Hey everyone,
Just had an academic/theory question for you guys. I have a few E6400 XFRs which I use for work which have always driven me insane with how slow they ran- you know how it is in the shop- the last thing to get worked on is the tech's. :lol:
I hope the length of this post doesn't scare everyone off.
I had always suspected the HDD as the bottleneck on both systems but I had never done anything to really verify it (aside from looking at the specs of the HDDs installed- both were WB Blues sporting an embarrassing 5400 RPM and 8 MB cache). Tonight, I finally did verify it- under the typical working application load my computer sees (and where I usually have trouble)- I ran Moo0's System Monitor which identifies the system's bottleneck (if present) at any point in time.
The results were interesting (at least I thought they were)- on my higher end E6400 XFR- the one with the "newer" Core 2 Duo P9700 and separate business graphics card, my System Monitor was showing the bottleneck being predominantly on the CPU side (100% CPU utilization that would stay pegged until you ultimately would give up and restart the computer out of frustration). Only rarely would the System Monitor flag the bottleneck as the HDD being the culprit. In contrast, my other E6400 XFR, a lower end one in comparison sporting the slightly older P8700 with integrated graphics and less RAM, actually flagged the HDD as the bottleneck almost constantly- even under similar work loads.
I decided to do a bit of an experiment with my "higher end" E6400 XFR- I wanted to see how performance (a subjective term determined by my own user experience/perception) would change if I put an upgraded HDD in there anyway, even if it supposedly wasn't the bottleneck. Since I had a Momentus XT "solid state hybrid" drive laying around (500 GB- 5400 RPM with 64 MB cache), I cloned the old HDD and dropped the Momentus XT in. What a world of difference. I no longer notice any appreciable lag or the infuriating downtime I used to when the PC seemed to just ignore my inputs. And I really don't think it's a placebo effect- I have the numbers to prove it.
All of this leads me to my question: why?
Obviously the load time for applications is reduced thanks to the SSHD effect, but if the HDD wasn't the bottleneck in the first place, why should it be any faster? I suppose you could argue that how the
What's even more interesting to me is that the CPU bottleneck also went away with the installation of the better hard drive. Whereas CPU usage used to sit at 100% during those periods of ignoring my inputs and otherwise averaged ~50% utilization, it now only very rarely peaks to 80% and tends to sit at 5-20% on average. How could a HDD upgrade eliminate (or at least significantly reduce) a CPU bottleneck? Is it because the seek time is better and the CPU doesn't have to devote processor power to constantly polling a slow HDD waiting for information?
Looking forward to your thoughts and insights,
Techniker
Just had an academic/theory question for you guys. I have a few E6400 XFRs which I use for work which have always driven me insane with how slow they ran- you know how it is in the shop- the last thing to get worked on is the tech's. :lol:
I hope the length of this post doesn't scare everyone off.
I had always suspected the HDD as the bottleneck on both systems but I had never done anything to really verify it (aside from looking at the specs of the HDDs installed- both were WB Blues sporting an embarrassing 5400 RPM and 8 MB cache). Tonight, I finally did verify it- under the typical working application load my computer sees (and where I usually have trouble)- I ran Moo0's System Monitor which identifies the system's bottleneck (if present) at any point in time.
The results were interesting (at least I thought they were)- on my higher end E6400 XFR- the one with the "newer" Core 2 Duo P9700 and separate business graphics card, my System Monitor was showing the bottleneck being predominantly on the CPU side (100% CPU utilization that would stay pegged until you ultimately would give up and restart the computer out of frustration). Only rarely would the System Monitor flag the bottleneck as the HDD being the culprit. In contrast, my other E6400 XFR, a lower end one in comparison sporting the slightly older P8700 with integrated graphics and less RAM, actually flagged the HDD as the bottleneck almost constantly- even under similar work loads.
I decided to do a bit of an experiment with my "higher end" E6400 XFR- I wanted to see how performance (a subjective term determined by my own user experience/perception) would change if I put an upgraded HDD in there anyway, even if it supposedly wasn't the bottleneck. Since I had a Momentus XT "solid state hybrid" drive laying around (500 GB- 5400 RPM with 64 MB cache), I cloned the old HDD and dropped the Momentus XT in. What a world of difference. I no longer notice any appreciable lag or the infuriating downtime I used to when the PC seemed to just ignore my inputs. And I really don't think it's a placebo effect- I have the numbers to prove it.
All of this leads me to my question: why?
Obviously the load time for applications is reduced thanks to the SSHD effect, but if the HDD wasn't the bottleneck in the first place, why should it be any faster? I suppose you could argue that how the
What's even more interesting to me is that the CPU bottleneck also went away with the installation of the better hard drive. Whereas CPU usage used to sit at 100% during those periods of ignoring my inputs and otherwise averaged ~50% utilization, it now only very rarely peaks to 80% and tends to sit at 5-20% on average. How could a HDD upgrade eliminate (or at least significantly reduce) a CPU bottleneck? Is it because the seek time is better and the CPU doesn't have to devote processor power to constantly polling a slow HDD waiting for information?
Looking forward to your thoughts and insights,
Techniker
Facebook
Twitter
Instagram