Been thinking about upgrading my laptop memory. According to the memory checker CPUID-CPUZ, my laptop is currently running with a CL latency of 21 at 1.20 volts. If I drop down to CL's of 16, say for example using Crucial Ballistic memory, it looks like the voltage on the chips ups to 1.35.

Is this voltage increase due to the lower CL or are there motherboards that run at a higher voltage?

Do I have to stick with only chips that run at 1.2 volts or could I go for better CL timings and run 1.35 volts?

My laptop is a Lenovo Legion series 5 (17IMH05H) with an 10th Gen i7 10750H processor and dual channel DDR4-3200 SODIMM ram.

I know we got some computer experts here on the forum and I'm wondering if you could steer me in the right direction.

Thanks,

Jeff

Just checking, can you remove and replace your processor? Some laptop builders used to solder in parts making them impossible to remove by non-techs.

You have me thinking now Matt. I don't think I have ever seen a laptop with a ZIF socket. Always soldered.

I might be able to do that Matt, but since it's a 10th Gen i7 I don't think I would gain much other than to go to an i9 or 11th Gen.

Jeff

Jeff, although I have built all my computers since 1983, I've never experimented with overclocking or anything comparable, so I can't help with that kind of question. I value stability over speed in digital audio.

Just curious again, even if you can change the CPU, have you looked at a CPU comparison chart to see how much speed you would really get with a different CPU, without changing the motherboard and RAM? I'm guessing less that twice what you have now, or in other words, not worth doing it.

And won't your current setup run Windows 11 ok?

It's possible that lowering the CAS Latency value might require upping the voltage, as the RAM will be working harder and might need a little more juice. Googling that particular Lenovo, it specifies RAM voltage at 1.2v so I'd say you're better off sticking with that - because as Matt said:

Stability is far more important than any minor gain in speed you might get. I have literally never seen any noticeable speed increase in any audio software by tweaking RAM timing or overclocking or anything - it really comes down to amount of ram and speed of storage (SSD or RAID-0).



There are some. I've had older Thinkpads and HP Elitebooks that had ZIF sockets - was quite nice to get a cheap dual-core laptop and be able to toss in a quad-core i7. I haven't seen a laptop with a socketed CPU in years though, I kinda assumed they stopped making them that way.

Appreciate all the help here guys. I'm a tweaker by nature and looking to get the last remaining bit out of this new laptop.

Where I'm at issues with this is that this new laptop, even though running a 10th Gen i7 with an NVME SSD and a separate mechanical HD; it actually runs slower that my 7th Gen i7 desktop. The desktop also has an NVME SSD and mechanical HD. While buss speeds are the same, the desktop has better CL timings, but slower DDR4 RAM.

The desktop also has 32 gig of ram whereas the laptop has 16. With all of my music software the laptop never seems to use more than 25% of the RAM...but why would it be slower??

Got me stumped.

Jeff

Could be a power management setting somewhere. The 10750H benchmarks around 30% higher than the 7700K.

Have you run any benchmarks on it to see where the bottleneck is? Run a CPU benchmark, GPU, and disk speed test on both machines, that might yield the answer.

Could also be preloaded software - many new computers come with a lot of bloat. Could be worth uninstalling some things, or reinstall the OS fresh. That's a bit of a heavy-handed solution though.

Simon, it has been many years since PG Music posted some minimum scores to run BIAB well. What do you recommend using now?

This might make a few of you cringe, but my work desktop has an i5-3550 which has a Passmark score of 4,760, with 16gb ram, 256gb SATA-3 SSD, and the integrated Intel graphics. It runs BIAB and RealBand quite nicely, so I'd say anything above that should be good.

Could go even lower though - I briefly tried BB on a very "lightweight" laptop not too long ago, which had a Intel Core m3-6Y30 sub-1ghz CPU, benchmark is around 1,947 and it seemed fine.

By comparison, Jeff's i7-10750H benchmarks at 12,522 and the 7700K is around 9,708.

Some good suggestions there Simon...back to the studio to try and figure this out!

Thanks!

Jeff

Anyone still using the Geekbench 4 single-core score as a standard?

Simon, I haven't seen Passmark mentioned in a long time. I use to recommend Passmark hardware and software packages to students for use in field testing pc components and I/O ports. Their USB port test can recognize a USB port with a weak power or data connection better than any other test I'm aware of.

No prob Jeff, let us know how it goes!



I don't typically use Geekbench scores, though since I'm a Mac guy I see them often on EveryMac.com which I use as a rough comparison when troubleshooting someone's system. I only look at the Passmark site since it basically lists every CPU from the last - I dunno - 20 years or so? It offers a good comparison.



Interesting device, looks really cool with the USB monitoring. A bit expensive for me though - I bought one of the testers below, which works well enough for me. Gotta be careful with it though, since it's possible to dial up a 4.5amp load which can easily fry a computer USB port - USB2 ports are usually limited to 500ma max and USB3 can be as low as 900ma.