So you're telling me that this
www.overclockers.co.uk/showproduct.php?prodid=HD-342-WD&groupid=701&catid=14&subcat=940 is no faster than it's sata2 counterpart?
Reviews tell me that drives like these should be getting massive burst rate values.
They are the same speed and EXACTLY the same inside.
Have you turned on the relevant caches on the drive? Only really Intel controllers get big burst rates.
Rubbish, absolute rubbish.
I'm only concerned about the burst rate which, as benchmarks show, should be a lot faster.
No, it shouldn't, burst speed is the speed the CACHE on the drive, and ONLY the cache can empty its data across the bus.
The problem is, the cache is depending on the drive 16-64mb, the problem is, once data is in the cache sitting there, it can be emptied incredibly quickly.
The burst speed that a few incredibly old benchmarks give, do NOT measure constant burst speeds, its the very peak of the highest piece of burst speed data. That means at 1 second throughout the benchmark the cache is full, manages to empty itself across the cable in 0.1seconds and the benchmark records that as 100mb's a second.
The cache is filled up by reading data off the platters, the platters are not faster, the transfer of data from the platter to the cache is no faster, the cache doesn't magically fill itself and send data faster, it waits on data, if it flushes all 64mb of data instantly, it will simply have to wait for the drive to fill it up before it can stream data again.
What I'm trying to say is, burst speed is nothing, its not any measure of performance, it literally means NOTHING. One drive could read 1mb burst and another could measure 1billion mb/s, it wouldn't tell you anything at all. There has never been a less useless piece of data for comparing drive speeds.
Its hard to explain because I can't think of a decent example to compare it to. Ok, its like having a car that does 100mph max, it drives 1000 miles averages 80mph due to speeding up and slowing down at points, it can never go faster than that, except every now and then as it speeds up the tyres spin, for a split second, and when the tyres spin unimpeded by contact they spin at 1000mph, but because it happens once every 100 miles, for 0.01seconds, it makes no difference. Thats burst speed, it makes absolutely and unequivically no difference.
There is exactly no difference in performance between a Sata 2 and Sata 3 drive with the same model and internals and they will never show a performance difference. Once the cache is empty it can ONLY fill the cache as fast as the platter can turn data, and as its filling with data it can again only empty it at a speed it receives the data, the platters on both drives provide this data at an identical speed.
Even your own benchmarks show something, the burst rate went up, it was capable of providing almost double the burst performance and yet your average speed basically didn't go up, your access speed and lowest data rate were significantly worse and overall performance would have been indistinguishable because hdd performance is almost impossible to replicate one benchmark to the next, each of your benchmarks has been noticeably different each time.
As for burst rates, on a decent motherboard controller most have been able to get embarassingly high burst speeds for 5 years, massively ahead of the 175mb's you got.
Also yes, you've been conned, the drive is no faster and the pci-e card is completely worthless for that drive. The, and I mean this quite literally, the ONLY drive its worth getting a Sata 3 card for is the Crucial C300 SSD at the moment, or the other rate situation that you've got so many drives all working hard constantly on the motherboard controller its maxed out performance wise and you would get more bandwidth when using another controller.
