Intel Optane 900P

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

beginner99

Diamond Member
Jun 2, 2009
5,314
1,756
136
It isn't just about having insane low QD performance.

I think the reason Star Citizen shows any benefit is because the game and its world are so massive. I can't see many applications falling into that category. Even then, there's likely fixed time paths that can't be sped up.

I would not buy this drive for game loading times but more for responsiveness. Even flash-based ssds can have some minor stutters especially when in steady-state. But yeah the value for the price is very questionable. I would probably invest the money in a larger flash based drive than this.
 

coercitiv

Diamond Member
Jan 24, 2014
7,164
16,693
136
Even flash-based ssds can have some minor stutters especially when in steady-state. But yeah the value for the price is very questionable. I would probably invest the money in a larger flash based drive than this.
Steady-state performance is not relevant when comparing products with such high price disparity: one simply sets 10-20% space provisioning on a high end SSD and steady-state performance is no longer a factor, while price still is even with the somewhat higher price/GB induced by provisioning.

Also, in relation with the Star Citizen test, I'd like to see Intel comparing themselves with the 960 Pro drives in tests, not with the EVO. Assuming the Pro performed in a similar fashion with the far less expensive EVO, what good does it do to Intel to compare their expensive flagship drive with the more value oriented product of the competition?
 

beginner99

Diamond Member
Jun 2, 2009
5,314
1,756
136
Steady-state performance is not relevant when comparing products with such high price disparity: one simply sets 10-20% space provisioning on a high end SSD and steady-state performance is no longer a factor, while price still is even with the somewhat higher price/GB induced by provisioning.

Show me that review that proofs this. Over-provisioning only increases steady-state performance which is still a lot slower than the performance showed in all the benches. AT used to have nice charts showing them but for some reason they got dropped / not tested anymore. I wonder why? Pressure from manufacturers? Because the steady-state performance is what matters for consumers. in fact all benches should only be done in steady-state. That is the actual performance you will have for like 95% of the drives life-time. An ssd with 20% OP will still be sub 20k IOPS compared to 500k for the optane drive. and if you over-provision by 20% your ssd drive is now 20% more expensive per GB.

But in the end of course you need to have the use-case for the optane drive to make it worth it's price. Even we as enthusiasts mostly won't feel a huge difference at least not one worth the price.
 

Billy Tallis

Senior member
Aug 4, 2015
293
146
116
AT used to have nice charts showing them but for some reason they got dropped / not tested anymore. I wonder why? Pressure from manufacturers? Because the steady-state performance is what matters for consumers. in fact all benches should only be done in steady-state.

No, consumer benchmarks mostly should not be done at steady state. Real-world consumer usage involves drives that aren't 100% full and do get TRIM commands and that get tons of idle time, so their SLC caches actually help and the drives aren't experiencing any thermal throttling. Steady-state tests are an unrealistic worst-case scenario. They're relevant only as a lower bound on performance or as a test of a consumer drive's suitability for enterprise workloads. They can offer an interesting window into the drive's garbage collection behavior, but no longer one that is representative of how the drive behaves in typical real-world usage.
 
Reactions: coercitiv

beginner99

Diamond Member
Jun 2, 2009
5,314
1,756
136
No, consumer benchmarks mostly should not be done at steady state. Real-world consumer usage involves drives that aren't 100% full and do get TRIM commands and that get tons of idle time, so their SLC caches actually help and the drives aren't experiencing any thermal throttling. Steady-state tests are an unrealistic worst-case scenario. They're relevant only as a lower bound on performance or as a test of a consumer drive's suitability for enterprise workloads. They can offer an interesting window into the drive's garbage collection behavior, but no longer one that is representative of how the drive behaves in typical real-world usage.

OK I see. I have to admit that I meant something different with steady-state, namely once the drive has been full written to, eg all flash cells which doesn't mean the drive is full. This comes from initial ssd days when TRIM wasn't always working correctly. Some people say depending on implementation and drive it still doesn't work, eg. some drives don't actually clear the cells. And such a cell that has been written to needs first be cleared hence making the drive slower in this state but of course faster than being filled 100%.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
Real-world consumer usage involves drives that aren't 100% ful

No SSD drives are truly full because there are parts you can't access due to overprovisioning.

Some people say depending on implementation and drive it still doesn't work, eg. some drives don't actually clear the cells. And such a cell that has been written to needs first be cleared hence making the drive slower in this state but of course faster than being filled 100%.

PCPer has one case of that happening: https://www.pcper.com/reviews/Stora...VMe-HHHL-SSD-Review-Lots-3D-XPoint/Performa-4

The stutters do exist, not just in the scenario mentioned above, but even in general usage. It is indeed a rare occurence though, so the question is whether you want a "pure" storage like 900P where it has virtually no caching. Some people are annoyed by such rare occurences, but some like it just because of almost an ideological reason of having a pure setup.
 

Billy Tallis

Senior member
Aug 4, 2015
293
146
116
No SSD drives are truly full because there are parts you can't access due to overprovisioning.

Steady-state write tests will get you pretty close, by attempting to write faster than the drive can perform garbage collection. That's why performance falls off a cliff shortly after the first full drive write on the way to steady state. That's the point at which the drive no longer has empty blocks and has to start reclaiming partially-invalidated blocks.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
Steady-state write tests will get you pretty close, by attempting to write faster than the drive can perform garbage collection.

I'm aware of this. But because of overprovisioning, it avoids the catastrophic scenario of having it truly full. If 128GB flash had 128GB user space, the performance would be chaos. By having say, 10GB for overprovisioning, a "full" drive means 118/128 full. A 10GB free drive would mean 108/128GB full, which is a lot less smaller gap in performance compared to say 128/128 vs 118/128.
 

Fir

Senior member
Jan 15, 2010
484
194
116
I have the 480GB version (PCI card) and have to say it's the fastest storage device I've had my hands on.
Prior I was using a 960 Pro and even had 3 in RAID0.
This drive is the bees knees. It can go full out pedal to the metal continously without slowing down.
I will get more when the price goes down but for this kind of performance $579 out the door for a 480GB drive isn't bad. (providing you actually can use it!)
When you factor in the need for > 200GB of space using a ramdisk, it's a bargain!
 

dahorns

Senior member
Sep 13, 2013
550
83
91
I have the 480GB version (PCI card) and have to say it's the fastest storage device I've had my hands on.
Prior I was using a 960 Pro and even had 3 in RAID0.
This drive is the bees knees. It can go full out pedal to the metal continously without slowing down.
I will get more when the price goes down but for this kind of performance $579 out the door for a 480GB drive isn't bad. (providing you actually can use it!)
When you factor in the need for > 200GB of space using a ramdisk, it's a bargain!

I'm glad reality is somewhat reflecting the synthetics. That being said, your first sentence is obvious. It is the fastest storage anyone has ever had their hands on. I mean, unless you work at Area 51 and have access to alien storage mediums.

What are you doing with it?
 

Bier667

Member
Oct 31, 2017
35
1
11
What are you doing with it?

Browsing Pr0n like everyone...


TBH in my daily usage i could not tell a difference between my 280GB AIC version and my old 850 Evo SATA with 250GB.

PCM is still the best tech to look forward
 

Fir

Senior member
Jan 15, 2010
484
194
116
I'm glad reality is somewhat reflecting the synthetics. That being said, your first sentence is obvious. It is the fastest storage anyone has ever had their hands on. I mean, unless you work at Area 51 and have access to alien storage mediums.

What are you doing with it?

Data stream scrubbing.
It's a waste to use one of these on a desktop aside from e-peen perhaps.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |