HomeAncelus TopicsAncelus Basics

Ancelus Hybrid Memory/Disk Database

The Ancelus Revolution has Started!

Ancelus introduces the first new theory of data management in 50 years.  It's the patented third-generation system from the team that invented the in-memory database in 1986. Now this hybrid memory/disk system delivers the first ACID compliant database to combine extreme speed, complexity handling, scalability and operational simplicity. 

Combined with Threaded Query Language (TQL), Ancelus supports both rapid development and extreme execution speed. 

Big Data is overhauled with embedded executable procedures, streaming statistics, multiple live schemas and dynamic revision.  A3:Ancelus Adaptive Analytics changes how data scientists will develop, detect and deliver the intelligence deeply obscured in the tsunami of raw data from the Internet of Things.  And the emerging Big Data crisis is averted. 

The bottom line is reduced first cost and operating cost.

 Bloor Research White Paper  See what London-based Bloor Research had to say about Ancelus.  Hint "...these are staggering [benchmark] numbers."

Now we've updated those numbers on the Intel Broadwell server.  Every other data management method is obsolete obsolete.  Latest Benchmarks.

Speed:  Speed = $$$.  The faster your database, the bigger the savings.  Unless you bailed out an old, bloated system by stacking servers 10 deep.  If you believe your current technology stack is fast enough, you've just decided to fund technology bloat.

Ancelus is the fastest available.  Period.  Orders of magnitude faster than anything else on the market.  Constant read/write latency under 100 nanosecond regardless of size (up to 16 EB). Stream into 32 fields at 256 million transactions per second on a single 4 socket server.  Table joins that require hours or days in other systems now respond in microseconds.

Complexity Tamed: Multi-key tables.  Multiple schema's in a single database. Nested lists (block chain format), recursive lists are the native structure of Ancelus. Integrated multiple views.  If you can draw the entity relationship diagram, Ancelus accepts it directly.  No transforms or performance compromise.  Temporary schema's allow experimental analysis without reloading data. 

Throughput - Table Joins: 44 concurrent 7-table-joins in a 271 GB database in 11 seconds.  10 billion TPM on a single 2U server.  See the latest benchmarks for details.  No one in the industry has even attempted this feat.

Search: Billion row search, 2.4 microsecond. Add ~500 nanoSec for each 10X DB size. Binary tree always balanced.

Stream Indexing: All indexing is updated on every transaction.  No more re-indexing downtime.

Non-Stop Operation: ZERO downtime for schema changes, re-index, binary tree re-balance, column compress, data refresh, data archive, or extending the database.

Scale: Ancelus offers a single logical image up to 16 EB regardless of the hardware stack. No synchronization challenge when you have "multiple versions of the truth." 

Lean:  No data duplication or sparse matrix. Reduces dataset size by up to 20X. True dataspace utilization of 85% vs. 4% typical for traditional systems, 2% for big data.

Fast Development:  Eliminates 15 DBA tasks including normalization.  Ancelus is 100% normalized and 100% de-normalized at the same time.

ACID Compliance: Only Ancelus delivers "Always-On" ACID compliance.  No option to turn it off.  All benchmarks are ACID compliant.

True Real-Time: It's not only how fast you can insert.  It's how soon you can query!  100 nanoseconds is a lot faster than you can blink.

Referential Integrity Checking: Always on, no option to turn it off.  All benchmarks are quoted with referential integrity on.

Reliability: 24x800 high-volume manufacturing with zero faults. Telecom engineer: "Carrier grade software runs forever."

Security: Unique random key for each data element, no stored key (some performance penalty).  Field level credential control.

Operational and Analytics in a single system: Stream in and complex query at the same time.  Stream and query against different schemas.


Big Data no longer requires retro solutions!


47 light yard







Go to top View Our Stats