#17289
Posted 03 May 2015 - 10:24 PM
I'm sure I've explained this before, but their performance seems to be a common misconception. In effect, the Core 2 was nothing more than a glorified Mobile Pentium III/Celeron with a dysfunctional cache. To say it was slow at everything is untrue, for reasons I can't be bothered to understand they perform rather well in most games and therefore, the majority of benchmarks which put more weight on such tests, but you can't judge them solely on a single task because by definition, x86 is a general purpose architecture.
The main problems I have are;
> Painfully slow performance. There is a noticeably wait when hot switching applications in Windows, for extra lulz, install a 64-Bit OS and use Google Chrome with more than three tabs open, it does NOT perform well. For some reason putting the CPU in 64-Bit mode makes it worse, it also has problems with 32-Bit code when in this mode, rendering it unreliable and prone to crashes. They also are not very good at compression or encoding related tasks, data streams can halt the core and cache temporarily too, which makes them useless for video editing - which is what I am doing, so it doesn't work for me.
> Heat. At 3.4GHz, my NetBurst Pentium D 940 is still barely exceeding 35°C at full load, in fact, it did not complain when I tested at 4.2GHz... Hardly a processor known to run cool, but this same heatsink cannot be used on a Core 2 E8200 (2.66GHz) because the CPU will quickly pass 50°C and set off the overheat alarm. The Core 2 also uses significantly more electricity - which makes sense, given how much heat it is dissipating. Due to the previously mentioned performance problems, it is therefore a very inefficient processor... But then it is from 1995, so what did we expect really?
> Supporting hardware was terrible; Generally, Core 2 motherboards have several problems. The vast majority seem to be based on P35/ICH9 combos. The P35 is a terrible chipset, it runs hot and performs badly, the ICH9 has possibly the worst RAID controller ever devised - I don't understand how they messed this up, because you'd think it was just the same as the ICH7, but it is not. The ICH9 loses around 100MB/s of throughput from my RAID0 array vs. the old ICH7. It may be the PCI-E on these chipsets at fault though, because some PCI-E cards notice a drop too, unfortunately I can't just drop another RAID card in, because I am using the PCI-E slots for the video capture hardware, which is what requires the array in the first place. By this time, I don't know if Via and SiS were still making chipsets, I have never seen one. I believe there were nForce boards out there but that would just be pointless because they don't work. It seems everyone bought the absolute worst back then too, because if you look for 775 motherboards all you will find are ASUS, Gigabyte, MSI, BioStar and very little else, all for £100 or more, which is a bad joke, I seriously doubt the ASUS P5Q was worth that even when new, let alone today, and anybody that paid that kind of money for such a board is a complete and utter buffoon.
Overall, they are slow and unstable. I have had several and they all yield the same problems, so I can only conclude that they never worked. Some of the problems were confirmed by scanning through documents, such as the errata, which Intel themselves published. It was just a bad time for hardware, the original Phenom was just as poor if not slightly worse because it doesn't last as long.
For reasons I don't understand the Core 2 had, and seems to still have, one of the biggest fanboy movements ever seen by the world, but that suits me fine because those people have always been fun to annoy - and they usually annoy very easily. All one has to do is press Render in Sony Vegas and observe as the supposedly slower NetBurst chip outstrips their piece of junk whilst using considerably less electricity (It's still horribly inefficient in reality)... They get mad and accuse you of cheating. I remember running Crysis too, they used to get mad because they had spent all that money on a new Core 2 box only for my crappy old thing to show up and keep pace with it... Usually on an inferior GPU to the one they had, it was the one game I could catch them up in and I have never known why. It was made worse when I then turned around and had a "meh" attitude to it, saying "I don't game anyway, so whatever." before firing up Vegas, which is something their machine won't even run stably.
What does it matter anyway? It's payday, a Xeon E3 1271V3 is on its way to me and once I've gotten the rest of the gear (A board to plug it into, etc) the Core 2, and even the Pentium D, will look like a Texas Instruments calculator by comparison. I already know this one is good because I specced it years ago, my friend then built one either because he stole my design or because he arrived at similar conclusions when looking for hardware. This proved to be good though as it gave me a chance to see the system running before I started building one myself. His is slightly anemic compared to mine though, or so I say, because he seems to substitute my Kingston parts for Corsair ones, as well as not using the mighty SeaSonic PSU I had specced... In reality those parts are pretty much identical, his PSU looks suspiciously familiar and I'm almost certain that if you pulled the heatsink off the RAM it would be the same ICs, probably even the same PCB, underneath... He does have a marginally slower CPU (1241), but there isn't much in it. Same motherboard, same hard drives, same OS... Probably the same GPU (750 Ti).
0
Help
Duke4.net
DNF #1
Duke 3D #1
High Treason, on 01 May 2015 - 09:10 PM, said:
