TR Forums

Waco wrote:

The end.

if only, my friend

DancinJack wrote:

Waco wrote:

The end.

if only, my friend

I think we're done here. One can hope we are, anyway

I upset Glorious by stating that the dispute between him and Krogoth was of no value. I did not mean of no value whatsoever, I meant, of no value to the thread, since it's off-topic and unrelated to the Meerkt's questions that raised the issue of spindle speeds, as well as unrelated to the thread title. It's just another off-topic internet argument and there's rarely much merit for anyone involved in pursuing those to the messy conclusion.

Exactly. The ironic part is that those dated tests prove that mainstream units have pretty much caught up outside of special cases. The trend started in 2006 and "fast" 7200RPM units eventually caught up around the same time as last generation of Raptor units.

I don't understand why Glorious is getting all pedantic about some silly rusty media and pulls out dated charts to prove some equally silly point. Take a chill pill for goodness sake.

FYI, I had a second generation 74GB Raptor back in its heyday and retired it because mainstream 7200RPM units caught up for my usage patterns and offer greater capacities.

DancinJack wrote:

Waco wrote:

The end.

if only, my friend

Wishful thinking on my part.

Krogoth wrote:

Exactly. The ironic part is that those dated tests prove that mainstream units have pretty much caught up outside of special cases.

Except that they never really did. Even the ones on the market today are still faster (significantly) than 7200 RPM drives in streaming throughput.

Anyway - can we drop it now? They can't really go slower for 3.5" drives that are helium filled because the heads will smack the platters. 5400/5900 RPM drives are pretty much going to be it until we go with a different form factor or a different filling gas. The benefits of higher RPM drives (even 7200 RPM) are relatively moot for most consumers.

Krogoth wrote:

The ironic part is that those dated tests prove that mainstream units have pretty much caught up outside of special cases.

Because Western digital hasn't released a new one in *SEVEN YEARS*

The DEC Alpha wasn't faster because, now, in 2019, the I7-9700k or whatever utterly annihilates it?

Huh?

Krogoth wrote:

The trend started in 2006 and "fast" 7200RPM units eventually caught up around the same time as last generation of Raptor units.

No. You aren't making an argument, just the same assertion. And it's wrong, as I have demonstrated and as many informed people have told you.

Krogoth wrote:

I don't understand why Glorious is getting all pedantic about some silly rusty media and pulls out dated charts to prove some equally silly point

You posted, incorrectly, about "silly rusty media" before I did.

You're only saying it's "silly" now because you're wrong, as usual, and are unable to admit it, as usual.

Krogoth wrote:

Take a chill pill for goodness sake.

I wasn't the one who revived a discussion that *THREE* posters declared dead so I could continue a ridiculous LARP.

Especially not to revive a personal conflict with another poster as if that declaration of "JUST LET IT BE FINISHED" was a vindication of my dispute with that poster.

Wow.

----

Sorry if it hurts your enormous pride, but no, I'm not going to stop. I'm tired of you continually making stuff up ENTIRELY and then refusing to back down in the slightest.

I came here 17 years ago because they were interested in "reproducible results" and "in-depth investigations".

If that's "silly" and "dated" to you, your loss.

But yes, ironically, your point of view is winning. This site has had troubles for years now because, much like you, a lot of people have either moved on or just don't value it anymore.

So maybe in a few more years I will "chill out" as this site fades away. At that point, you know what? You're not going to find me on youtube disputing your nonsense, and let's be honest, the comments there are more in your stride anyway.

Enjoy!

I TOLD YOU GUYS.

I AM CLAIRVOYANT

Chrispy_ wrote:

I upset Glorious by stating that the dispute between him and Krogoth was of no value. I did not mean of no value whatsoever, I meant, of no value to the thread, since it's off-topic and unrelated to the Meerkt's questions that raised the issue of spindle speeds, as well as unrelated to the thread title. It's just another off-topic internet argument and there's rarely much merit for anyone involved in pursuing those to the messy conclusion.

Honestly, what upset me more was the "Willy-waving" remark.

I know it's an article of faith among some that "GLORIOUS JUST HAS TO BE RIGHT" but that's simply not true. At least not in that sense.

I can, and will, shutdown absolutely if someone demonstrates that I am mistaken.

It's not hard to do, Waco even did it once (actually twice in the same conversation). I was absolutely wrong about something, and he was absolutely right. I admitted that immediately and the only remaining discussion was me trying to self-discover how I got the contrary idea.

Because I care if I am *CORRECT*.

So, yes, in a sense I *DO* have to be right, that is, if I am *WRONG* I admit that and try to understand why I was mistaken.


---

I don't get into it with Krogoth (and others) because I just hate Krogoth. There are times in which I agree with Krogoth, and say so strongly.

I don't do this because I enjoy "messy conclusions" and "off-topic" strife. I do this because this place isn't very useful to me (and others in my admittedly less-than-humble opinion) if what I read on it is just LARPing nonsense where people just make up random and absurd BS for reasons I can't even be bothered to understand.

I can get that ANYWHERE else, and in fact, that mostly *IS* what I get EVERYWHERE else.

EVERYWHERE ELSE is currently winning, -NOT- me. I'd lose EVERY ARGUMENT I'VE HAD in the most embarrassing way possible if I could only maintain the informed discourse and intelligent opinion we (mostly) have here.


You don't have to believe it, but that's my plea.

I wasn't talking about 10K+, only consumer 7200 vs 5400.
Assuming same platter density (and no sector interleave ) a 7200's STR must be 5400 + 33%.

At 1.33TB/platter there are both 5400 and 7200 models from everyone (Toshiba's only 7200, and for them rml527 calls it 1.3-1.5TB/platter for some reason).
Next up is 2TB/platter, only SMR, relatively new, and also there you have 7200rpm.

So until now it doesn't seem like 7200rpm necessitates lower densities.

Usacomp2k3 wrote:

Would there be value in a lower spindle speed drive? Would that allow for more density?

The next best thing: SMR.

I should probably also clarify something else:

Krogoth, you are an institution here. In a good way. Honestly, I'm not. I'm not part of the culture here in the way you are.

Even if I didn't think that, my (negative) character judgment is not be trusted. I don't need to be right, and here I am arguing that in some things I'm more than likely wrong.

For instance: Sweatshopking is an institution here too. He even contributes to the site.

I once thought he should be banned FOR-EVAR. PLEASE GOD let it be so!

That would have been a huge mistake.

So if I think that I am disparaging you personally, look, I'm even telling you to disregard whatever I say. I'm probably wrong.

---

So, on that note, let's have a NON "messy conclusion" to this.

OK?

meerkt wrote:

I wasn't talking about 10K+, only consumer 7200 vs 5400.
Assuming same platter density (and no sector interleave ) a 7200's STR must be 5400 + 33%.

At 1.33TB/platter there are both 5400 and 7200 models from everyone

There are so many other variables, spindle speed isn't even listed any more on e.g. the Seagate 3.5" consumer PC (non-NAS, non-Surveillance, non-Prosumer, non-SSHD) drives webpage.

The spec sheet for these comingles the 2 RPM categories, and also states that the slower ones have a reasonaly-close STR of 185-190 MB/s while the faster are 210-220 MB/s.

Personally, I'm more interested if there's a decent warranty period, UBER, POH, and TB/Y listed...

Official "up to" specs are difficult to correlate with real performance.
Yeah, for some models they've been hiding RPM figures, and it's up to people to investigate and collect that info.

What other variables might affect sequential speed other than density and rpm?

meerkt wrote:

Assuming same platter density (and no sector interleave ) a 7200's STR must be 5400 + 33%.

Let's just stipulate that such a situation is true.

At this point, most consumers are (or at least SHOULD) be getting SSDs. If you have *one* drive, that should be it. Especially if that one drive is now 512GB or 1TB. For most consumers that's "enough".

If you are a consumer with *two* drives, that second one is almost certainly external.

And if it is external, these days it's increasing difficult to figure out what the RPM even is. Often it is variable these days too. So consumers clearly don't care (because they aren't even MARKETED this way!), but when it comes to power/noise alone, yeah, the OEMs sticking them in enclosures probably do.

Then let's talk about people here, like me, who have 12 spinning INTERNAL (that is, bare-drive purchases).

Guess what? At this point I'd prefer 5400 RPM ones. Less power. The performance for my use case almost doesn't matter, and if I didn't have a 10G point-to-point connection for NASing it literally wouldn't matter at all: above 110MB/s I would get zero benefit in any of my scenarios.

The last drives I bought, for instance, were bus powered EXTERNAL 2.5 drives.

So, where is the market for 7200rpm drives?

Regular consumers either don't use them at all or use them as external backup drives and have no idea and therefore don't care (I mean, is it noisy? Hot? Like to fail? ---They care. How fast? They're probably copying music, photos, documents---don't really benefit and aren't really going to notice "33%" longer anyway)

Prosumers, like me, I'd rather just have the 5400rpm/5900rpm drives.

See?

Glorious wrote:

So, where is the market for 7200rpm drives?

Improved Raid rebuild speeds. 200MB/s is way better than 120MB/s. I presume you're using some form of redundancy with those 12 drives? What's your rebuild strategy?

Glorious, I'm all for choice. For me it's 7200rpm. The bottleneck to actually accessing all your data is already there, no need to make it worse. So it would be nice if 7200rpm remained an option, and too bad it's partially becoming upmarket.

LAN is one thing (2.5 - 5Gbps is starting to happen, BTW), but there's also SATA and USB3.

I am all for choice too, I am just explaining the market.

Descriptive, not normative.

I'll talk about my array stuff later but the short answer is that I am not doing this commercially (or discussing this in that context) so my response is: what do I do? Wait!

meerkt wrote:

The next best thing: SMR.

Unfortunately packing tracks tighter together does not increase linear areal density, therefor it grants you nothing in terms of head/platter transfer speeds.

SMR is difficult to recommend outside of archival or multimedia streaming usage patterns.

Krogoth wrote:

SMR is difficult to recommend outside of archival or multimedia streaming usage patterns.

Given that you generally can't buy SMR drives any more, I'm not sure that's much of a concern at the moment.

Waco wrote:

Krogoth wrote:

SMR is difficult to recommend outside of archival or multimedia streaming usage patterns.

Given that you generally can't buy SMR drives any more, I'm not sure that's much of a concern at the moment.

I wouldn't count out SMR drives being unobtainium just yet.

Krogoth wrote:

I wouldn't count out SMR drives being unobtainium just yet.

No manufacturer is making them in drive-managed form (and they are unlikely to do so again). They haven't for quite a while.

Care to add any substance to your claim?

Waco wrote:

Krogoth wrote:

I wouldn't count out SMR drives being unobtainium just yet.

No manufacturer is making them in drive-managed form (and they are unlikely to do so again). They haven't for quite a while.

Care to add any substance to your claim?

Backup plan just in case HAMR/MAMR don't pan out for whatever reason or as a cheaper to make "alternative" for bulk storage units.

Krogoth wrote:

Backup plan just in case HAMR/MAMR don't pan out for whatever reason or as a cheaper to make "alternative" for bulk storage units.

I didn't say that they wouldn't be making SMR drives.

I said they wouldn't be making drive-managed drives (like they aren't today). Go Google that, then let's talk.

Waco wrote:

Krogoth wrote:

I wouldn't count out SMR drives being unobtainium just yet.

No manufacturer is making them in drive-managed form (and they are unlikely to do so again). They haven't for quite a while.

Are you sure about that? And what is "quite a while"? The ST4000DM004 drives I bought last summer certainly act like they're SMR. Maybe they were "new old stock"? Will take a look at the manufacture date when I get home.

just brew it! wrote:

Are you sure about that? And what is "quite a while"? The ST4000DM004 drives I bought last summer certainly act like they're SMR. Maybe they were "new old stock"? Will take a look at the manufacture date when I get home.

There are backlogs of drives in the channels, but Seagate hasn't been making those drives (the M004 line of SMR drives) since early last year.

For new large purchases they've been EOS for almost 2 years. There are no new models with drive-managed shingles.

Waco wrote:

just brew it! wrote:

Are you sure about that? And what is "quite a while"? The ST4000DM004 drives I bought last summer certainly act like they're SMR. Maybe they were "new old stock"? Will take a look at the manufacture date when I get home.

There are backlogs of drives in the channels, but Seagate hasn't been making those drives (the M004 line of SMR drives) since early last year.

For new large purchases they've been EOS for almost 2 years. There are no new models with drive-managed shingles.

Must be a huge backlog. Amazon is still selling them: https://www.amazon.com/Seagate-Barracud ... B071WLPRHN

just brew it! wrote:

Must be a huge backlog. Amazon is still selling them: https://www.amazon.com/Seagate-Barracud ... B071WLPRHN

Wow. I thought most people were getting screwed by pulling these from external enclosures, I didn't know they were still on the market as "new" drives with zero verbiage about them being shingled!

Anything ending in M004 is shingled AFAIK from Seagate.

So, what's the deal right now?

I had been buying HGST drives for anything 3-4TB and up, but I have read/heard some stories about WDC mucking that up. What's everyone suggest for an internal 8-10TB drive right now? It'd likely be in a PC or NAS for its entire lifetime being utilized as a media serving drive.

HGST and Seagate are my goto companies for large drives. They don't make SMR drives over 8 TB, so you're at least safe from that debacle.

just brew it! wrote:

Waco wrote:

just brew it! wrote:

Are you sure about that? And what is "quite a while"? The ST4000DM004 drives I bought last summer certainly act like they're SMR. Maybe they were "new old stock"? Will take a look at the manufacture date when I get home.

There are backlogs of drives in the channels, but Seagate hasn't been making those drives (the M004 line of SMR drives) since early last year.

For new large purchases they've been EOS for almost 2 years. There are no new models with drive-managed shingles.

Must be a huge backlog. Amazon is still selling them: https://www.amazon.com/Seagate-Barracud ... B071WLPRHN

FWIW the ones I bought were manufactured in April of 2018. So they were still producing them as of early Q2, at least. Good riddance, these things are a dumpster fire!

DancinJack wrote:

What's everyone suggest for an internal 8-10TB drive right now? It'd likely be in a PC or NAS for its entire lifetime being utilized as a media serving drive.

Seagate, sadly. Exos 7E8.

WD: has mucked up the HGST portfolio and comingled, some Ultrastars are HGST lineage derivatives while other ones are relabelled WD Golds; UBER has gone down to consumer-tier 10^-14; spec sheets are now inadequate. A hot mess at the mo.

Toshiba: compared to the other 2 bigger players, the warranty is more annoying to deal with, and the spec sheets' claimed reliability (AFR/MTBF) is 1/3 lower which doesn't inspire confidence.