Still wrestling with performance issues processing ORFs on my machine. The setup: 2.9GHZ Core i7 860 (4 core), 8GB RAM, 256GB SSD, Windows 7. The problem: operations in the Develop module become laggy, anything re-rendering previews results in 3-4 second pauses. My basic test is importing 100 10MP ORFs from the disk, and rendering 1:1 previews of them (other things are harder to measure). What I've tried: Different versions of LR. I've now tried each major released version of LR, from 1 to 5. Aside from a slight improvement between 1.4 and 2.7, I've not seen any difference in performance. Adding an SSD. Good for browsing photos in the catalog. No effect on import/export/develop performance. Different system. A colleague let me try his system for an evening. 3.4GHZ Xeon 5677 (8 cores), 16GB RAM, Windows 7. System was actually slower than mine, albeit marginally. Overclocking. I was able to run my machine at 3.3GHZ long enough to do a test. No real-world difference, although according to my stopwatch import times dropped 4%. Disabling extra cores. While disabling 3 of 4 cores slowed things down, it did not do so by a factor of 4 but by a factor of 2. That suggests to me LR isn't using the extra cores very effectively. Disabling hyperthreading got about a 5% speedup. Using different files. Using 10MP NEFs, I got a 40% speedup in rendering times. 10MP RW2s were similarly faster. Subjectively, develop performance seemed much improved, although still not 'snappy.' Using DNGs converted from the ORFs, I get no speedup at all. I'm not really sure where to go at this point. Dropping LR is the most obvious choice, but performance aside I find it a good tool, and none of the other RAW converters I've tried are as effective for my workflow. In terms of hardware, more cores clearly is not the answer. Faster cores could help, but even though my machine is 3.5 years old, the current generation looks to gain at most 30% in clock speed, and maybe 10% in architectural improvement. Any thoughts?