WITH C AS
(
	SELECT *, ROW_NUMBER() OVER(ORDER BY date ASC) rn
          FROM TableName
)
SELECT *
  FROM C A
       JOIN C B
         ON A.rn + 1 = B.rn;

CREATE TABLE [Sample] 
(
	seq_nbr INTEGER NOT NULL PRIMARY KEY
);

INSERT INTO [Sample]
VALUES (1), (5), (7), (8), (15), (50);

SELECT seq_nbr
  FROM [Sample]
 ORDER BY seq_nbr ASC;

/*====== Sample Data =========
seq_nbr
-----------
1
5
7
8
15
50


============ Wanted Result ==========
seq_nbr     seq_nbr
----------- -----------
1           5
5           7
7           8
8           15
15          50
50          NULL

*/

WITH C(seq_nbr, rnk) AS
(
	SELECT seq_nbr,
	       ROW_NUMBER() OVER(ORDER BY seq_nbr ASC)
	  FROM [Sample]
)
SELECT A.seq_nbr, B.seq_nbr
  FROM C AS A 
       LEFT JOIN C AS B
         ON A.rnk + 1 = B.rnk;

SET STATISTICS IO ON
SET STATISTICS TIME ON

-- Ms65g
;WITH C(seq_nbr, rnk) AS
(
	SELECT seq_nbr,
	       ROW_NUMBER() OVER(ORDER BY seq_nbr ASC)
	  FROM [Sample]
)
SELECT A.seq_nbr, B.seq_nbr
  FROM C AS A 
       LEFT JOIN C AS B
         ON A.rnk + 1 = B.rnk;

-- Peso
SELECT		s.seq_nbr,
		f.seq_nbr
FROM		dbo.[Sample] AS s
OUTER APPLY	(
			SELECT TOP(1)	w.seq_nbr
			FROM		dbo.[Sample] AS w
			WHERE		w.seq_nbr > s.seq_nbr
			ORDER BY	w.seq_nbr
		) AS f

SET STATISTICS TIME OFF
SET STATISTICS IO OFF

SET STATISTICS IO ON
SET STATISTICS TIME ON

-- Ms65g 1
;WITH C(seq_nbr, rnk) AS
(
	SELECT seq_nbr,
	       ROW_NUMBER() OVER(ORDER BY seq_nbr ASC)
	  FROM [Sample]
)
SELECT A.seq_nbr, B.seq_nbr
  FROM C AS A 
       LEFT JOIN C AS B
         ON A.rnk + 1 = B.rnk;

-- Peso 1
SELECT		s.seq_nbr,
		f.seq_nbr
FROM		dbo.[Sample] AS s
OUTER APPLY	(
			SELECT TOP(1)	w.seq_nbr
			FROM		dbo.[Sample] AS w
			WHERE		w.seq_nbr > s.seq_nbr
			ORDER BY	w.seq_nbr
		) AS f;

-- Peso 2
WITH C(seq_nbr, rnk) AS 
(
    SELECT	seq_nbr,
		ROW_NUMBER() OVER(ORDER BY seq_nbr)
    FROM	dbo.[Sample]
) 
SELECT		MAX(CASE WHEN flag = 1 THEN seq_nbr END) AS seq_nbr,
		MAX(CASE WHEN flag = 2 THEN seq_nbr END) AS seq_nbr
FROM		(
			SELECT		seq_nbr, 
					rnk + 1 AS grp_fct,
					1 AS flag
			FROM		C
			UNION ALL
			SELECT		seq_nbr,
					rnk,
					2
			FROM		C
		    ) AS D
GROUP BY	grp_fct
HAVING		MAX(CASE WHEN flag = 1 THEN seq_nbr END) IS NOT NULL;

--Ms65g 2
;WITH C AS
(
SELECT *, 
       ROW_NUMBER() OVER(ORDER BY seq_nbr ASC) AS rownum
  FROM [Sample]
)
SELECT MAX(CASE WHEN k = 1 THEN seq_nbr END) AS seq_nbr,
       MAX(CASE WHEN k = 2 THEN seq_nbr END) AS seq_nbr
  FROM C
    CROSS JOIN
      (VALUES (1), (2)
      ) AS D(k)
GROUP BY (CASE k WHEN 1 THEN rownum + 1
                 WHEN 2 THEN rownum
          END)
HAVING MAX(CASE WHEN k = 1 THEN seq_nbr END) IS NOT NULL;


SET STATISTICS TIME OFF
SET STATISTICS IO OFF

/*
====== Ms65g 1
Table 'Sample'. Scan count 2, logical reads 15

====== Peso 1
Table 'Sample'. Scan count 7, logical reads 14

====== Peso 2
Table 'Sample'. Scan count 2, logical reads 4

====== Ms65g
Table 'Sample'. Scan count 1, logical reads 2
*/

quote:

---

Originally posted by ms65g

As you see my second solution is most efficient and needs half of your best number of reads.

---

;WITH cte
AS (
	SELECT		seq_nbr,
			d,
			ROW_NUMBER() OVER (ORDER BY seq_nbr) / 2 AS theGrp
	FROM		[Sample]
	CROSS JOIN	(
				VALUES	(0),
					(1)
			) AS x(d)
)
SELECT		MIN(seq_nbr) AS col1,
		MAX(seq_nbr) AS col2
FROM		cte
GROUP BY	theGrp
HAVING		MAX(d) = 1

-- Peso 2 (my own version)
Table 'Worktable'. Scan count 0, logical reads 0, physical reads 0.
Table 'Sample'. Scan count 1, logical reads 7, physical reads 0.

 SQL Server Execution Times:
   CPU time = 0 ms,  elapsed time = 53 ms.


-- Ms65g #2
Table 'Worktable'. Scan count 0, logical reads 0, physical reads 0.
Table 'Sample'. Scan count 1, logical reads 14, physical reads 0.

 SQL Server Execution Times:
   CPU time = 0 ms,  elapsed time = 69 ms.

;WITH C1 AS
(
SELECT *, 
       ROW_NUMBER() OVER(ORDER BY seq_nbr ASC) AS rownum
  FROM [Sample]
  
), C2 AS
(SELECT (CASE k WHEN 1 THEN rownum + 1
                 WHEN 2 THEN rownum 
          END) grp_fct, seq_nbr
  FROM C1
      CROSS JOIN
      (VALUES (1), (2)
      ) AS D(k)
)
SELECT MIN(seq_nbr) AS seq_nbr,
       MAX(seq_nbr) AS seq_nbr
  FROM C2
GROUP BY grp_fct
HAVING grp_fct > 1;

Peso 2#
Table 'Sample'. Scan count 1, logical reads 10
Query Cost: 83%

Ms65g 2#
Table 'Sample'. Scan count 1, logical reads 10
Query Cost: 17%

;WITH C AS
(
SELECT seq_nbr, 
       ROW_NUMBER() OVER(ORDER BY seq_nbr ASC) + k AS grp_fct
  FROM [Sample]
        CROSS JOIN
        (VALUES (0), (1)
        ) AS D(k)
)
SELECT MIN(seq_nbr) AS seq_nbr,
       MAX(seq_nbr) AS seq_nbr
  FROM C
GROUP BY grp_fct
HAVING grp_fct > 1;