Tag: Performance

Posted on

20x Faster Time to First Token: The HPE Alletra X10000 Edge for AI

In March 2026, HPE, NVIDIA, Kamiwaza and Signal65 published a paper (check here and here) showing  about a 20x acceleration for both time-to-first-token (TTFT) and token generation rate, using an HPE Alletra X10000 to store the KV cache. The storage system used S3 over RDMA to achieve this (and is in fact the first NVIDIA-certified object storage system).

The test has, as of the date of this writing, the most complete benchmark disclosure of all the KV Cache competitive tests I could find. It also consists of an extremely hard workload with high concurrency.

I will explain why all this is important, why it’s different from competitor numbers, plus provide some insights about what this means regarding overall system efficiency.

Because the goal isn’t just to keep GPUs busy. It’s to keep them busy generating new stuff, not recalculating old stuff.

The benefits with this solution are numerous:

  • Far more workload becomes possible but also…
  • One could approach it as a lot less infrastructure is needed due to far more efficient use of the hardware, which means…
  • Lower power and rackspace requirements, which all leads to…
  • Lower Watt/token and lower $/token.
Continue reading “20x Faster Time to First Token: The HPE Alletra X10000 Edge for AI”
Posted on

The Architectural Benefits of HPE Alletra MP – Plus R4 Coolness

When we first released the new HPE Alletra MP platforms, I wrote a few articles going over the benefits and how the flexible new hardware platform manifests into different “personalities” for high end block and file solutions.

This time I want to take a deeper dive into the architectural benefits of our approach and how the new R4 software for Alletra MP Block enables certain things no other vendor can come close to – plus give a taste of what may be possible in the future given the amazing flexibility of the underlying architecture (it’s a blog, I can’t provide roadmaps here).

I will cover things like fractional multi-dimensional scaling (that allows things impossible with other vendors like adding a single controller node without needing to add capacity) but also resiliency in the face of simultaneous failures that would cripple all other storage systems I’m aware of. It’s not meant to be a comprehensive coverage of everything, but hopefully enough to give you a taste.

Let’s go!

Continue reading “The Architectural Benefits of HPE Alletra MP – Plus R4 Coolness”
Posted on

HPE GreenLake for File Storage

A critical part of the recent April 4th, 2023 announcements from HPE Storage was the scale-out HPE GreenLake for File Storage.

For the foundational piece explaining the common hardware between the various offerings please go here. For the Block storage piece, here.

The new HPE File offering is based on the HPE Alletra Storage MP hardware, and uses a common management interface for both File and Block, providing a seamless, centralized, multiprotocol management experience.

For the people that like looking at boxes, a small one would look like this:

A Small HPE GreenLake for File Storage System – Compute Separate from Capacity
Continue reading “HPE GreenLake for File Storage”
Posted on

Beware of Cloud Sizing Tools and avoid Reliability Angst

Some time ago I wrote about the dangers of taking certain things for granted with new technologies.

This time I wanted to use a more specific enterprise application example to show that customers need to be extra careful when comparing solutions, especially for mission-critical apps.

Sometimes being too high-level means missing the unspeakable horrors lurking under the covers. And ignorance doesn’t mean bliss… just nasty surprises.

To summarize: Avoid bait-and-switch so you avoid surprise costs and pain.

  • Ensure all components in any sizing tools reflect your business requirements.
  • The simpler the infrastructure, the more reliable. If one must do things like stripe across many volumes in order to get decent performance even for medium-sized solutions, then that may be a warning sign that the solution is lacking.
  • Ensure all the underlying components in any pricing you see would fit your company’s mission-critical needs. For instance – what reliability and resiliency are the storage components rated for? And is that sufficient for your needs?
  • Ensure you are accounting for the right number of systems (Production-spec vs non, times number of applications, etc). This can quickly add up with certain apps.

Continue reading “Beware of Cloud Sizing Tools and avoid Reliability Angst”

Posted on

HPE Alletra 9000 – Primera Evolved

I’m excited to announce that the evolution of the HPE Primera (which was the evolution of 3PAR) is now available.

It’s called the HPE Alletra 9000 and is the mission-critical Tier-0 complement to the Tier-1 Alletra 6000 (which in turn is the evolution of Nimble).

It retains the rich feature set of Primera and the 100% uptime guarantee. The main enhancement vs Primera is the increased speeds, and the fact that all the performance is possible in just a 4U configuration, making it the most performance-dense full-feature Tier 0 system in the world (by far). It is managed via the HPE Data Services Cloud Console.

A welcome enhancement (that is also coming to Primera) is that of Active Peer Persistence, which allows a LUN to be simultaneously read from and written to from two sites synchronously replicating. This means that each site can do local writes to a sync replicated LUN without the hosts needing to cross the network to the other site.

Continue reading “HPE Alletra 9000 – Primera Evolved”
Posted on

HPE Alletra 6000 – Nimble Evolved

HPE Alletra

I’m very excited to announce that the evolution of HPE Nimble Storage is now available.

The new line is called Alletra 6000, and is completely interoperable with Nimble and manageable both from its own GUI/CLI/API and Data Services Cloud Console. For the Alletra 9000 (the evolution of Primera) there will be a separate post.

All the usual goodies of Nimble are still there (6-nines guaranteed uptime, fancy direct-to-L3 support, InfoSight with infrastructure AI recommendations, 100% headroom even if running on one controller, SCM cache, Triple+ RAID, Cascade Multistage Checksums etc). What’s different is mostly the vastly increased speeds in real-world workloads, and the shorter form factor.

Continue reading “HPE Alletra 6000 – Nimble Evolved”
Posted on

The Harsh Realities of PCIe Lane Shortage in Storage Systems

There are a lot of myths and misinformation, plus more than a modicum of misunderstanding, regarding how storage systems can use available bandwidth, especially with certain newer kinds of media.

I wanted to explain some of the harsh facts of storage system design in the real world, and why one shouldn’t just add up drive speeds to estimate performance.

Continue reading “The Harsh Realities of PCIe Lane Shortage in Storage Systems”

Posted on

When Terrified Vendors Attack: The Dell Edition

It recently came to my attention that Dell is now advertising some kind of benchmark that shows one of their platforms can be faster than Nimble in some very specific test of their own concoction.

While I don’t doubt that’s possible (indeed, we could do it the other way around), it may be worthwhile investigating what’s prompting the attack.

I also wanted to point out the various technically fishy points of the benchmark.

Continue reading “When Terrified Vendors Attack: The Dell Edition”

Posted on

Progress Needs Trailblazers

I got the idea for this stream-of-consciousness (and possibly too obvious) post after reading some comments regarding new ways to do high speed I/O. Not something boring like faster media or protocols, but rather far more exotic approaches that require drastically rewriting applications to get the maximum benefits of radically different architectures.

The comments, in essence, stated that such advancements would be an utter waste of time since 99% of customers have absolutely no need for such exotica, but rather just need cheap, easy, reliable stuff, and can barely afford to buy infrastructure to begin with, let alone custom-code applications to get every last microsecond of latency out of gear.

Continue reading “Progress Needs Trailblazers”

Posted on

Practical Considerations for Implementing NVMe Storage

Before we begin, something needs to be clear: Although dual-ported NVMe drives are not yet cost effective, the architecture of Nimble Storage is NVMe-ready today. And always remember that in order to get good benefits from NVMe, one needs to implement it all the way from the client. Doing NVMe only at the array isn’t as effective.

Continue reading “Practical Considerations for Implementing NVMe Storage”

Posted on

The Well-Behaved Storage System: Automatic Noisy Neighbor Avoidance

This topic is very near and dear to me, and is one of the big reasons I came over to Nimble Storage.

I’ve always believed that storage systems should behave gracefully and predictably under pressure. Automatically. Even under complex and difficult situations.

It sounds like a simple request and it makes a whole lot of sense, but very few storage systems out there actually behave this way. This creates business challenges and increases risk and OpEx.

Continue reading “The Well-Behaved Storage System: Automatic Noisy Neighbor Avoidance”

Posted on

Proper Testing vs Real World Testing

The idea for this article came from seeing various people attempt product testing. Though I thought about storage when writing this, the ideas apply to most industries.

Three different kinds of testing

There are really three different kinds of testing.

The first is the incomplete, improper, useless in almost any situation testing. Typically done by people that have little training on the subject. Almost always ends in misleading results and is arguably dangerous, especially if used to make purchasing decisions.

The second is what’s affectionately and romantically called “Real World Testing”. Typically done by people that will try to simulate some kind of workload they believe they encounter in their environment, or use part of their environment to do the testing. Much more accurate than the first kind, if done right. Usually the workload is decided arbitrarily 🙂

The third and last kind is what I term “Proper Testing”. This is done by professionals (that usually do this type of testing for a living) that understand how complex testing for a broad range of conditions needs to be done. It’s really hard to do, but pays amazing dividends if done thoroughly.

Let’s go over the three kinds in more details, with some examples.

Continue reading “Proper Testing vs Real World Testing”

Posted on

NetApp Enterprise Grade Flash

June 23rd marked the release of new NetApp Enterprise Flash technology. The release consists of:

  • New ONTAP 8.3.1 with significant performance, feature and usability enhancements
  • New Inline Storage Efficiencies
  • New All-Flash FAS systems (AFF works only with SSDs)
  • New, aggressive pricing model
  • New maintenance options (price protection on warranty for up to 7 years, even if you buy less warranty up front)

Continue reading “NetApp Enterprise Grade Flash”