4J0AV) >>> Support Assistant product page

---

Its motherboard

           https://support.hp.com/hk-en/document/c06638240

---

HP Desktop-PCs - Technische Daten der Hauptplatine Stark | HP® Customer Support

It came with a "HP Stark 8860 A (SMVB)" motherboard with H570 chip set (instead of a HP Baker with H470 chip set like the one you linked), but the RAM specs are identical.

---

It has 

8 GB DDR4-2933 SDRAM (1 x 8 GB)

---

It came equiped with16 GB DDR4-3200 RAM (1x16 GB, R1x8, unbuffered) [Samsung M378A2G43AB3-CWE]

The RAM is clocked @ 1600 MHz powered with 1.2V.

Due to the motherboards identical RAM specs, it should be the same result and it'll end up with 2933MHz dual channel clock.

This should still be quicker / more responsive then a single 1600MHz DIMM

I've done a test at userbenchmark.com (click to view) and tried 5 Games & a CAD-Project with my standard settings.

I'll repeat this test after I've got my Kingston Fury Beast 3200 (2x16GB) kit today and report the difference for future reference.

thanks and regards

marmic

[marmic]

heya, reporting as promised 😉

Got my Kingston Fury dual kit (2x16GB DDR4-3200 CL16)

They work steady and reliable but... only with 2400 MHz.
That's not even close to the 2933 MHz from the motherboards specifications and I can't find anything within the UEFI to change that. 

Anything I can do to get them working with at least the 2933MHz from the specs without buying a new mobo, tower & PSU?
Due to the fact that no other motherboard will fit into the pavilion case and 3 months ago I've payed 1400€ for this PC, I'm a little disapointed.

I realy don't understand why a manufacturer should decide something like this.

[OmenGamer79]

When dealing with HP’s that are locked bios using Intel processors its best to just use samsung ram.

If you have an HP with a Ryzen processor you can then change clock speed and frequecy XMP using Ryzen master.

MSI after burner cannot bipass HP’s locked bios with intel chips. Rendering MSI afterburner useless unless just using the sensors to determine performance and FPS.

I only use SAMSUNG on all my HP’s with intel processors. Bench mark results show the RAM out performs Hynix, Kingston,G Skills, and Corsair.

[marmic]

Hi OmenGamer79,

2 days ago, I've ordered a Samsung DIMM identical to the supplied one. Should arrive on monday (tomorrow).
We'll see how it works 😉
In general I'm really satisfied with my little Pavilion. I've changed the CPU cooler (came with intel stock cooler and worked at 85-88°C with full load) with a Noctua Tower Cooler (now the CPU is working at 65-69°C with full load).

If the Samsung RAM will work wit 2933 in DC, the only "disturbing" fact ist the GPU's temp of 85-88°C while operating at 100%. 
If I got this fixed/ reduced, I will be pleased for the next 2 years I think.

[OmenGamer79]

For what ever reason HP’s and Dells love Samsung Ram.

I have a Dell Vostro 470with Samsung Ram, I have a HP Elite Tower with Samsung Ram, and HP/OMEN laptop with Samsung Ram.

when ever I tried otherRAM sticks including the stock RAM sticks HP and Dell used on the other towers being either Hynix or Kingston, the benchmarks sucked and my systems receive horrible scores of underperform usually under the 40th percentile or lower.

For whatever reason I use Samsung RAM with systems with Intel i7 processors and the scores go up to 65th percentile or higher.

my Omen laptop scores in the 85th percentile and over performs omen laptops with kingston or Gskill.

Has something to do with the intel cpu’s,  and samsung RAM. Use dual channel sockets and it performs phenomenal. 

[marmic]

That sound promissing.

I hope my Pavilion will do the same 😉
I really don't understand why a Manufacturer will cut the customers choice of hardware like this. I mean, If I were up to pay for restrictions, I would buy a rotten apple 😉 It can't be because of reliability, because most RAM DIMMs are working very reliable.

I've done a couple of performance tests with the Kinston RAM (userbenchmark, Forza Horizon, Brakpoint, Photoshop, Fusion360 etc) and will repeat the exakt same test with the Samsung DIMM.

Well, after I received and tested the Samsung RAM today I'll report the results.

thanks and regards

marmic

[marmic · Accepted solution]

well, I tested the Samsung RAM today an again I was surprised..

Expecting a RAM clock of 2933MHz due to the linked specsheet within the product page of my system,

But Windows performance monitor and CPU-Z gave me a full 3200MHz RAM clock. So I think the spec sheet isn't analog to the latest BIOS update.

CAD Tools (Fusion360, PaletteCAD, InnoPlus) as well as Photoshop, Illustrato &  InDesign responding as expected with the Samsung RAM (3194 MHz). A 713MB *.indd print project was loaded ~20% quicker as the Kingston does. The CAD Tools reacting without a delay to every input with both of them, small differences only noticable while importing larger models.

In Games there is a small performance gain, but it depends on the game how much.

In Forza Horizon 4 (4K HDR, ultra preset) I can play with  constant 60fps v-sync and no lags while streaming to discord (kingston was lagging a little).

Below are 2 ingame benchmarks and the result from 3DMark TimeSpy Demo.

All benchmarks are made with a resolution of 3440x1440p (UWQHD) and with both RAMS:

Kingston (2x16GB, R2x8, 2400MHz, CL17)

Samsung M378A2G43AB3-CWE (2x16GB, R1x8, 3200MHz, CL22)

Ghost Recon Breakpoint (Vulkan) (integrated benchmark, average out of 3 runs)

setting: Kingston result / Samsung result

very high: 15845 / 16151 pts

ultra: 13854 / 14034 pts

ultimate: 10768 / 11530 pts

Forza Horizon 4 (integrated benchmark, average out of 3 runs)

setting: Kingston result / Samsung result
Ultra (preset): average 79 / 86 fps

all ultra or highest possible: average 63 / 74 fps
all setting at max possible: average 60 / 71 fps

Futuremark 3DMark TimeSpy

setting: Kingston result / Samsung result

default settings: 9868 / 10053

UserBenchmark.com (average out of 3)
category: Kingston result / Samsung result

Gaming: 114  / 130

Desktop: 99 / 104 

Workstation: 103 / 120 

CPU: 89,3% / 102%

Graphics Card: 126% / 126%

Drive1 (Nvme): 64% / 207%

Drive2 (HDD): 76,9% / 91,1%

Memory Kit (RAM): 84% / 99,6%

I got these alerts from userbenchmark in all 3 runs with the Kingston RAM kit (no such alerts with samsung):

PC Status: Overall this PC is performing below expectations (30th percentile).
CPU: "performing way below expectation (13th percentile)

Drive2 (HDD): Performing below expectations (33rd percentile)

Memory Kit (RAM): Performing below expectations (31st percentile)

I hope that helps someone with a similar issue to decide right insted of buying the wrong RAM kit like me.

regards

marmic

[OmenGamer79]

Yup, sound about right. That's why all MAC pro's are samsung RAM stock.

To get a really good bench mark on the gaming, you can use games that require alot of rendering, such as cyber punk, Shadow of the Tomb Raider, Battle Field 5, and of course Doom Eternal on highest settings. Will require 85 to 90% of CPU usage and then the cpu ultimately resorts to hyper threading and turbo.

What this will do is require more CPU bound processing, and then the processor has to go into turbo, maximizing the RAM.

You can see alot of that action though in CAD where you were seeing faster 20% loadtimes.

Samsung RAM is usually the most sound investment when it comes to RAM and its performance advantages out of the box, with no XMP modifications needed.

Everything on youtube when it comes to corsair, Kingston Fury, Gskills its all commercialized hype and advertising, because out of the box it doesn't perform even to its out of the box latency timings. Samsung is tried and true and meets or exceeds its latency timings. I've been bench marking RAM for a decade, spent alot of wasted money on these bench marks to find out almost every time samsung is the top dog when it comes to affordable RAM that outperforms. I now only use samsung and never buy anything else.