• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

New Seagate drive, SATA II

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
I was also interested in this drive, and am thinking about purchasing, although i wouldn't be of much help since i'm upgrading from a IDE 100 40gb 2mb cache piece of poopie
 
Theoretical Max Transfers are never achieved by any single drive & that your PCI BUS is not wide enough to handle even half of what the SATAII standard is theoretically capable of doing. Comparing a 7200rpm drive to a 10,000rpm drive is no contest.
 
Odd, the link indicates SATA II, but the interface rate is is still listed as SATA 150. It does have NCQ, but performance will still be significantly lower than either Raptor. STRs are listed from 32 to 58MB/s and combining seek times with latency you get disk service times around 12.6ms. Compare this to the 36GB Raptor, with similar STRs and disk access times of 8.6ms and it's "no contest." Going to second generation 74GB Raptors gets STRs between 54 and 72MB/s and disk access down to 8.1ms. the warranty is nice though.
 
Its Native Command Queueing. I was reading a little about it. From my understanding, its only supported by the new Intel chipsets. So I don't think there would really be any gain from me using it. :(
 
That review is of the ST3160023AS, not ST3160027AS. But, if we can find one of the 27, or someone posts benchmarks of their own on it, then we'll have something great to compare it to.
 
you're right, my bad. hehe. but i think while the numbers aren't exact it should give a good basis of comparison vs. a standard HDD
 
Are there even any motherboards out that have Serial ATA II on them? otherwise it is pointless to stick this generation II on a generation I serial ATA connection

Serial ATA II compatible chipsets will be able to overcome the PCI bandwidth issues i would assume.
 
As long as the controller is planar mounted(nonPCI attached), it will have dedicated bandwidth to the southbridge. This eliminates the 133MB/s limitation on the PCI bus. None of the boards I've seen available have SATA II yet. As mentioned previously NCQ is being implemented by new Intel chipsets. This will allow a bit more efficiency in the disk subsystem. As for the viability of going to a 300MB/s bus with a single drive attached, it will give excellent burst rates, but short of a RAM drive, it will have little practical effect.
 
They need to add larger cache to these drives. I see a lot of bandwidth going to waste with SATA1 let alone 2. We need to have drives with larger cache buffers, with perhaps even battery backup. 128mb cache per drive, and now you will see performance. I would expect this to hand about 32 dollars to cost of drive, but retail would be about 60 more, which I would pay.
 
darthdana said:
They need to add larger cache to these drives. I see a lot of bandwidth going to waste with SATA1 let alone 2. We need to have drives with larger cache buffers, with perhaps even battery backup. 128mb cache per drive, and now you will see performance. I would expect this to hand about 32 dollars to cost of drive, but retail would be about 60 more, which I would pay.

While that would be good for sequential reads(RAID-0 or single drive), it wouldn't have an appreciable effect for usage on disks with server or OS disk access patterns. This would really limit the market for these drives and firmware optimizations would need to be specifically developed for the desired application. A better use would be DDR RAM disks or other high bandwidth and low latency solid state storage.
 
Xaotic said:
While that would be good for sequential reads(RAID-0 or single drive), it wouldn't have an appreciable effect for usage on disks with server or OS disk access patterns. This would really limit the market for these drives and firmware optimizations would need to be specifically developed for the desired application. A better use would be DDR RAM disks or other high bandwidth and low latency solid state storage.

Actually if you notice the part about the battery, I was more concerned with writes. Wouldnt need to have battery backup for cached reads now would you. And if you look at any caching raid controller, or even a EMC SAN with 32GB cache, any positive hits in the cache are better than waiting for the drive's moving parts hardware. Heck my 3ware cards average about 40 percent hit on the cache on reads (5 cards on a variety of machines), not bad, take same tech, drop it drive level. Heck, leave the algorithm in flash, and let the OS'es/filesystem update the code on the fly for whatever purpose it needs.
 
I agree with those proposed improvements, but it looks like a nightmare from current cost and development standpoints. That said, I'd happily pay for it for my usage.
 
Back