Precision 7560 & 7760 Owners' Thread

Did the Crucial Ballistix not work for you?

 

Could you post AIDA latency?

 

Crucial Ballistix (2x32GB) seems to have 5% less bandwidth (might be the wrong term) than the Kingston HyperX, but it has 5% lower latency.

LOOK AT MY MORE RECENT POST WHERE I RETESTED RIGHT AFTER A RESTART

 

Are these the 3200CL16 Ballistix SODIMMs? Those scores are comparable to or worse than my hyperx 3200CL20.

 

Yes they are 3200mhz CL16 Crucial Ballistix
Could you post your HyperX scores?

 

Either way the latency is really poor. Clearly there is a bottleneck somewhere in both this and AW x17; and we know it's not the 11the gen platform itself. One hypothesis is that Dell/AW BIOSes force slower command clocks on the modules, hamstringing them severely - resulting in basically token improvements from third party RAM....

 

Yeah, no way to do legacy boot on these new systems.

Never heard of Lenovo whitelisting RAM modules but they do have whitelists for other components (Wi-Fi & GPU cards for example). Dell doesn't actively support upgrades like this but they don't go out of their way to break it either.

With all this chat about RAM latency, that's something but I'm more interested to see how much it translates to a notable performance improvement in actual applications, rather than looking at very specific memory benchmarks. Someone would have to do a specific before/after test. I suspect the difference will be a fair bit less drastic for most applications.

 

Here is my 2x32GB HyperX 3200CL20 results. Benchmark performance is effectively the same between JEDEC and XMP profiles. No surprise, both are CL20-22-22-42

Are you able to confirm the timings with HWINFO or similar? I wonder if the Crucial set is using a 1.2v profile rather than the 1.35v CL16-18-18-36.

 

For faster ram you're paying a premium price for a mostly unnoticeable increase in performance. We're talking about maybe 5 FPS in a shooter that's getting 150'ish. For Illustrator/Photoshop/InDesign - the apps I work in daily - there is no difference whatsoever. It's not quite a scam, but GD close

 

People in x17 forums reported 10% improvements in gaming scenarios and benchmarks, particularly when using memory-intensive Optimus. Basically most applications operating on data sets much larger than the L3 cache size has the potential to benefit greatly. The reason is that in that case memory becomes the bottleneck, i. e. DRAM latency stars to dominate
computation times.