Navigation
Topics Register • News • History • How to • Sequences statistics • Template prototypes

PrimeNet summary report

From Prime-Wiki
Jump to: navigation, search

The PrimeNet summary report, also known as the 'Status report' or 'Work distribution map' shows the GIMPS virtual supercomputer's operating status, automatically updated by the server every hour, at the start of the hour. It also shows the current distribution of exponents.

Above the line

In the section above the horizontal divider are four sections (three tabular data, one graphical).

Aggregate Computing Power

The left table shows the computing power that PrimeNet has observed during the past day, week, and month; and what is the potential power of the active machines. Since not all machines are contributing 24/7 this number is not close to 100%. GHz-days and TFLOPS are used.

Resources Registered

The middle table lists teams (groups of users that have joined together), user ID's (individuals that are or have been contributing to the project), computers (individual machines, including manual assignment machines, that have been registered, not only those currently active {many computers are no longer contributing to the project}), and work units (the sum of all current assignments).

Recently Active and Work Done

Broken down by activity in the past 24 hours, 7 days, and 30 days.

  • user IDs: How many individuals have had contact with PrimeNet (receiving assignments, turning in results, updating assignments, etc.) during the period listed.
  • computers: The count of machines that have contacted PrimeNet during the period.
  • work units: The number of assignments that have been turned in during the period. For the 30 day period, this is higher than the total number in the 'Resources Registered' section, because many assginments take much less than a month.

The Graph

Plotted average for each hour over the last 7 days is the TFLOPS throughput of GIMPS. The vertical axis auto scales and the horizontal axis is oldest to current, left to right.

Below the line

In the section below the horizontal divider is a data table entitled Exponent Status Distribution. The table details the status of exponents of the Mersenne numbers by ranges. It is broken down into 4 sets of vertical columns. The exponents below 10,000,000 are treated slightly differently than those above.

The rows of the table indicate ranges of one million numbers. Each row starts at the number listed in the left-most column and continues up to but excluding the number that starts the next row. The first row is for the numbers 0 - 999,999. The second row is for numbers 1,000,000 - 1,999,999. Etc.

Exponent Range

The section to the left of the first vertical divider is the Exponent Range section and is broken into 3 columns.

  • Start: The lowest potential exponent for the range.
  • Count: The number of regular prime numbers in the range. This is not the number of Mersenne primes in the range. This is the numbers that need to be tested; all others need not be tested, because it is impossible to have a Mersenne prime with an exponent that is not prime (see: Mersenne composite. The values in this column will not change, since it is trivial to find and count all primes in these ranges.
  • P: The number of Mersenne primes found (so far and verified) in this range.

Composite / Status Unproroven

The second section is broken into parts numbers proven composite and those tested, but not proven. The sum of numbers in each row of this section will equal Count less Primes.

  • Composite: The exponents that have been proven to be composite, by one of the following
    • F: Factored. At least one factor has been for these exponents.
    • LL-D: Lucas-Lehmer double checked. These exponents have had two independent LL test performed, with matching residues. Thus they are proven to be composite.
    • Status Unproven: The exponents that have yet to be proven either prime or composite.
    • LL: Lucas-Lehmer tested. These exponents have been tested at least once and don't have a matching residue and had no significant errors durring the test.
    • LLERR: Lucas-Lehmer with error. These exponents have been test at least once and had a significant error of some sort durring the test and don't have a matching residue.
    • NO-LL: No Lucas-Lehmer test has been performed on these exponents.

Assigned

These are the exponents that PrimeNet shows as currently being assigned to various types of work.

Note: some individuals may be doing work on exponents and have not registered them with PrimeNet.

Assignment types/columns:

  • ECM: (For exponents below 10,000,000 only) Elliptic curve method factoring. These exponents have already been proven to be composite. The ECM assignments are handed out solely to either find a factor for known composites with no known factor or to complete the factorization of an exponent.
  • TF: (For exponents 10,000,000 and above, practically for those above 50,000,000) Trial factoring. Attempting to find a factor for exponents with no known factor and have not been LL tested. These include normal TF and LMH-TF assignments.
  • P-1: P-1 factorization method assignments. These exponents have had a reasonable amount of trial factoring done, but no factor has been found nor have they been LL tested (except for those under 10,000,000, those have a known status, but like ECM are being assigned solely to find factors).
  • LL: First time Lucas-Lehmer tests. These also may include re-assigned exponents that had 'bad' errors on the intial LL test.
  • LL-D: Double (or triple, if the residues did not match) checking of exponents. These have been LL tested at least once before.

Available

These are the exponents that are currently un-assigned and ready to be tested. The column headings are identical to those of the preceeding section.

  • ECM: (For exponents below 10,000,000 only) See ECM above.
  • TF: (For exponents 10,000,000 and above, practically for those above 50,000,000) Trial factoring. The first type of work that an exponent under goes. Once an exponent has all of the TF work done on it (the level is determined to best benefit the project), it then will become availble for other work.
  • P-1: As P-1 above. This assignment follows TF. (Currently the sequence is TF to final bit level minus one, P-1, then final bit level TF).
  • LL: These exponents have had all of the requisite TF and P-1 work done, but have not had a first time LL test.
  • LL-D: These have been LL tested once before and are ready to be double checked.