Seagate’s 7200rpm Medalist Pro series has rapidly ascended to the status of the "premiere" ATA drive when it comes to performance. The Maxtor DiamondMax 2880 outperforms the Medalist Pro in many tests, yet doesn’t quite seem to enjoy the same novel reputation as the Seagate drive. IBM’s Deskstar 14GXP promises to combine the spindle speed of the Medalist Pro with the areal density of the DiamondMax, yet still seems to be available to only a handful of OEM system manufacturers. Quantum and Western Digital are also stirring with improved areal densities in their own ATA drive lines, albeit still at 5400rpm. So, Seagate owns the perch. There was just one terrible problem - heat.

The 9.1 gigabyte ST39140A is among the hottest drives we’ve ever tested, ATA or SCSI. Even mounted in a drive cooler, after a bit of use the drive was reminiscent of, well, the Cheetah. Since then, through e-mail comments we received from readers (always appreciated) and from various postings we saw in usenet newsgroups, we came across many reports of the 7200rpm Medalist Pro running quite cool. On the flip side, however, we’ve received messages and seen newgroup postings of the drive indeed running hot and even failing. Though its beyond the Storage Review’s capacity to test a large enough representative sample of Medalist Pros to determine what percentage of drives runs hot, we’ve decided to take the opportunity to test the 6.5GB ST36530A. We’ve received many requests for a set of figures on the drive, and though the smaller unit can be expected to perform quite similarly to its larger brother, we’ve taken the opportunity to add yet another drive to our comparative database.

The spindle speed of the ST36530A is, of course, 7200rpm. Seagate specifies seek time at 9.5ms, while the buffer is a roomy-for-ATA 512k. Data is spread out over 3 individual platters, meaning that areal density calculates out to 2.16 gigs per platter. Note that this comes out a bit differently than either the larger 9.1GB or the smaller 4.5GB Medalist Pros. This may indicate that the data is spread out a little more per square inch; on the other hand, the data may indeed be packed at the same density with the inner most tracks on each platter not being used. I lean towards the latter theory - more on this later.

The drive was an OEM unit (if there’s a retail version available anywhere, we haven’t seen it yet) that, while not arriving in a fancy colored box, was accompanied  by Seagate’s fold-out instruction poster (a bit more intimidating looking than that provided by Western Digital or Maxtor, but also more comprehensive), some screws, and Seagate’s Disc Wizard overlay and data-transfer software.

ZDBop’s WinBench 98 along with Adaptec’s Threadmark 2.0 were both run on the unit in Windows 95 OSR 2.1 and Windows NT Workstation 4.0 The drive was partitioned into a single volume of maximum size. The average of 5 trials is presented below.

As expected, the ST36530A’s performance is similar to that of the 9.1 gig version. Note the higher inner-track ("Disk/Read Transfer Rate – End) transfer rate of the 6530A when compared to that of the 9140A. The 6530A’s faster showing there along with the lower seek time lead us to believe that the inner most tracks of each platter are being left unused. This parallels the situation in Maxtor’s DiamondMax 2880 line, where the jumps between drive sizes are smaller than the amount of data one disk side (1440 megs) holds. Thus, the SR database will list this drive’s areal density as 2.3 gigs per platter, since this most likely is accurate in reflecting the density of the used portion of the platters.

Overall performance of the 6530A when compared to the DiamondMax 2880 5.7gb drive is quite similar, with only Win95’s WinMark tests showing a greater than 5% gap (the nod going to the Medalist pro: 9% faster in Business, 6% faster in the high-end).

Noise wise, the 6530A is identical to the 9140A. There’s no noticeable sound at idle, thanks to the drive’s unique fluid-based bearings. The absence of the metal-on-metal contact of ball bearings eliminates even the slightest high-pitched whine that plagues other drives with high spindle speeds. Seek noise is also similar to the larger MP, a hollow metallic sound not unlike that of the IBM Ultrastar 9LP.

Heat! How hot is the drive? Running in a drive cooler, the unit runs fairly cool, a marked change from the 9140A that we tested earlier. Outside of a cooler, the drive runs quite warm. In our particular test bed case (a very roomy, server-style enclosure) the heat is manageable. This is in a 5.25" bay with no drives above or below it. Mounted in a 3.5" bay in a crowded mini-tower, however, I’m not so sure. Make no mistake, this drive is definitely hotter than all other ATA drives, save only the larger 9140A.

In conclusion, this Medalist Pro 36530A provides excellent ATA performance with a bit less temperature than than the larger 9.1GB unit we tested. Whether these results are peculiar only to our experiences or translate across Seagate’s line in general are undetermined. In any case, the drive still runs significantly warmer than other ATA offerings. The 7200rpm Medalist Pro series –does- provide top Windows 95 performance if you can’t justify the transition to the highest-speed SCSI drives. Under NT, however, the DiamondMax meets and often exceeds the MP’s performance, and at a lower price to boot. This, again, while running quite cool.